The Past, Present, and Future of End-Of-Life Care

CHILDREN AND FAMILIES The RAND Corporation is a nonprofit institution that helps improve policy and EDUCATION AND THE ARTS decisionmaking through research and analysis. ENERGY AND ENVIRONMENT

HEALTH AND HEALTH CARE This electronic document was made available from www.rand.org as a public service

INFRASTRUCTURE AND of the RAND Corporation. TRANSPORTATION

INTERNATIONAL AFFAIRS LAW AND BUSINESS Skip all front matter: Jump to Page 16 NATIONAL SECURITY

POPULATION AND AGING

PUBLIC SAFETY Support RAND

SCIENCE AND TECHNOLOGY Browse Reports & Bookstore

TERRORISM AND Make a charitable contribution HOMELAND SECURITY

For More Information Visit RAND at www.rand.org Explore the Pardee RAND Graduate School View document details

Limited Electronic Distribution Rights This document and trademark(s) contained herein are protected by law as indicated in a notice appearing later in this work. This electronic representation of RAND intellectual property is provided for non- commercial use only. Unauthorized posting of RAND electronic documents to a non-RAND website is prohibited. RAND electronic documents are protected under copyright law. Permission is required from RAND to reproduce, or reuse in another form, any of our research documents for commercial use. For information on reprint and linking permissions, please see RAND Permissions. This product is part of the Pardee RAND Graduate School (PRGS) dissertation series. PRGS dissertations are produced by graduate fellows of the Pardee RAND Graduate School, the world’s leading producer of Ph.D.’s in policy analysis. The dissertation has been supervised, reviewed, and approved by the graduate fellow’s faculty committee. Dissertation

Go Gentle into That Good Night The Past, Present, and Future of End-of-Life Care

Adam E. Singer

This document was submitted as a dissertation in May 2015 in partial fulfillment of the requirements of the doctoral degree in public policy analysis at the Pardee RAND Graduate School. The faculty committee that supervised and approved the dissertation consisted of Karl Lorenz (Chair), Daniella Meeker, and Joan Teno.

C O R P O R A T I O N Dissertation

Go Gentle into That Good Night The Past, Present, and Future of End-of-Life Care

Adam E. Singer

PARDEE RAND GRADUATE SCHOOL The Pardee RAND Graduate School dissertation series reproduces dissertations that have been approved by the student’s dissertation committee.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.

R® is a registered trademark.

Permission is given to duplicate this document for personal use only, as long as it is unaltered and complete. Copies may not be duplicated for commercial purposes. Unauthorized posting of RAND documents to a non-RAND website is prohibited. RAND documents are protected under copyright law. For information on reprint and linking permissions, please visit the RAND permissions page (http://www.rand.org/publications/permissions.html).

Published 2015 by the RAND Corporation 1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138 1200 South Hayes Street, Arlington, VA 22202-5050 4570 Fifth Avenue, Suite 600, Pittsburgh, PA 15213-2665 RAND URL: http://www.rand.org/ To order RAND documents or to obtain additional information, contact Distribution Services: Telephone: (310) 451-7002; Fax: (310) 451-6915; Email: [email protected] Abstract

End‐of‐life care has received increasing attention in recent years as the baby boomers age and health care costs continue to rise. This attention has brought with it remarkable growth in the field and improvement in care, but there remains work to be done in order to more consistently deliver high quality, compassionate, and patient‐ and family‐centered end‐of‐life care. In this dissertation, I examine the past, present, and future of end‐of‐life care in order to shed light on the most effective ways to organize and deliver it.

In the first analysis, I describe changes in symptom prevalence in the last year of life from 1998 to 2010. I find that the many important and troubling end‐of‐life symptoms, including pain and depression, became more common over that time frame, and that no symptom became less common. These changes occurred at the same time national efforts were underway to improve end‐of‐life care, which calls into question the nature of these changes and highlights the need to reexamine aspects of the delivery of end‐of‐life care.

In the second analysis, I review the current evidence for palliative health services interventions to identify the populations that are appropriate for palliative care and the interventions that are effective in improving patient and family quality of life and reducing health care use and costs. I find that there are a broad range of currently available interventions that span a variety of goals, settings, and providers; that certain types of interventions are more or less effective for certain quality of life outcomes; and that health care use and cost outcomes are poorly studied overall. Some of these results enhance the existing understanding of what makes for effective end‐of‐life care. This information can be used in designing end‐ of‐life care programs and in organizing research priorities for the field.

In the third analysis, I simulate how cost and quality of life outcomes could fare through 2040 if three of the most effective palliative health services interventions identified in the second analysis were implemented today. I find that all three interventions lower health care costs, reduce mortality, improve quality‐adjusted life years, and reduce pain, depression, and activities of daily living dependencies. These results highlight the benefit to individuals and to society of implementing evidence‐based approaches to end‐of‐life care.

iii

Table of Contents

Abstract ...... iii Abbreviations ...... ix Acknowledgements ...... xi Chapter 1. Introduction ...... 1 Figure 1.1: Projected Federal Medicare Spending ...... 2 Figure 1.2: Prevalence of Palliative Care in US Hospitals with 50 or More Beds ...... 3 References ...... 5 Chapter 2. Symptom Trends in the Last Year of Life, 1998‐2010: A Cohort Study ...... 6 Abstract ...... 6 Introduction ...... 8 Methods ...... 9 Setting and Study Participants ...... 9 Symptom Outcomes ...... 9 Symptom Covariates ...... 9 Decedent Categories ...... 10 Statistical Analysis ...... 10 Results ...... 13 Discussion...... 15 References ...... 18 Tables ...... 20 Table 1. Characteristics of Study Population ...... 20 Table 2. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 in the Entire Population ...... 23 Table 3. Adjusted Total Percent Change in Prevalence of Symptoms between 1998 and 2010 ...... 25 Appendix Table 1. HRS Survey Questions Used to Construct Symptom Outcomes ...... 27 Appendix Table 2. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 among Decedent Categories (Fully Adjusted Models) ...... 28 Appendix Table 3. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 among Decedent Categories (Minimally Adjusted Models) ...... 31 Appendix Table 4. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 among Decedent Categories (Fully Adjusted Models Excluding 2002 Survey Wave) ...... 34

Appendix Table 5. Adjusted Total Percent Change in Prevalence of Symptoms between 1998 and 2010 (Fully Adjusted Models Excluding 2002 Survey Wave) ...... 38 Appendix Table 6. Adjusted Average Yearly Percent Change in Prevalence of Symptoms between 1998 and 2010 (Fully Adjusted Models) ...... 39 Chapter 3. Populations and Interventions for Palliative and End‐of‐Life Care: A Systematic Review ...... 40 Abstract ...... 40 Introduction ...... 42 Methods ...... 43 Data Sources and Searches ...... 43 Study Selection ...... 44 Data Extraction and Quality Assessment ...... 45 Data Synthesis and Analysis ...... 45 Results ...... 47 Literature Flow ...... 47 Intervention Populations and Characterizing Advanced Illness ...... 47 Conditions of Study Participants ...... 47 Severity of Illness ...... 48 Intervention Elements ...... 49 Intervention Personnel ...... 49 Multidisciplinary Teams ...... 49 Intervention Settings...... 50 Supporting Technology ...... 51 Quality of Life‐Relevant Outcomes ...... 51 Patient Outcomes ...... 51 Caregiver Outcomes ...... 52 Relationships among Key Study Characteristics and Select Quality of Life‐Relevant Outcomes ... 52 Economic Outcomes ...... 54 Health Care Use ...... 54 Health Care Costs ...... 55 Discussion...... 56 References ...... 59 Tables and Figures...... 67

Table 1. Patient Populations ...... 67 Table 2. Intervention Components ...... 68 Table 3. Patient and Caregiver Quality of Life‐Relevant Outcomes ...... 70 Table 4. Economic Outcomes ...... 72 Appendix Figure 1. Literature Flow ...... 73 Appendix Table 1. Literature Search Strategy ...... 74 Appendix Table 2. Acronyms Used in Evidence Tables ...... 91 Appendix Table 3. Evidence Tables ...... 97 Chapter 4. Estimating the Value of Palliative Care for Older Adults: What Does the Evidence Support? 345 Abstract ...... 345 Introduction ...... 347 Methods ...... 348 Future Elderly Model ...... 348 Evidence‐Based Scenarios ...... 348 Status Quo Scenarios ...... 349 Palliative Care Team Scenario ...... 349 Pain Management Education Scenario ...... 350 Case Management Scenario ...... 350 All Scenarios ...... 351 Outcomes ...... 351 Measures ...... 352 Analysis ...... 352 Results ...... 354 Pain, Depression, and ADL Dependencies ...... 354 Population Size, Median Age, and Mortality ...... 354 Health Care Costs ...... 355 QALYs ...... 357 Discussion...... 358 References ...... 363 Tables and Figures...... 365 Table 1. Intervention Effects on Depression, Pain, and ADL Dependencies ...... 365

vii

Figure 1a. Projections of Per Capita Societal Costs in Status Quo and Intervention Scenarios, 2016‐ 2040 ...... 366 Figure 1b. Projections of Per Capita OOP Costs in Status Quo and Intervention Scenarios, 2016‐2040 ...... 367 Figure 2. Projections of Per Capita QALYs in Status Quo and Intervention Scenarios, 2016‐2040 .. 368 Appendix Figure 1a. Projections of Percent Change in Per Capita Societal Costs in Intervention vs. Status Quo Scenarios, 2016‐2040 ...... 369 Appendix Figure 1b. Projections of Percent Change in Per Capita OOP Costs in Intervention vs. Status Quo Scenarios, 2016‐2040 ...... 370 Appendix Figure 2. Projections of Percent Change in Per Capita QALYs in Intervention vs. Status Quo Scenarios, 2016‐2040 ...... 371 Appendix Table 1. FEM Regression Model for Mortality ...... 372 Appendix Table 2. FEM Regression Models for Costs ...... 374 Chapter 5. Conclusions ...... 376 Summary ...... 376 Implications ...... 377 Improving Palliative Care Programs ...... 377 Research Priorities ...... 378 Funding Priorities ...... 379 Closing Thoughts ...... 380 References ...... 382

viii

Abbreviations

ACOVE Assessing Care of Vulnerable Elders ADL Activity of daily living AMSTAR Assessment of Multiple Systematic Reviews ASSIST Assessing Symptoms Side Effects and Indicators of Supportive Treatment CAPC Center to Advance Palliative Care CBO Congressional Budget Office CES‐D Center for Epidemiologic Studies Depression scale CHF Congestive heart failure CM Case management COPD Chronic obstructive pulmonary disease EHR Electronic health record EQ‐5D EuroQol‐5 Dimensions ESLD End‐stage liver disease ESRD End‐stage renal disease FEM Future Elderly Model HRQOL Health‐related quality of life HRS Health and Retirement Study IOM Institute of Medicine IQR Interquartile range NIH National Institutes of Health OOP Out‐of‐pocket OR Odds ratio PCT Palliative care team PME Pain management education QALY Quality‐adjusted life year QOL Quality of life RCT Randomized controlled trial SSDI Social Security Disability Insurance

See Appendix Table 2 in Chapter 3 for abbreviations used in evidence tables.

Acknowledgements

I would like to extend my sincerest thanks to my dissertation committee: Karl Lorenz, Daniella Meeker, and Joan Teno. I have learned an immense amount from each of them, and I will take these lessons with me as I look ahead to the future. Their expertise is second to none, and their input over the years has taught me a great deal about conducting research and has helped me hone my interests and skills. I am truly grateful for their guidance.

I would also like to thank all of my mentors, collaborators, coauthors throughout my time at the Pardee RAND Graduate School: Joanne Lynn, June Lunney, Sydney Dy, Joy Goebel, Sangeeta Ahluwalia, Yan Kim, Anne Walling, Kelly Chong, Aneesa Motala, Chris Skeels, Duncan Leaf, Megan Clifford, Elizabeth Dzeng, Claire O’Hanlon, Jaime Goldberg, Claudia Ochotorena, Tayla Ash, Mike Cui, Roberta Shanman, and Keegan Wood. I have greatly enjoyed working with them and have benefitted from their knowledge and insights.

The Trajectories and Palliation Study at RAND Health funded the vast majority of the research in this dissertation. It was funded primarily by a National Institute of Nursing Research grant and supplemented by a Cambia Health Foundation Sojourns Award and the California HealthCare Foundation. I am grateful to all of these institutions for their support.

This dissertation also received support from the Dana G. Mead and Anne and James Rothenberg Dissertation Awards. I was fortunate to be awarded these scholarships and would like to thank their donors and the Pardee RAND Graduate School.

Lastly, I would like to thank the Pardee RAND Graduate School and its faculty and administration for providing me with this opportunity. The last few years have broadened my horizons in ways I could never have envisioned.

Chapter 1. Introduction

In a seminal 1997 report on the state of end‐of‐life care, the IOM described extensive patient and family suffering and emphasized the need for better care at the end of life.1 It highlighted an array of concerning aspects of end‐of‐life care at the time, including that clinicians could do better to prevent or relieve suffering with existing knowledge and therapies, and that the education and training of health care professionals failed to provide them with the attitudes, knowledge, and skills required to care well for dying patients. It also called for strengthening the accountability of the quality of end‐of‐life care through the use of better data and tools for evaluating outcomes important to patients and families.

Many of the IOM’s indictments can be addressed through high‐quality palliative care. Palliative care is defined as “an approach that improves the quality of life of patients and their families facing the problems associated with life‐threatening illness, through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual.”2 The administration of palliative care has been shown to improve quality of life among end‐of‐life patients.3‐6

Palliative approaches also have the potential to reduce health care costs at the end of life, although the evidence supporting these reductions is not as strong as for improvements in quality of life.7 The end of life is expensive: 30% of lifetime Medicare expenditures are spent in the last year of life.8 High spending near the end of life is not necessarily a problem in and of itself, as end‐of‐life patients have more complex needs that must be addressed by the health care system. However, as the population continues to age, the number of patients contributing to end‐of‐life spending continues to increase, which places an ever larger burden on Medicare. Indeed, Medicare expenditures are projected to rise (Figure 1.1) to potentially unsustainable levels, in part due to this effect.9 In light of this, any cost reductions resulting from palliative approaches would be beneficial.

Figure 1.1: Projected Federal Medicare Spending

Source: CBO10

Palliative care in the United States has grown considerably since the IOM report: less than 25% of hospitals with 50 or more beds had a palliative care program in the 2000; that number was projected to be 84% in 2014 (Figure 1.2).11 In addition, palliative medicine has been certified by the American

Board of Medical Specialties, and palliative medicine fellowships have been approved by the

Accreditation Council for Graduate Medical Education. Policy and practice have emphasized better management of pain and other symptoms.12,13 Hospice use more than doubled from 2000 to 2013, with

47% of Americans receiving hospice before death in 2013.14

Figure 1.2: Prevalence of Palliative Care in US Hospitals with 50 or More Beds

Source: CAPC11

The growth of palliative care is a promising direction for end‐of‐life care, but challenges remain in designing and organizing it and in understanding its impact. This dissertation addresses these issues, which are detailed below. It is comprised of five chapters, including this introduction.

First, the growth of the field itself is not necessarily tantamount to improved patient and family outcomes. Because palliative care is known to improve quality of life and because it has grown so much, there is reason to believe that, if it is implemented in the right way, it can lead to population‐level improvement in quality of life. This latter component is the final link in the chain, but as of yet it has remained largely unaddressed. Knowing whether growth in the field has led to improved patient and family outcomes has implications for the design and implementation of palliative care. In Chapter 2 of

3 the dissertation, I address this issue by analyzing trends in the prevalence from 1998 to 2010 of eight common symptoms experienced in end‐of‐life patients, which are a core component of quality of life.

Second, palliative care has traditionally been defined as an approach to care (exemplified in the definition above), but in practice it must be operationalized as specific sets of health services. Thus far, the design of these services has largely focused on inpatient settings and physician‐led multidisciplinary teams, but there are a broad range of currently available palliative health services that might be more effective for various patient populations and quality of life and economic outcomes. It would be valuable for policymakers, payers, and providers to know which services to implement for their patient populations when designing or refining palliative care programs. In Chapter 3 of the dissertation, I address these issues by synthesizing the current evidence for which populations are appropriate for palliative care and which interventions are most effective in improving quality of life outcomes and reducing health care use and costs.

Third, policymakers and payers could benefit from the knowledge of how effective palliative services might affect quality of life and economic outcomes in the future. Tasked with designing and implementing large‐scale, long‐term programs, these decision‐makers need to know how best to deploy scare resources so as to maximize benefit to individuals and society. Projections of the future quality of life and economic impact of palliative care have been missing from the field, and research is needed that focuses on these questions. In Chapter 4 of the dissertation, I address these issues by simulating the impact through 2040 of three of the most effective palliative interventions identified in Chapter 3 on health care costs, QALYs, mortality, and pain, depression, and ADL dependencies.

In Chapter 5 of the dissertation, I summarize the conclusions of this work and highlight its policy implications.

References

1. Approaching Death: Improving Care at the End of Life. Washington, D.C.: Institute of Medicine;1997. 2. WHO Definition of Palliative Care. http://www.who.int/cancer/palliative/definition/en/. Accessed April 17, 2015. 3. Brumley R, Enguidanos S, Jamison P, et al. Increased satisfaction with care and lower costs: results of a randomized trial of in‐home palliative care. J Am Geriatr Soc. 2007;55(7):993‐1000. 4. Morrison RS, Penrod JD, Cassel JB, et al. Cost savings associated with US hospital palliative care consultation programs. Arch Intern Med. 2008;168(16):1783‐1790. 5. Gade G, Venohr I, Conner D, et al. Impact of an inpatient palliative care team: a randomized control. trial J Palliat Med. 2008;11(2):180‐190. 6. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non‐ small‐cell lung cancer. N Engl J Med. 2010;363(8):733‐742. 7. May P, Normand C, Morrison RS. Economic impact of hospital inpatient palliative care consultation: review of current evidence and directions for future research. J Palliat Med. 2014;17(9):1054‐1063. 8. Hogan C, Lunney J, Gabel J, Lynn J. Medicare beneficiaries' costs of care in the last year of life. Health Aff. 2001;20(4):188‐195. 9. Potetz L. Medicare Spending and Financing: A Primer. Menlo Park, CA: The Henry J. Kaiser Family Foundation;2011. 10. The 2014 Long‐Term Budget Outlook. Washington, DC: Congressional Budget Office;2014. 11. Growth of Palliative Care in U.S. Hospitals: 2013 Snapshot. New York, NY: Center to Advance Palliative Care;2013. 12. Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research. Washington, D.C.: Institute of Medicine;2011. 13. Qaseem A, Snow V, Shekelle P, et al. Evidence‐based interventions to improve the palliative care of pain, dyspnea, and depression at the end of life: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2008;148(2):141‐146. 14. March 2015 Report to the Congress: Medicare Payment Policy Chapter 12. Hospice Services. Washington, DC: Medicare Payment Advisory Commission;2015.

Chapter 2. Symptom Trends in the Last Year of Life, 1998‐2010: A Cohort Study

Abstract

Background: Research in the 1990s described serious symptoms at the end of life, and a 1997 IOM report called for improvement. Palliative care and hospice have grown considerably since then, but it is not known whether symptoms have improved.

Objective: To describe changes in pain intensity and symptom prevalence during ther last yea of life from 1998 to 2010.

Design: Observational study.

Setting: HRS, a nationally representative longitudinal survey of community‐dwelling US residents aged

51 and older.

Participants: 7,204 HRS participants who died while enrolled in the study, and their family respondents.

Measurements: Proxy‐reported pain during the last year of life; depression, confusion, dyspnea, incontinence, fatigue, anorexia, and vomiting for a period of at least one month during the last year of life. Trends in pain intensity and symptom prevalence were analyzed among all decedents and those within cancer, congestive heart failure or chronic lung disease, frailty, and sudden death.

Results: Between 1998 and 2010, proxy reports of the prevalence of any pain (mild, moderate, or severe) increased for all decedents by 11.9% (95% CI: 3.1%, 21.4%). Reported prevalence of depression increased for all decedents by 26.6% (14.5%, 40.1%), for congestive heart failure or chronic lung disease by 27.0% (8.1%, 49.3%), and for frailty by 39.4% (9.9%, 79.8%). Reported prevalence of periodic confusion increased for all decedents by 31.3% (18.6%, 45.1%), for congestive heart failure or chronic lung disease by 24.9% (6.0%, 47.6%), for frailty by 20.3% (5.9%, 39.1%), and for sudden death by 45.7%

(5.9%, 106.1%). Trends in the reported prevalence of most other symptoms in most groups of decedents were positive but not significant. Moderate or severe pain, severe fatigue, anorexia, and frequent vomiting did not show significant changes in any group of decedents. There were no significant changes for cancer.

Limitation: Proxy reports, mostly yes/no symptom questions.

Conclusion: Despite national efforts to improve end‐of‐life care, proxy reports of pain and other alarming symptoms in the last year of life increased from 1998 to 2010.

Introduction

Symptoms are among the most distressing aspects of the end‐of‐life experience for patients and families. Although we have lacked definitive population‐based data, systematic reviews show the high prevalence of many symptoms, including pain, dyspnea, and depression, across multiple advanced diseases.1 Pain is among the most prevalent and troubling of symptoms, and patients and families fear and wish to avoid it near the end of life.2,3

Expert panels and consensus statements highlight the appropriate identification of and care for many end‐of‐life symptoms, and several feasible, valid, and reliable quality indicators and measures have been developed to assess the occurrence and management of symptoms near the end of life.4‐8

Evidence supports a variety of effective interventions to alleviate pain, dyspnea, and depression, among other symptoms; these interventions often promote comfort across a range of conditions.5,9‐11

Palliative care and hospice have grown considerably since the 1997 IOM report, but research also suggests that there remain actionable gaps regarding the management of symptoms in cancer and other conditions.12,13 In short, it is not known whether national efforts to improve end‐of‐life care since the IOM report have led to changes in the prevalence of commonly occurring end‐of‐life symptoms. I therefore examined nationally representative trends in end‐of‐life symptom prevalence from 1998 to

2010 for the population as a whole and for different groups of decedents. Given the strength of practice evidence and policy attention to both cancer and pain,6,14 I expected that overall trends would be better for patients with cancer than with other conditions, and that trends in the prevalence and severity of pain would be better than for other symptoms.

Methods

Setting and Study Participants

I used data from HRS, a nationally representative longitudinal survey of community‐dwelling adults aged 51 or older in the contiguous United States.15,16 The mortality rate of the HRS population is comparable to that of the overall United States population of adults aged 51 and older. HRS participants are interviewed every two years until their deaths. After each participant’s death, HRS interviews a proxy informant, typically a family member, who was most familiar ewith th health, family, and financial situation of the participant. I included participants who died while enrolled in HRS and whose proxy informant provided a postmortem interview within two years of death.

Symptom Outcomes

I evaluated symptom prevalence using eight yes/no questions that asked about the presence of pain, depression, periodic confusion, dyspnea, severe fatigue, incontinence, anorexia, and frequent vomiting. All symptom questions excluding pain asked if the decedent experienced the symptom for a period of at least one month during the last year of life; for pain only, the question asked if the decedent were often troubled with pain in the last year of life. In addition, a follow‐up question asked about the degree of pain (mild, moderate, or severe) if pain were present. I analyzed any pain and moderate or severe pain separately. Appendix Table 1 provides the exact wording for all symptom questions.

Symptom Covariates

I employed a model of whole person distress, which included demographics as well as clinical, psychological, and social domains, and which I modified to include proxy factors. Demographics included age at death, gender, and race/ethnicity. Clinical covariates included number of nights spent in a hospital in the last two years of life, nursing home residency at the time of death and length of nursing home residency prior to death (regardless of site of death), number of ADL dependencies (0‐6 scale comprised of help with dressing, walking, bathing, eating, transferring, and toileting), and diagnosis of

9 arthritis (included in models for pain only). The psychological covariate was depression reported in the interview wave prior to death, as indicated by a score of four or more on an eight‐item subset of the 20‐ item Center for Epidemiologic Studies Depression Scale.17,18 The social covariates included highest level of education, household wealth, and household income. Proxy covariates included relationship of proxy to decedent (spouse, son or daughter, or other); time elapsed between decedent’s death and proxy’s interview; average hours per week of care given to decedent from all informal carers, including care given by proxy; highest level of education (only available if proxy was spouse [approximately 33% of proxies]); and English/Spanish preference for interview. In addition, I also included the year that the decedent entered the HRS sample.

Decedent Categories

I employed a sequential categorization scheme based on prior work to resolve decedents into four mutually exclusive categories: sudden death, cancer, CHF or chronic lung disease, and frailty.16,19‐21

Decedents were categorized as (1) sudden death, if they died within one day of receiving their terminal diagnosis; (2) cancer, if their proxy reported cancer as their cause of death; (3) CHF or chronic lung disease, if they or their proxy reported a diagnosis of either or both of these conditions; or (4) frailty, if they died in a nursing home, they or their proxy reported physician‐diagnosed memory impairment, or their proxy reported hip fracture in the last two years of life.

Questions that asked about diagnoses of CHF, chronic lung disease, and physician‐diagnosed memory impairment were asked both of proxies in postmortem interviews and of decedents themselves in each survey wave while they were alive. I considered each of these conditions to be present if either the decedent or the proxy reported its presence.

Statistical Analysis

All analyses accounted for complex survey design using sampling weights, providing estimates representative of the entire US population. I used multivariable logistic regression models to investigate

10 the association between each symptom outcome and time of death. I employed two sets of models for each outcome: a fully adjusted model that included all symptom covariates described above, and a minimally adjusted model that included only demographics and time elapsed between decedent’s death and proxy’s interview. I conducted all analyses for the cohort as a whole and in each of the four mutually exclusive decedent categories. All decedents were included in the entire population analysis, and decedents who qualified for membership in one of the four categories were also analyzed separately in that category.

I report adjusted prevalence of each symptom by year of death from both fully and minimally adjusted models. I estimated the percent change in the adjusted prevalence of each symptom using first and last death dates from all regressions, and used bootstrapping with bias correction to estimate confidence intervals.22 I report adjusted percent change in the prevalence of each symptom from both fully and minimally adjusted models.

I imputed missing values using multiple imputation by chained equations with 15 imputed datasets using Stata’s “ice” command;23 analyses in imputed data were carried out with “mi estimate” commands or manually using Rubin’s rules.24 I constructed a multiple imputation model that included time of death, decedent categories, all symptom outcome variables, and all symptom covariates described above except for the year that the decedent entered the HRS sample. Two sets of variables were imputed using forms different from those included in regression analyses: (1) pain was captured with one categorical variable that recorded no pain or mild, moderate, or severe pain, and the outcomes of any pain and moderate or severe pain were created from this categorical variable after imputation;

(2) the six ADL questions were included separately in the imputation model and combined into a scale after imputation.

Of the 31 total variables in the imputation model, 12 had no missing values and were not themselves imputed. The remaining 19 variables were imputed (numbers of missing values are

11 presented here): pain categorical variable (241), depression (225), periodic confusion (53), dyspnea (54), incontinence (128), severe fatigue (123), anorexia (212), frequent vomiting (99), race/ethnicity (7), number of nights spent in a hospital in the last two years of life (582), help with dressing (31), help with walking (26), help with bathing (31), help with eating (33), help with transferring (43), help with toileting

(44), diagnosis of arthritis (4), depression in interview wave prior to death (485), and household wealth

(75).

In the 2002 survey wave only, approximately 41% of proxies were erroneously not asked non‐ pain symptom questions due to faulty survey logic. Upon discovering the issue, HRS corrected the survey logic but did not go back and collect the missing data. As a result, those responses were permanently lost. In my main analyses, I excluded those 41% of proxies in the 2002 survey wave only. As a sensitivity analysis, I also replicated these analyses excluding the entire 2002 survey wave to investigate the effect of potentially non‐random missing data due to faulty survey logic.

I also performed an additional sensitivity analysis to evaluate whether there was substantial year‐to‐year variation in my outcomes. I replicated each minimally adjusted multivariable logistic regression using a continuous variable that captured each survey year rather than a variable that captured time of death. I report average yearly percent change and confidence intervals.

I conducted analyses with Stata 12.1 IC (StataCorp, College Station, Texas). The RAND IRB approved the study.

Results

8,641 HRS participants died between 1998 and 2010, for which 8,089 proxy interviews were completed. Of those, 7,204 proxy interviews were completed within two years of the decedent’s death.

The latter comprised my study cohort. The response rates for the six survey waves from 2000 to 2010 were 86%, 85%, 91%, 88%, 92%, and 85%, respectively. Table 1 presents the weighted characteristics of the cohort as a whole and by timing of death. Their mean age at death was 79.1 years. Fifty‐four percent were women; 17% were nonwhite. Twenty‐two percent were reported to have had cancer; 33%, CHF or chronic lung disease; 16%, frailty; and 16%, sudden death. Fourteen percent reported none of these categories. Approximately 50% of decedents were reported to have had moderate or severe pain, depression, periodic confusion, dyspnea, or incontinence; approximately 60% were reported to have had any pain, severe fatigue, or anorexia; and approximately 12% were reported to have had frequent vomiting.

Table 2 presents the adjusted prevalence of symptoms from both fully and minimally adjusted models for all decedents between 1998 and 2010; Appendix Tables 2 and 3 present this information for cancer, CHF or chronic lung disease, frailty, and sudden death. These estimates are largely similar between the two sets of models and show the high prevalence of most symptoms in the entire population and in all decedent categories. Most estimates also suggest positive trends in prevalence over time.

Table 3 presents the adjusted percent change in prevalence of symptoms from both fully and minimally adjusted models among all decedents and for cancer, CHF or chronic lung disease, frailty, and sudden death. In fully adjusted models, proxy reports of the prevalence of any pain (mild, moderate, or severe) increased for all decedents by 11.9% (95% CI: 3.1%, 21.4%) between 1998 and 2010. Reported prevalences of depression and periodic confusion each increased for all decedents and in multiple decedent categories by large percentages (between 20.3% and 45.7%). Incontinence also increased for

13 all decedents, and dyspnea increased for sudden death. Trends in the reported prevalence of most other symptoms in most groups of decedents were positive but not significant. Moderate or severe pain, severe fatigue, anorexia, and frequent vomiting did not show significant changes in any group of decedents. There were no significant changes for cancer.

As compared with fully adjusted models, minimally adjusted models generally showed larger trends, and more trends were significant. Proxy reports of moderate or severe pain increased for all decedents by 20.9% (95% CI: 9.7%, 32.9%), and reports of any pain increased for all decedents by 23.3%

(13.4%, 33.9%), for CHF or chronic lung disease by 22.5% (7.0%, 41.7%), and for sudden death by 32.8%

(5.3%, 68.9%). Increases in the reported prevalences of depression and periodic confusion for all decedents and in multiple decedent categories were similar to those in fully adjusted models. Dyspnea, severe fatigue, and anorexia also increased for all decedents, and dyspnea and incontinence increased for sudden death. As with fully adjusted models, there were no significant changes for cancer.

I performed two sensitivity analyses for fully adjusted models: estimating symptom prevalence and percent change excluding the 2002 survey wave (Appendix Tables 4 and 5), and investigating yearly changes rather than 12‐year changes (Appendix Table 6). Overall results do not differ appreciably between analyses with and without the 2002 survey wave, and the magnitude and significance of yearly changes generally comport with those of 12‐year changes.

Discussion

I analyzed bereaved family interviews conducted on behalf of decedents in a large, nationally representative survey in order to evaluate changes in end‐of‐life symptom burden in the United States between 1998 and 2010. Over the time‐frame of my study, proxy reports of many symptoms increased in prevalence, including pain, depression, and periodic confusion. Consistent with my hypothesis, I found no significant trends in any symptom in cancer.

Proxy reports of worsening symptom prevalence raise concerns about shortcomings in end‐of‐ life care despite increasing national attention and resources devoted to it. Indeed, recent studies of health care performance tsuggest tha there remain persistent gaps in addressing symptoms near the end of life.25‐27 It is particularly concerning that proxy reports of pain have increased, as pain is among the most visible and well‐studied aspect of the end‐of‐life experience, has received policy attention, and significantly impacts HRQOL.6 Moderate and severe pain, for example, has an HRQOL impact more pronounced than many health and sociodemographic factors.28,29 It is encouraging, however, that trends in cancer pain prevalence and severity may have stabilized. This should be monitored in the face of growing public concern about prescription opioid abuse, which may create resistance to using opioids from both clinicians and patients in otherwise appropriate scenarios.30

My results indicate that symptom burden is high near the end of life, and my findings are generally concordant with population‐level studies available from other countries. Cancer outpatients in

Canada assessed between diagnosis and death reported the prevalence of pain, dyspnea, depression, and anorexia between 45% and 60%, and fatigue at 75%.31 Dyspnea prevalence among hospice patients in Australia at three months prior to death was 50% overall, 65% for heart failure, and 88% for end‐stage pulmonary disease.32

There are many factors that could contribute to the persistence and potential increase in prevalence of these troubling end‐of‐life symptoms. Recent reports demonstrate that the intensity of

15 treatment and the rate of adverse transitions have been increasing near the end of life.33 Hospice is often “tacked on” to this more intense late life care: even though hospice use doubled from 2000 to

2009, the median stay is less than three weeks.33,34 Some patients who have short stays may not realize the full benefits of hospice for symptomatic relief. Palliative care services remain more common in hospitals, and patients may not have consistent access to palliative services in outpatient, home, and long‐term facility settings, where most of the course of a terminal illness takes place. Effective interventions can sometimes mitigate the symptoms I have highlighted (e.g., opioid regimens for pain and dyspnea),35 but there remain significant gaps in their delivery near the end of life.12,13 This suggests that interventions may not be reaching the right patients in the right ways.

In addition, increased attention to end‐of‐life care generally and symptoms specifically may have increased proxy reporting of symptoms over the time frame of my study. For example, if clinicians became more likely to ask about symptoms, proxies may have become more aware of them and thus more likely to report them.

Proxy reports inevitably reflect both the patient’s and proxy’s experiences. They can provide invaluable information, but further research is needed to improve their validity, particularly with regard to the reporting of subjective symptoms, which proxies tend to overestimate.36 The evidence base is inconsistent with respect to the impact of a variety of factors on the validity of proxy reports, including the contributions of caregiver distress and the proxy’s relationship to the decedent.36 Improved understanding of proxy reports is especially important now that they are being used in hospice and other settings for quality assessment and improvement.9

I faced several limitations. First, I used proxy reports of outcomes, which could affect the validity of subjective symptoms,36 but is unlikely to explain symptom trends, since proxies were used in all survey waves. Second, I used mostly yes/no questions, which might have masked variation in symptom intensity. Although I could not assess changes in the severity of non‐pain outcomes, the increase in their

16 prevalence raises concerns. Third, due to limitations in the survey itself, I could not capture all constructs relevant to evaluating symptom trends in end‐of‐life patients, particularly hospice enrollment and site of death. The attenuation of many of my results in fully adjusted models as compared with minimally adjusted models suggests that changes in many of the proxy and decedent characteristics were partially responsible for the changes in reported symptoms I observed. It is therefore important to recognize that residual confounding may remain due to other factors I could not account for in my models.

In summary, between 1998 and 2010, proxy reports of serious pain and many other distressing symptoms became more common near the end of life. Given our knowledge of best practices and continued gaps in applying them, there is an urgent need simply to benchmark current practice against current knowledge. Future research should evaluate settings that provide better and worse end‐of‐life symptom management in order to offer insight into promoting best practices. Improving care at the end of life will necessitate further investment to understand the trends I identified, and steps will be required to reverse them.

References

1. Solano JP, Gomes B, Higginson IJ. A comparison of symptom prevalence in far advanced cancer, AIDS, heart disease, chronic obstructive pulmonary disease and renal disease. J Pain Symptom Manage. 2006;31(1):58‐69. 2. Steinhauser KE, Christakis NA, Clipp EC, McNeilly M, McIntyre L, Tulsky JA. Factors considered important at the ende of lif by patients, family, physicians, and other care providers. JAMA. 2000;284(19):2476‐2482. 3. Steinhauser KE, Clipp EC, McNeilly M, Christakis NA, McIntyre LM, Tulsky JA. In search of a good death: observations of patients, families, and providers. Ann Intern Med. 2000;132(10):825‐832. 4. Qaseem A, Snow V, Shekelle P, et al. Evidence‐based interventions to improve the palliative care of pain, dyspnea, and depression at the end of life: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2008;148(2):141‐146. 5. Lorenz KA, Lynn J, Dy SM, et al. Evidence for Improving Palliative Care at the End of Life: A Systematic Review. Ann Intern Med. 2008;148(2):147‐159. 6. Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research. Washington, D.C.: Institute of Medicine;2011. 7. Mularski RA, Campbell ML, Asch SM, et al. A review of quality of care evaluation for the palliation of dyspnea. Am J Respir Crit Care Med. 2010;181(6):534‐538. 8. Rayner L, Price A, Hotopf M, Higginson IJ. Expert opinion on detecting and treating depression in palliative care: A Delphi study. BMC Palliat Care. 2011;10:10. 9. NIH State‐of‐the‐Science Conference Statement on improving end‐of‐life care. NIH Consens State Sci Statements. 2004;21(3):1‐26. 10. Priorities for the National Quality Strategy: National Priorities Partnership;2011. 11. Future Directions for the National Healthcare Quality and Disparities Reports. Washington, D.C.: Institute of Medicine;2010. 12. Walling AM, Asch SM, Lorenz KA, et al. The quality of supportive care among inpatients dying with advanced cancer. Support Care Cancer. 2012;20(9):2189‐2194. 13. Walling AM, Tisnado D, Asch SM, et al. The quality of supportive cancer care in the veterans affairs health system and targets for improvement. JAMA Intern Med. 2013;173(22):2071‐2079. 14. Improving Palliative Care for Cancer: Summary and Recommendations. Washington, D.C.: Institute of Medicine;2001. 15. Juster FT, Suzman R. An overview of the Health and Retirement Study. J Hum Resour. 1995;30(Suppl):S7‐S56. 16. Lunney JR, Lynn J, Hogan C. Profiles of older Medicare decedents. J Am Geriatr Soc. 2002;50(6):1108‐1112. 17. Reyes‐Gibby CC, Aday LA, Anderson KO, Mendoza TR, Cleeland CS. Pain, depression, and fatigue in community‐dwelling adults with and without a history of cancer. J Pain Symptom Manage. 2006;32(2):118‐128. 18. Steffick DE. Documentation of Affective Functioning Measures in the Health and Retirement Study. Ann Arbor, MI: University of Michigan;2000. 19. Lunney JR, Lynn J, Foley DJ, Lipson S, Guralnik JM. Patterns of functional decline at the end of life. JAMA. 2003;289(18):2387‐2392. 20. Hogan C, Lunney J, Gabel J, Lynn J. Medicare beneficiaries' costs of care in the last year of life. Health Aff (Millwood). 2001;20(4):188‐195. 21. Smith AK, Cenzer IS, Knight SJ, et al. The epidemiology of pain during the last 2 years of life. Ann Intern Med. 2010;153(9):563‐569.

22. Efron B. Better bootstrap confidence intervals. J Am Stat Assoc. 1987;82(397):171‐185. 23. Royston P. Multiple imputation of missing values: further update of ice, with an emphasis on categorical variables. The Stata Journal. 2009;9(3):466‐477. 24. Rubin DB. Multiple imputation after 18+ years. J Am Stat Assoc. 1996;91(434):473‐489. 25. Walling AM, Asch SM, Lorenz KA, et al. The quality of care provided to hospitalized patients at the end of life. Arch Intern Med. 2010;170(12):1057‐1063. 26. Dy SM, Asch SM, Lorenz KA, et al. Quality of End‐of‐Life Care for Patients with Advanced Cancer in an Academic Medical Center. J Palliat Med. 2011;14(4):451‐457. 27. Malin JL, O'Neill SM, Asch SM, et al. Quality of supportive care for patients with advanced cancer in a VA medical center. J Palliat Med. 2011;14(5):573‐577. 28. Langley PC, Liedgens H. The impact of pain severity and frequency on HRQoL in the big 5 european union countries2010. 29. Lorenz KA, Shapiro MF, Asch SM, Bozzette SA, Hays RD. Associations of symptoms and health‐ related quality of life: findings from a national study of persons with HIV infection. Ann Intern Med. 2001;134(9 Pt 2):854‐860. 30. Fine RL. Ethical and practical issues with opioids in life‐limiting illness. Proc (Bayl Univ Med Cent). 2007;20(1):5‐12. 31. Barbera L, Seow H, Howell D, et al. Symptom burden and performance status in a population‐ based cohort of ambulatory cancer patients. Cancer. 2010;116(24):5767‐5776. 32. Currow DC, Smith J, Davidson PM, Newton PJ, Agar MR, Abernethy AP. Do the trajectories of dyspnea differ in prevalence and intensity by diagnosis at the end of life? A consecutive cohort study. J Pain Symptom Manage. 2010;39(4):680‐690. 33. Teno JM, Gozalo PL, Bynum JP, et al. Change in end‐of‐life care for Medicare beneficiaries: site of death, place ofd care, an health care transitions in 2000, 2005, and 2009. JAMA. 2013;309(5):470‐477. 34. NHPCO Facts and Figures: Hospice Care in America. Alexandria, VA: National Hospice and Palliative Care Organization;2013. 35. Ben‐Aharon I, Gafter‐Gvili A, Leibovici L, Stemmer SM. Interventions for alleviating cancer‐ related dyspnea: a systematic review and meta‐analysis. Acta Oncol. 2012;51(8):996‐1008. 36. McPherson CJ, Addington‐Hall JM. Judging the quality of care at the end of life: can proxies provide reliable information? Soc Sci Med. 2003;56(1):95‐109.

Tables

Table 1. Characteristics of Study Populationa

% of Decedentsb Interval of Death All 1998‐2000 2000‐2002 2002‐2004 2004‐2006 2006‐2008 2008‐2010 Characteristic (n = 7,204) (n = 1,243) (n = 1,226) (n = 1,144) (n = 1,186) (n = 1,212) (n = 1,193) Age at death < 65 12.3 13.7 10.4 7.8 14.6 14.6 12.0 65‐74 19.6 20.4 21.2 18.1 18.3 18.5 21.3 75‐84 32.6 33.0 33.2 39.1 31.5 30.4 29.2 > 84 35.5 32.9 35.2 35.0 35.6 36.5 37.5

Women 53.6 51.3 53.0 54.4 51.4 55.9 55.4

Ethnicity Non‐Hispanic White 83.2 84.4 82.3 83.7 81.9 83.1 83.8 Non‐Hispanic Black 10.3 10.3 10.7 9.9 10.6 10.6 9.9 Hispanic 4.6 3.7 4.9 4.4 4.9 4.7 4.9 Other 1.9 1.6 2.1 2.0 2.7 1.6 1.4

Decedent categories Cancer 22.2 23.8 21.8 21.2 20.9 22.5 22.9 CHF or chronic lung disease 32.5 27.7 30.9 32.9 34.8 33.5 34.6 Frailty 15.5 14.5 15.0 15.4 16.1 16.9 14.8 Sudden death 15.7 18.7 14.9 15.7 13.4 16.1 15.7

Symptoms Moderate or severe pain 50.6 46.7 47.7 49.9 51.9 51.5 54.8

Any pain 57.5 53.1 53.9 56.8 58.7 57.9 63.6 Depression 51.4 44.6 49.5 51.8 52.0 53.2 55.5 Periodic confusion 47.9 41.7 42.8 46.1 49.6 54.1 49.1 Dyspnea 52.6 49.8 51.9 52.4 50.9 53.5 56.2 Incontinence 45.8 41.9 43.2 47.6 47.1 46.3 46.8 Severe fatigue 62.4 59.2 57.9 63.0 66.6 62.7 62.6 Anorexia 64.0 62.1 61.1 62.1 65.0 67.6 64.1 Frequent vomiting 11.8 11.4 11.4 11.5 11.4 13.5 11.1

Comorbid conditions Depression 26.4 25.9 27.8 27.1 25.7 26.1 26.0 Arthritis 67.6 57.8 65.3 70.5 69.0 70.4 71.2

2+ ADLs (0‐6 scale) 70.6 66.8 67.8 70.5 73.6 71.2 72.7

Health care utilization Median hospital nights in last two years 7 5 7 6 7 7 7 of life (IQR) (0‐20) (0‐19) (0‐21) (0‐20) (0‐21) (0‐20) (0‐20) Nursing home residency at time of death 28.0 28.8 29.9 29.2 26.9 29.2 24.2

Median hours of informal care received per 15 13 16 14 18 14 21 week in last two years of life (IQR) (0‐93) (0‐75) (0‐90) (0‐88) (0‐106) (0‐111) (1‐103)

Highest level of education Some high school or less 55.4 60.2 60.3 54.1 54.2 52.3 52.5 High school graduate 29.4 27.8 26.2 30.2 30.3 30.4 30.8 Some college or more 15.2 12.1 13.5 15.8 15.5 17.2 16.7

83,000 72,000 81,000 94,000 89,000 83,000 102,000 Median household wealth (IQR), $ (8,000‐ (7,000‐ (8,000‐ (8,000‐ (3,000‐ (6,000‐ (8,000‐ 273,000) 209,000) 220,000) 259,000) 265,000) 330,000) 336,000)

20,000 17,000 18,000 18,000 21,000 21,000 24,000 Median income (IQR), $ (11,000‐ (10,000‐ (10,000‐ (11,000‐ (12,000‐ (12,000‐ (13,000‐ 36,000) 32,000) 34,000) 32,000) 39,000) 38,000) 45,000)

Relationship of proxy to decedent Spouse 32.9 34.0 33.7 30.3 36.2 31.8 31.3 Son or daughter 44.8 42.9 41.9 44.0 43.7 46.5 49.0 Other 22.3 23.1 24.4 25.8 20.1 21.7 19.6

Mean months between decedent’s death 11.3 11.5 12.6 11.0 10.9 10.7 11.5 and proxy’s interview (95% CI) (11.1‐11.5) (11.1‐11.9) (12.1‐13.0) (10.5‐11.5) (10.4‐11.3) (10.2‐11.1) (11.1‐11.9)

Highest spouse level of education (if spouse was proxy) Some high school or less 43.4 42.4 45.4 44.5 42.2 45.3 41.0 High school graduate 37.1 41.0 36.7 39.7 38.7 33.3 33.5 Some college or more 19.5 16.6 17.9 15.8 19.1 21.4 25.4

Interview conducted in English 98.3 98.4 98.2 98.4 98.2 98.3 98.1 a Reported values account for complex survey design b Percentages are rounded and may not sum to 100%

Table 2. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 in the Entire Populationa

% of Decedents (95% CIs) Year of Death Models adjusted for demographics and clinical, psychological, social, and proxy characteristicsb Outcome 1998 2000 2002 2004 2006 2008 2010 48.7 49.3 49.9 50.5 51.1 51.7 52.4 Moderate or severe pain (45.8, 51.6) (47.0, 51.5) (48.2, 51.6) (49.0, 52.0) (49.5, 52.7) (49.6, 53.8) (49.6, 55.2) 54.3 55.4 56.5 57.5 58.6 59.7 60.8 Any pain (51.6, 57.1) (53.3, 57.5) (54.9, 58.1) (56.2, 58.9) (57.1, 60.1) (57.7, 61.6) (58.2, 63.4) 45.0 47.0 49.0 50.9 52.9 54.9 57.0 Depression (42.3, 47.7) (44.9, 49.1) (47.3, 50.6) (49.4, 52.4) (51.1, 54.7) (52.5, 57.2) (53.9, 60.0) 41.1 43.2 45.3 47.4 49.5 51.7 53.9 Periodic confusion (38.5, 43.6) (41.2, 45.2) (43.8, 46.8) (46.3, 48.5) (48.4, 50.7) (50.2, 53.1) (51.9, 56.0) 50.2 50.9 51.7 52.4 53.2 54.0 54.8 Dyspnea (48.1, 52.2) (49.4, 52.5) (50.5, 52.9) (51.3, 53.6) (51.8, 54.6) (52.1, 55.8) (52.3, 57.2) 43.0 43.8 44.7 45.5 46.4 47.2 48.1 Incontinence (40.0, 46.0) (41.6, 46.1) (43.1, 46.2) (44.4, 46.7) (45.0, 47.7) (45.2, 49.2) (45.3, 51.0) 60.7 61.2 61.7 62.2 62.7 63.2 63.7 Severe fatigue (58.0, 63.4) (59.2, 63.2) (60.3, 63.1) (61.0, 63.4) (61.3, 64.1) (61.2, 65.1) (61.0, 66.4) 62.2 62.7 63.2 63.8 64.3 64.8 65.4 Anorexia (59.5, 64.9) (60.7, 64.7) (61.8, 64.6) (62.5, 65.0) (62.7, 65.9) (62.5, 67.1) (62.3, 68.4) 12.3 12.1 11.9 11.7 11.6 11.4 11.2 Frequent vomiting (10.4, 14.2) (10.7, 13.5) (10.9, 12.9) (11.0, 12.5) (10.8, 12.4) (10.3, 12.5) (9.7, 12.7) Minimally adjusted modelsc Outcome 1998 2000 2002 2004 2006 2008 2010 45.7 47.3 48.8 50.4 52.0 53.6 55.3 Moderate or severe pain (42.9, 48.5) (45.1, 49.4) (47.1, 50.6) (48.8, 52.0) (50.2, 53.8) (51.2, 55.9) (52.3, 58.3) 51.5 53.5 55.5 57.5 59.5 61.4 63.4 Any pain (48.8, 54.1) (51.4, 55.5) (53.9, 57.1) (56.0, 59.0) (57.7, 61.2) (59.1, 63.7) (60.6, 66.3) 44.7 46.8 48.8 50.9 53.0 55.0 57.2 Depression (42.2, 47.2) (44.9, 48.7) (47.3, 50.4) (49.4, 52.5) (51.1, 54.9) (52.6, 57.5) (54.1, 60.4) Periodic confusion 41.6 43.5 45.5 47.5 49.5 51.4 53.6

(38.9, 44.3) (41.4, 45.6) (43.9, 47.1) (46.2, 48.7) (48.1, 50.8) (49.6, 53.3) (51.1, 56.1) 49.2 50.3 51.3 52.4 53.4 54.5 55.6 Dyspnea (47.1, 51.3) (48.7, 51.8) (50.1, 52.6) (51.2, 53.6) (51.9, 55.0) (52.5, 56.5) (53.0, 58.3) 43.3 44.1 44.8 45.6 46.3 47.0 47.8 Incontinence (40.4, 46.2) (42.0, 46.2) (43.4, 46.3) (44.4, 46.7) (44.8, 47.8) (44.8, 49.3) (44.8, 50.9) 59.5 60.4 61.3 62.1 63.0 63.8 64.7 Severe fatigue (56.7, 62.4) (58.4, 62.5) (59.8, 62.7) (60.9, 63.4) (61.4, 64.6) (61.6, 66.1) (61.7, 67.7) 61.0 61.9 62.8 63.7 64.6 65.4 66.3 Anorexia (58.4, 63.7) (60.0, 63.9) (61.4, 64.3) (62.3, 65.1) (62.8, 66.3) (63.1, 67.7) (63.3, 69.4) 11.4 11.5 11.6 11.7 11.8 11.9 12.0 Frequent vomiting (9.6, 13.2) (10.1, 12.8) (10.6, 12.6) (10.9, 12.4) (11.0, 12.6) (10.8, 13.0) (10.4, 13.5) a Analysis accounts for complex survey design b Adjusted prevalence is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, number of nights spent in a hospital in the last two years of life, nursing home residency at the time of death and length of nursing home residency prior to death (regardless of site of death), number of ADL dependencies (0‐6 scale), diagnosis of arthritis (pain models only), depression reported in interview wave prior to death, highest level of education, household wealth, household income, proxy relationship to decedent, time elapsed between decedent’s death and proxy’s interview, average hours per week of care given to decedent from all informal carers, highest level of education if proxy was spouse, English/Spanish preference for interview, and year that the decedent entered the HRS sample c Adjusted prevalence is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, and time elapsed between decedent’s death and proxy’s interview

Table 3. Adjusted Total Percent Change in Prevalence of Symptoms between 1998 and 2010a

Adjusted % Change (95% CIs) Models adjusted for demographics and clinical, psychological, social, and proxy characteristicsb CHF or Chronic Lung Entire Population Cancer Frailty Sudden Death Outcome Disease (n = 7,204) (n = 1,546) (n = 1,175) (n = 1,161) (n = 2,293) 7.6 4.8 4.2 6.8 ‐3.0 Moderate or severe pain (‐1.9, 18.4) (‐10.9, 24.0) (‐10.6, 23.3) (‐21.8, 43.9) (‐26.2, 28.5) 11.9 7.9 12.0 12.4 8.0 Any pain (3.1, 21.4) (‐6.7, 25.5) (‐2.0, 29.0) (‐13.3, 45.5) (‐13.9, 35.7) 26.6 8.8 27.0 39.4 17.0 Depression (14.5, 40.1) (‐11.6, 33.4) (8.1, 49.3) (9.9, 79.8) (‐10.4, 53.0) 31.3 26.3 24.9 20.3 45.7 Periodic confusion (18.6, 45.1) (‐1.6, 61.1) (6.0, 47.6) (5.9, 39.1) (5.9, 106.1) 9.2 4.4 0.5 8.9 36.7 Dyspnea (‐1.0, 19.9) (‐15.8, 27.0) (‐8.7, 11.8) (‐28.1, 56.2) (2.3, 85.9) 11.9 ‐4.4 10.0 2.8 29.3 Incontinence (1.0, 23.6) (‐26.1, 21.7) (‐7.2, 30.5) (‐14.1, 21.7) (‐5.1, 82.4) 4.9 7.0 ‐2.3 ‐1.8 16.4 Severe fatigue (‐2.9, 13.7) (‐5.5, 21.4) (‐13.2, 10.6) (‐22.1, 25.4) (‐10.5, 51.7) 5.1 7.4 0.8 ‐7.5 13.9 Anorexia (‐2.4, 13.2) (‐3.7, 18.8) (‐10.6, 13.7) (‐23.2, 10.4) (‐16.4, 50.3) ‐8.8 11.4 ‐30.5 ‐26.4 72.5 Frequent vomiting (‐31.2, 21.5) (‐26.0, 72.8) (‐60.2, 26.8) (‐66.0, 69.4) (‐30.4, 305.7) Minimally adjusted modelsc CHF or Chronic Lung Entire Population Cancer Frailty Sudden Death Outcome Disease (n = 7,204) (n = 1,546) (n = 1,175) (n = 1,161) (n = 2,293) 20.9 11.2 15.9 13.3 24.3 Moderate or severe pain (9.7, 32.9) (‐5.8, 31.2) (‐1.3, 37.0) (‐16.6, 53.9) (‐6.6, 66.4) 23.3 13.0 22.5 18.4 32.8 Any pain (13.4, 33.9) (‐2.6, 31.3) (7.0, 41.7) (‐8.5, 55.2) (5.3, 68.9)

28.0 11.5 25.0 37.6 23.1 Depression (15.7, 42.0) (‐9.8, 36.7) (6.4, 46.7) (8.2, 78.0) (‐7.8, 64.6) 28.8 26.7 16.5 20.1 49.4 Periodic confusion (15.6, 43.6) (‐2.4, 61.5) (‐2.4, 39.0) (5.0, 39.0) (3.4, 114.8) 13.1 5.0 2.5 9.1 42.4 Dyspnea (2.5, 24.5) (‐15.4, 27.9) (‐7.0, 13.1) (‐26.2, 56.1) (6.6, 92.6) 10.4 ‐9.9 3.0 ‐1.9 48.2 Incontinence (‐0.6, 23.3) (‐29.6, 16.6) (‐14.4, 23.4) (‐18.7, 16.8) (4.9, 113.5) 8.7 3.4 0.7 ‐2.2 21.3 Severe fatigue (0.3, 17.6) (‐8.8, 17.7) (‐10.8, 14.1) (‐23.1, 24.3) (‐6.6, 59.3) 8.7 8.1 1.2 ‐6.0 21.2 Anorexia (0.5, 17.5) (‐3.3, 20.4) (‐10.5, 14.3) (‐21.9, 12.0) (‐11.6, 63.9) 5.1 25.7 ‐29.4 ‐16.3 105.5 Frequent vomiting (‐20.8, 40.3) (‐16.9, 91.4) (‐60.9, 27.2) (‐63.2, 93.6) (‐13.9, 457.9) a Analysis accounts for complex survey design b Adjusted percent change is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, number of nights spent in a hospital in the last two years of life, nursing home residency at the time of death and length of nursing home residency prior to death (regardless of site of death), number of ADL dependencies (0‐6 scale), diagnosis of arthritis (pain models only), depression reported in interview wave prior to death, highest level of education, household wealth, household income, proxy relationship to decedent, time elapsed between decedent’s death and proxy’s interview, average hours per week of care given to decedent from all informal carers, highest level of education if proxy was spouse, English/Spanish preference for interview, and year that the decedent entered the HRS sample. Confidence intervals are bootstrapped. c Adjusted percent change is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, and time elapsed between decedent’s death and proxy’s interview. Confidence intervals are bootstrapped.

Appendix Table 1. HRS Survey Questions Used to Construct Symptom Outcomes

Outcome HRS Variable Question to Proxy Was [he/she] often troubled with pain? I want a general idea of [his/her] pain level during the Presence of pain C104 last year or so of life. Degree of pain C105 How bad was the pain most of the time: mild, moderate or severe? Was there a period of at least one month during the last year of [his/her] life when [he/she] had Dyspnea C198 the following problems: difficulty breathing? Was there a period of at least one month during the last year of [his/her] life when [he/she] had Anorexia C199 very little appetite or desire for food? Frequent Was there a period of at least one month during the last year of [his/her] life when [he/she] had C200 vomiting frequent vomiting? Was there a period of at least one month during the last year of [his/her] life when [he/she] had Depression C202 depression? Periodic Was there a period of at least one month during the last year of [his/her] life when [he/she] had C203 confusion periodic confusion? Was there a period of at least one month during the last year of [his/her] life when [he/she] had Severe fatigue C204 severe fatigue or exhaustion? Was there a period of at least one month during the last year of [his/her] life when [he/she] had Incontinence C208 loss of control of bowel or bladder?

Appendix Table 2. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 among Decedent Categories (Fully Adjusted Models)a

% of Decedents (95% CIs)b Year of Death Category Outcome 1998 2000 2002 2004 2006 2008 2010 Moderate or 60.3 60.8 61.3 61.8 62.2 62.7 63.2 severe pain (54.4, 66.2) (56.5, 65.1) (58.1, 64.5) (58.8, 64.7) (58.5, 66.0) (57.5, 67.9) (56.3, 70.1) 63.7 64.6 65.4 66.2 67.1 67.9 68.7 Any pain (57.8, 69.6) (60.3, 68.9) (62.4, 68.4) (63.7, 68.8) (64.0, 70.2) (63.6, 72.2) (62.9, 74.6) 51.1 51.9 52.6 53.3 54.1 54.8 55.6 Depression (46.3, 56.0) (48.3, 55.5) (49.8, 55.4) (50.4, 56.3) (50.3, 57.9) (49.7, 59.9) (48.9, 62.3) Periodic 38.1 39.7 41.4 43.0 44.7 46.3 48.1 confusion (33.0, 43.2) (35.9, 43.6) (38.5, 44.3) (40.3, 45.7) (41.3, 48.0) (41.7, 50.9) (42.0, 54.3) 52.8 53.2 53.5 53.9 54.3 54.7 55.1 Cancer Dyspnea (48.0, 57.5) (49.5, 56.8) (50.6, 56.5) (51.0, 56.8) (50.7, 57.9) (50.0, 59.3) (49.1, 61.1) 43.0 42.7 42.4 42.1 41.8 41.5 41.1 Incontinence (38.1, 47.9) (38.8, 46.6) (39.3, 45.5) (39.1, 45.0) (38.4, 45.1) (37.2, 45.7) (35.7, 46.6) Severe 75.5 76.4 77.3 78.2 79.1 79.9 80.8 fatigue (69.7, 81.2) (72.3, 80.5) (74.7, 80.0) (76.2, 80.2) (76.6, 81.6) (76.3, 83.5) (75.8, 85.7) 77.3 78.4 79.4 80.3 81.2 82.1 83.1 Anorexia (72.9, 81.7) (75.1, 81.6) (76.9, 81.8) (78.0, 82.6) (78.5, 84.0) (78.6, 85.7) (78.6, 87.5) Frequent 20.7 21.0 21.4 21.8 22.2 22.6 23.0 vomiting (15.3, 26.0) (16.9, 25.2) (18.3, 24.6) (19.3, 24.3) (19.6, 24.8) (19.1, 26.1) (18.2, 27.8) Moderate or 52.4 52.8 53.1 53.5 53.8 54.2 54.6 severe pain (47.7, 57.1) (49.2, 56.3) (50.5, 55.7) (51.3, 55.6) (51.4, 56.3) (50.9, 57.6) (50.0, 59.1) 57.3 58.5 59.6 60.8 61.9 63.0 64.2 Any pain CHF or (53.4, 61.2) (55.5, 61.5) (57.3, 61.9) (58.7, 62.8) (59.4, 64.3) (59.9, 66.1) (60.1, 68.2) chronic lung 50.0 52.3 54.6 56.8 59.0 61.2 63.5 Depression disease (45.2, 54.9) (48.5, 56.1) (51.7, 57.5) (54.4, 59.3) (56.4, 61.7) (57.9, 64.5) (59.3, 67.8) Periodic 45.0 46.8 48.7 50.6 52.4 54.2 56.2 confusion (40.5, 49.5) (43.4, 50.3) (46.1, 51.3) (48.4, 52.7) (50.1, 54.7) (51.2, 57.3) (52.1, 60.3) Dyspnea 75.1 75.1 75.2 75.3 75.3 75.4 75.5

(70.8, 79.4) (71.9, 78.4) (72.8, 77.6) (73.4, 77.2) (73.2, 77.4) (72.6, 78.2) (71.6, 79.3) 46.6 47.4 48.2 48.9 49.7 50.5 51.3 Incontinence (41.9, 51.3) (43.9, 50.9) (45.6, 50.7) (47.0, 50.9) (47.6, 51.8) (47.5, 53.4) (47.2, 55.4) Severe 69.3 69.1 68.8 68.5 68.3 68.0 67.7 fatigue (65.1, 73.6) (65.8, 72.3) (66.4, 71.2) (66.5, 70.6) (65.9, 70.6) (64.8, 71.2) (63.4, 72.0) 68.7 68.8 68.9 68.9 69.0 69.1 69.2 Anorexia (63.8, 73.5) (65.0, 72.5) (66.0, 71.7) (66.6, 71.3) (66.5, 71.6) (65.8, 72.4) (64.8, 73.6) Frequent 12.1 11.4 10.7 10.1 9.5 9.0 8.4 vomiting (8.5, 15.6) (8.9, 13.9) (9.0, 12.4) (8.8, 11.4) (8.1, 11.0) (7.0, 10.9) (5.8, 11.0) Moderate or 37.7 38.1 38.5 38.9 39.4 39.8 40.2 severe pain (30.3, 45.0) (32.6, 43.6) (34.7, 42.4) (36.1, 41.8) (36.1, 42.6) (35.1, 44.5) (33.6, 46.9) 43.9 44.8 45.7 46.6 47.5 48.4 49.3 Any pain (37.2, 50.6) (39.5, 50.0) (41.6, 49.7) (43.3, 49.9) (44.1, 50.8) (44.2, 52.6) (43.8, 54.9) 42.3 45.0 47.8 50.6 53.3 56.1 59.0 Depression (35.8, 48.8) (39.9, 50.1) (43.9, 51.6) (47.5, 53.6) (50.2, 56.4) (52.1, 60.0) (53.6, 64.3) Periodic 69.3 71.9 74.5 76.9 79.1 81.2 83.3 confusion (62.0, 76.5) (66.6, 77.3) (70.6, 78.4) (73.9, 79.9) (76.3, 82.0) (77.9, 84.6) (79.3, 87.4) 28.8 29.2 29.7 30.1 30.5 30.9 31.4 Frailty Dyspnea (22.6, 35.1) (24.5, 34.0) (26.3, 33.0) (27.7, 32.5) (28.0, 33.0) (27.4, 34.5) (26.2, 36.6) 65.3 65.6 65.9 66.2 66.5 66.8 67.2 Incontinence (59.2, 71.4) (60.8, 70.4) (62.1, 69.7) (63.0, 69.4) (63.2, 69.9) (62.7, 71.0) (61.8, 72.5) Severe 52.9 52.8 52.6 52.5 52.3 52.1 52.0 fatigue (46.5, 59.4) (48.0, 57.5) (49.2, 56.1) (49.4, 55.5) (48.4, 56.2) (46.7, 57.6) (44.6, 59.3) 68.6 67.7 66.9 66.1 65.2 64.3 63.4 Anorexia (60.0, 77.1) (61.3, 74.2) (62.3, 71.5) (62.7, 69.4) (61.5, 68.9) (58.9, 69.7) (55.5, 71.2) Frequent 7.5 7.1 6.8 6.5 6.1 5.8 5.5 vomiting (4.0, 11.0) (4.7, 9.6) (5.1, 8.5) (5.0, 7.9) (4.3, 8.0) (3.4, 8.3) (2.4, 8.7) Moderate or 42.0 41.8 41.6 41.4 41.2 40.9 40.7 severe pain (35.1, 48.8) (36.6, 46.9) (37.9, 45.2) (38.6, 44.1) (38.2, 44.1) (36.8, 45.1) (34.9, 46.6) 49.0 49.6 50.3 50.9 51.6 52.2 52.9 Sudden death Any pain (42.6, 55.4) (44.9, 54.3) (47.0, 53.6) (48.2, 53.6) (48.1, 55.0) (47.3, 57.1) (46.1, 59.7) 39.5 40.5 41.6 42.7 43.8 45.0 46.2 Depression (33.8, 45.1) (36.2, 44.9) (38.4, 44.9) (39.8, 45.6) (40.4, 47.3) (40.3, 49.6) (39.9, 52.5)

Periodic 26.0 27.8 29.7 31.7 33.6 35.7 37.9 confusion (20.1, 31.9) (23.1, 32.6) (26.1, 33.3) (29.0, 34.3) (31.2, 36.1) (32.4, 38.9) (33.2, 42.7) 34.7 36.7 38.7 40.8 42.9 45.1 47.4 Dyspnea (29.6, 39.7) (32.6, 40.7) (35.4, 42.0) (37.7, 43.8) (39.4, 46.4) (40.6, 49.5) (41.6, 53.2) 28.6 29.9 31.2 32.6 34.0 35.4 37.0 Incontinence (21.8, 35.4) (24.7, 35.1) (27.5, 35.0) (29.9, 35.4) (31.0, 37.0) (31.1, 39.7) (30.7, 43.2) Severe 43.4 44.6 45.8 46.9 48.1 49.3 50.5 fatigue (37.3, 49.6) (40.2, 49.0) (42.8, 48.7) (44.7, 49.2) (45.1, 51.1) (44.7, 53.8) (44.0, 57.0) 37.8 38.7 39.5 40.4 41.3 42.1 43.1 Anorexia (32.3, 43.4) (34.6, 42.8) (36.4, 42.7) (37.2, 43.6) (37.0, 45.5) (36.3, 47.9) (35.4, 50.8) Frequent 5.3 5.9 6.4 7.0 7.7 8.4 9.2 vomiting (2.6, 8.1) (3.6, 8.1) (4.8, 8.1) (5.8, 8.3) (6.3, 9.1) (6.1, 10.7) (5.6, 12.8) a Analysis accounts for complex survey design b Adjusted prevalence is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, number of nights spent in a hospital in the last two years of life, nursing home residency at the time of death and length of nursing home residency prior to death (regardless of site of death), number of ADL dependencies (0‐6 scale), diagnosis of arthritis (pain models only), depression reported in interview wave prior to death, highest level of education, household wealth, household income, proxy relationship to decedent, time elapsed between decedent’s death and proxy’s interview, average hours per week of care given to decedent from all informal carers, highest level of education if proxy was spouse, English/Spanish preference for interview, and year that the decedent entered the HRS sample

Appendix Table 3. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 among Decedent Categories (Minimally Adjusted Models)a

% of Decedents (95% CIs)b Year of Death Category Outcome 1998 2000 2002 2004 2006 2008 2010 Moderate or 58.4 59.5 60.6 61.7 62.8 63.9 65.0 severe pain (52.5, 64.4) (55.2, 63.8) (57.5, 63.8) (58.7, 64.8) (58.8, 66.8) (58.5, 69.2) (57.9, 72.1) 62.1 63.5 64.9 66.2 67.5 68.8 70.2 Any pain (56.3, 67.9) (59.3, 67.7) (61.9, 67.9) (63.6, 68.8) (64.3, 70.8) (64.4, 73.3) (64.3, 76.1) 50.4 51.4 52.3 53.3 54.3 55.2 56.2 Depression (45.7, 55.1) (47.8, 54.9) (49.5, 55.2) (50.4, 56.2) (50.6, 57.9) (50.3, 60.1) (49.9, 62.5) Periodic 38.0 39.7 41.3 43.0 44.7 46.4 48.2 confusion (33.3, 42.8) (36.1, 43.2) (38.6, 44.0) (40.4, 45.6) (41.3, 48.0) (41.8, 51.0) (42.1, 54.4) 52.6 53.0 53.5 53.9 54.3 54.8 55.2 Cancer Dyspnea (47.8, 57.5) (49.3, 56.8) (50.5, 56.5) (50.9, 56.9) (50.7, 58.0) (50.0, 59.5) (49.1, 61.4) 44.3 43.6 42.9 42.1 41.4 40.7 39.9 Incontinence (39.3, 49.4) (39.6, 47.5) (39.7, 46.1) (39.1, 45.2) (37.9, 44.9) (36.3, 45.1) (34.4, 45.5) Severe 77.0 77.4 77.8 78.3 78.7 79.1 79.6 fatigue (71.4, 82.5) (73.5, 81.3) (75.2, 80.4) (76.1, 80.4) (75.8, 81.6) (74.9, 83.3) (73.8, 85.3) 77.0 78.2 79.3 80.3 81.3 82.3 83.3 Anorexia (72.6, 81.5) (75.0, 81.4) (76.9, 81.7) (77.9, 82.7) (78.4, 84.3) (78.5, 86.1) (78.5, 88.1) Frequent 19.3 20.1 20.9 21.7 22.5 23.4 24.3 vomiting (14.4, 24.3) (16.2, 24.0) (17.9, 23.9) (19.2, 24.3) (19.8, 25.3) (19.7, 27.0) (19.3, 29.4) Moderate or 49.4 50.7 52.0 53.3 54.6 55.9 57.3 severe pain (45.0, 53.8) (47.5, 54.0) (49.6, 54.4) (51.0, 55.6) (51.5, 57.7) (51.7, 60.1) (51.7, 62.9) 54.4 56.5 58.6 60.6 62.6 64.6 66.6 Any pain CHF or (50.5, 58.4) (53.5, 59.5) (56.2, 60.9) (58.2, 63.0) (59.6, 65.6) (60.7, 68.4) (61.8, 71.5) chronic lung 50.5 52.6 54.8 56.9 58.9 61.0 63.1 Depression disease (45.4, 55.6) (48.7, 56.6) (51.7, 57.8) (54.3, 59.4) (56.2, 61.7) (57.6, 64.4) (58.7, 67.6) Periodic 46.9 48.2 49.5 50.7 52.0 53.3 54.7 confusion (41.7, 52.1) (44.3, 52.1) (46.6, 52.4) (48.4, 53.1) (49.4, 54.6) (49.8, 56.8) (49.9, 59.4) Dyspnea 74.3 74.6 74.9 75.2 75.5 75.8 76.1

(69.8, 78.8) (71.2, 78.0) (72.4, 77.5) (73.2, 77.2) (73.3, 77.7) (73.0, 78.6) (72.3, 79.9) 48.4 48.6 48.9 49.1 49.3 49.6 49.8 Incontinence (43.1, 53.6) (44.7, 52.6) (46.0, 51.7) (46.9, 51.3) (46.9, 51.8) (46.1, 53.1) (45.0, 54.7) Severe 68.2 68.3 68.4 68.4 68.5 68.6 68.7 fatigue (64.0, 72.4) (65.1, 71.5) (66.0, 70.7) (66.5, 70.4) (66.2, 70.8) (65.5, 71.7) (64.4, 73.0) 68.5 68.7 68.8 68.9 69.1 69.2 69.3 Anorexia (63.6, 73.4) (64.9, 72.4) (65.9, 71.6) (66.6, 71.3) (66.6, 71.6) (65.9, 72.5) (65.0, 73.7) Frequent 12.0 11.3 10.7 10.1 9.5 9.0 8.5 vomiting (8.2, 15.7) (8.7, 14.0) (8.9, 12.5) (8.8, 11.4) (8.0, 11.1) (7.0, 11.1) (5.8, 11.2) Moderate or 36.5 37.3 38.1 38.9 39.7 40.5 41.4 severe pain (29.9, 43.1) (32.3, 42.3) (34.4, 41.8) (35.7, 42.0) (35.9, 43.5) (35.3, 45.7) (34.3, 48.5) 42.7 43.9 45.2 46.5 47.8 49.1 50.5 Any pain (36.7, 48.6) (39.2, 48.7) (41.4, 49.0) (43.0, 50.0) (43.9, 51.7) (44.2, 54.0) (44.2, 56.8) 42.6 45.3 47.9 50.6 53.2 55.9 58.7 Depression (36.7, 48.6) (40.7, 49.9) (44.4, 51.4) (47.7, 53.5) (50.1, 56.4) (51.8, 60.0) (53.2, 64.2) Periodic 69.3 72.0 74.5 76.9 79.2 81.2 83.3 confusion (62.3, 76.3) (67.0, 77.0) (71.0, 78.1) (74.1, 79.8) (76.0, 82.3) (77.3, 85.1) (78.5, 88.1) 28.8 29.2 29.6 30.1 30.5 30.9 31.4 Frailty Dyspnea (22.4, 35.2) (24.3, 34.1) (26.1, 33.2) (27.4, 32.8) (27.6, 33.4) (27.0, 34.9) (25.8, 37.1) 67.0 66.7 66.5 66.3 66.1 65.9 65.7 Incontinence (59.9, 74.0) (61.1, 72.4) (62.0, 71.0) (62.5, 70.2) (62.1, 70.1) (61.0, 70.8) (59.4, 72.0) Severe 53.0 52.9 52.7 52.5 52.3 52.1 51.9 fatigue (46.4, 59.7) (48.1, 57.6) 49.4, 56.0) (49.4, 55.5) (48.1, 56.4) (46.1, 58.0) (43.8, 60.0) 68.0 67.3 66.7 66.0 65.3 64.7 63.9 Anorexia (59.9, 76.1) (61.2, 73.5) (62.2, 71.1) (62.5, 69.5) (61.5, 69.2) (59.2, 70.1) (56.3, 71.6) Frequent 6.7 6.5 6.4 6.2 6.0 5.8 5.6 vomiting (3.6, 9.9) (4.2, 8.9) (4.6, 8.1) (4.5, 7.8) (4.1, 7.9) (3.3, 8.3) (2.5, 8.8) Moderate or 37.0 38.4 39.9 41.3 42.8 44.3 45.9 severe pain (29.6, 44.3) (32.9, 43.9) (36.0, 43.7) (38.4, 44.3) (39.3, 46.3) (39.2, 49.4) (38.7, 53.2) 43.8 46.2 48.5 50.9 53.3 55.6 58.2 Sudden death Any pain (37.2, 50.5) (41.3, 51.0) (45.2, 51.9) (47.9, 53.9) (49.3, 57.3) (49.9, 61.4) (50.4, 65.9) 38.3 39.8 41.2 42.7 44.1 45.6 47.2 Depression (32.6, 44.0) (35.4, 44.1) (37.7, 44.7) (39.2, 46.1) (39.9, 48.4) (39.9, 51.3) (39.8, 54.6)

Periodic 25.7 27.6 29.6 31.7 33.8 36.0 38.4 confusion (20.2, 31.2) (23.2, 32.1) (26.2, 33.0) (28.8, 34.5) (30.7, 36.8) (31.8, 40.1) (32.6, 44.2) 33.8 36.1 38.4 40.7 43.1 45.6 48.2 Dyspnea (28.6, 39.1) (31.8, 40.4) (34.8, 42.0) (37.4, 44.1) (39.4, 46.9) (40.9, 50.2) (42.2, 54.2) 26.5 28.4 30.4 32.5 34.6 36.8 39.3 Incontinence (19.7, 33.3) (23.0, 33.8) (26.4, 34.5) (29.4, 35.6) (31.4, 37.8) (32.4, 41.3) (32.9, 45.6) Severe 42.4 43.9 45.4 46.9 48.4 49.9 51.5 fatigue (36.5, 48.3) (39.6, 48.2) (42.3, 48.4) (44.1, 49.6) (44.6, 52.1) (44.5, 55.2) (44.2, 58.7) 36.6 37.8 39.1 40.3 41.6 42.9 44.3 Anorexia (31.3, 41.9) (33.8, 41.8) (35.8, 42.3) (36.8, 43.9) (36.9, 46.3) (36.6, 49.2) (36.1, 52.6) Frequent 4.9 5.5 6.2 7.0 7.8 8.8 10.0 vomiting (2.4, 7.3) (3.4, 7.6) (4.6, 7.8) (5.7, 8.3) (6.3, 9.4) (6.4, 11.3) (6.1, 13.9) a Analysis accounts for complex survey design b Adjusted prevalence is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, and time elapsed between decedent’s death and proxy’s interview

Appendix Table 4. Adjusted Estimated Prevalence of Symptoms between 1998 and 2010 among Decedent Categories (Fully Adjusted Models Excluding 2002 Survey Wave)a

% of Decedents (95% CIs)b Year of Death Category Outcome 1998 2000 2002 2004 2006 2008 2010 Moderate or 49.5 50.0 50.5 51.0 51.4 51.9 52.4 severe pain (46.1, 52.9) (47.4, 52.6) (48.5, 52.5) (49.3, 52.6) (49.8, 53.1) (49.8, 54.1) (49.6, 55.3) 55.3 56.2 57.1 58.0 58.9 59.8 60.8 Any pain (51.9, 58.6) (53.6, 58.7) (55.2, 59.0) (56.5, 59.5) (57.4, 60.4) (57.8, 61.8) (58.1, 63.5) 45.5 47.3 49.1 50.9 52.7 54.5 56.4 Depression (42.6, 48.4) (45.1, 49.5) (47.4, 50.8) (49.4, 52.4) (51.0, 54.4) (52.2, 56.8) (53.3, 59.5) Periodic 42.0 43.9 45.8 47.7 49.6 51.5 53.5 confusion (38.9, 45.1) (41.5, 46.3) (44.0, 47.5) (46.4, 48.9) (48.5, 50.7) (50.0, 53.0) (51.4, 55.7) Entire 50.4 51.1 51.7 52.4 53.1 53.7 54.5 Dyspnea population (47.9, 52.8) (49.2, 52.9) (50.3, 53.2) (51.2, 53.6) (51.7, 54.5) (51.9, 55.6) (52.0, 56.9) 43.9 44.5 45.1 45.7 46.3 46.9 47.6 Incontinence (40.5, 47.4) (41.9, 47.1) (43.2, 47.0) (44.3, 47.1) (44.9, 47.7) (45.0, 48.9) (44.8, 50.4) Severe 61.9 62.2 62.4 62.7 62.9 63.2 63.5 fatigue (59.2, 64.7) (60.1, 64.2) (61.0, 63.9) (61.5, 63.9) (61.5, 64.4) (61.2, 65.2) (60.7, 66.2) 62.6 63.1 63.5 64.0 64.5 64.9 65.4 Anorexia (59.5, 65.7) (60.8, 65.4) (61.9, 65.2) (62.6, 65.4) (62.8, 66.1) (62.7, 67.1) (62.3, 68.4) Frequent 12.6 12.4 12.1 11.9 11.6 11.4 11.1 vomiting (10.3, 15.0) (10.6, 14.1) (10.9, 13.4) (11.0, 12.7) (10.8, 12.4) (10.3, 12.4) (9.5, 12.6) Moderate or 59.8 60.5 61.1 61.7 62.4 63.0 63.7 severe pain (52.3, 67.4) (54.9, 66.0) (57.1, 65.1) (58.6, 64.9) (58.7, 66.1) (57.8, 68.2) (56.6, 70.8) 62.1 63.3 64.6 65.8 66.9 68.1 69.3 Any pain (54.9, 69.3) (58.0, 68.6) (60.9, 68.2) (63.0, 68.5) (63.8, 70.0) (63.8, 72.4) (63.3, 75.3) Cancer 51.6 52.1 52.7 53.2 53.8 54.3 54.9 Depression (46.1, 57.0) (48.0, 56.2) (49.5, 55.8) (50.2, 56.2) (50.0, 57.6) (49.3, 59.4) (48.3, 61.6) Periodic 39.3 40.6 41.9 43.2 44.5 45.8 47.2 confusion (33.9, 44.6) (36.5, 44.6) (38.8, 44.9) (40.5, 45.9) (41.2, 47.8) (41.3, 50.3) (41.1, 53.3) Dyspnea 50.5 51.3 52.0 52.7 53.5 54.2 55.0

(45.1, 55.9) (47.1, 55.5) (48.7, 55.3) (49.6, 55.8) (49.9, 57.1) (49.6, 58.8) (49.0, 61.0) 42.2 42.0 41.8 41.5 41.3 41.1 40.8 Incontinence (36.6, 47.9) (37.5, 46.5) (38.2, 45.3) (38.4, 44.7) (37.9, 44.7) (36.8, 45.3) (35.4, 46.3) Severe 75.3 76.3 77.3 78.2 79.1 80.0 80.9 fatigue (69.1, 81.6) (71.8, 80.9) (74.2, 80.4) (75.9, 80.5) (76.6, 81.7) (76.5, 83.5) (76.1, 85.7) 77.1 78.2 79.3 80.3 81.3 82.3 83.2 Anorexia (72.3, 82.0) (74.6, 81.9) (76.5, 82.1) (77.8, 82.8) (78.5, 84.1) (78.7, 85.8) (78.8, 87.6) Frequent 19.9 20.4 21.0 21.6 22.2 22.8 23.4 vomiting (13.5, 26.2) (15.3, 25.5) (17.1, 24.9) (18.6, 24.6) (19.3, 25.0) (19.2, 26.3) (18.5, 28.4) Moderate or 53.1 53.3 53.5 53.8 54.0 54.2 54.5 severe pain (47.7, 58.5) (49.1, 57.6) (50.3, 56.8) (51.1, 56.4) (51.3, 56.7) (50.9, 57.6) (50.0, 58.9) 58.1 59.1 60.1 61.1 62.1 63.0 64.1 Any pain (53.1, 63.0) (55.2, 62.9) (57.1, 63.0) (58.6, 63.5) (59.5, 64.6) (59.9, 66.1) (60.0, 68.1) 49.7 52.0 54.3 56.5 58.8 60.9 63.3 Depression (44.2, 55.3) (47.7, 56.3) (51.0, 57.5) (54.0, 59.1) (56.1, 61.4) (57.7, 64.2) (58.9, 67.6) Periodic 46.2 47.8 49.4 51.0 52.7 54.3 56.0 confusion (40.1, 52.3) (43.2, 52.5) (46.1, 52.8) (48.7, 53.4) (50.4, 54.9) (51.2, 57.3) (51.6, 60.4) CHF or 75.2 75.2 75.2 75.2 75.2 75.2 75.2 chronic lung Dyspnea (70.4, 79.9) (71.5, 78.8) (72.5, 77.9) (73.1, 77.3) (73.1, 77.3) (72.5, 78.0) (71.5, 79.0) disease 50.0 50.1 50.1 50.1 50.1 50.1 50.1 Incontinence (45.0, 55.1) (46.1, 54.0) (47.2, 53.0) (47.8, 52.3) (47.9, 52.3) (47.2, 53.0) (46.1, 54.1) Severe 69.3 69.0 68.7 68.5 68.2 68.0 67.7 fatigue (65.0, 73.5) (65.8, 72.2) (66.4, 71.1) (66.5, 70.5) (65.9, 70.6) (64.7, 71.2) (63.3, 72.1) 69.4 69.4 69.3 69.3 69.3 69.2 69.2 Anorexia (63.8, 75.0) (65.0, 73.8) (66.0, 72.7) (66.7, 71.9) (66.7, 71.9) (66.0, 72.5) (64.8, 73.5) Frequent 12.8 11.9 11.0 10.2 9.5 8.8 8.1 vomiting (8.4, 17.2) (8.8, 15.0) (9.0, 13.1) (8.8, 11.7) (8.0, 11.0) (6.8, 10.8) (5.5, 10.7) Moderate or 41.4 41.1 40.8 40.5 40.2 39.9 39.6 severe pain (33.4, 49.4) (35.1, 47.2) (36.5, 45.1) (37.4, 43.6) (37.0, 43.5) (35.4, 44.5) (33.1, 46.1) 49.4 49.2 49.0 48.8 48.6 48.4 48.2 Frailty Any pain (41.6, 57.3) (43.0, 55.4) (44.3, 53.8) (45.1, 52.5) (45.2, 52.1) (44.3, 52.6) (42.6, 53.8) 44.6 46.8 49.0 51.2 53.4 55.6 58.0 Depression (38.0, 51.3) (41.6, 52.0) (45.1, 53.0) (48.1, 54.4) (50.2, 56.7) (51.5, 59.7) (52.5, 63.4)

Periodic 72.3 74.2 76.0 77.8 79.5 81.0 82.6 confusion (64.8, 79.8) (68.5, 79.9) (71.8, 80.3) (74.5, 81.1) (76.4, 82.5) (77.6, 84.5) (78.4, 86.8) 30.6 30.6 30.6 30.6 30.6 30.6 30.7 Dyspnea (23.6, 37.6) (25.3, 35.9) (26.9, 34.3) (28.0, 33.2) (28.1, 33.2) (27.0, 34.3) (25.3, 36.0) 64.2 64.7 65.1 65.6 66.1 66.6 67.1 Incontinence (57.2, 71.2) (59.3, 70.0) (61.1, 69.2) (62.4, 68.8) (62.9, 69.3) (62.4, 70.7) (61.5, 72.6) Severe 56.2 55.3 54.5 53.7 52.9 52.0 51.1 fatigue (48.4, 63.9) (49.5, 61.1) (50.3, 58.7) (50.3, 57.1) (48.9, 56.8) (46.5, 57.6) (43.5, 58.8) 70.9 69.6 68.3 67.0 65.6 64.2 62.7 Anorexia (61.9, 79.8) (62.7, 76.4) (63.5, 73.1) (63.6, 70.3) (62.0, 69.1) (58.8, 69.6) (54.6, 70.8) Frequent 7.1 6.8 6.5 6.3 6.0 5.7 5.5 vomiting (4.1, 10.2) (4.7, 9.0) (5.0, 8.1) (4.7, 7.8) (4.0, 7.9) (3.2, 8.3) (2.2, 8.7) Moderate or 40.5 40.7 40.8 40.9 41.0 41.1 41.3 severe pain (33.0, 48.1) (35.0, 46.4) (36.7, 44.8) (38.0, 43.8) (38.1, 43.9) (37.0, 45.2) (35.4, 47.2) 49.0 49.7 50.3 50.9 51.6 52.2 52.9 Any pain (41.8, 56.3) (44.3, 55.0) (46.5, 54.0) (48.0, 53.8) (48.0, 55.1) (47.2, 57.2) (45.8, 59.9) 38.7 39.8 40.9 42.0 43.1 44.3 45.5 Depression (32.3, 45.1) (34.9, 44.7) (37.3, 44.5) (39.2, 44.9) (39.9, 46.4) (39.7, 48.8) (39.2, 51.8) Periodic 26.4 28.1 29.8 31.5 33.4 35.2 37.2 confusion (19.4, 33.3) (22.5, 33.6) (25.6, 33.9) (28.6, 34.5) (30.8, 35.9) (31.8, 38.6) (32.1, 42.4) 34.6 36.6 38.7 40.9 43.0 45.2 47.6 Sudden death Dyspnea (29.0, 40.2) (32.2, 41.1) (35.2, 42.2) (37.8, 43.9) (39.6, 46.4) (40.8, 49.6) (41.7, 53.5) 28.7 29.9 31.1 32.3 33.6 34.9 36.3 Incontinence (20.8, 36.6) (23.7, 36.1) (26.6, 35.6) (29.1, 35.6) (30.6, 36.6) (30.8, 39.0) (30.2, 42.3) Severe 44.8 45.7 46.6 47.5 48.5 49.4 50.4 fatigue (37.7, 51.9) (40.4, 51.0) (43.0, 50.3) (44.9, 50.2) (45.5, 51.5) (44.9, 53.9) (44.0, 56.8) 37.5 38.4 39.2 40.1 41.0 41.9 42.8 Anorexia (30.9, 44.1) (33.5, 43.3) (35.6, 42.9) (36.8, 43.5) (36.7, 45.3) (36.0, 47.8) (34.9, 50.8) Frequent 6.4 6.7 7.0 7.4 7.8 8.2 8.7 vomiting (2.9, 9.8) (4.0, 9.4) (5.1, 9.0) (6.1, 8.8) (6.5, 9.2) (6.2, 10.3) (5.4, 11.9) a Analysis accounts for complex survey design

36 b Adjusted prevalence is predicted from regression models of each symptom on time of death, controlling for age at death, gender, race/ethnicity, number of nights spent in a hospital in the last two years of life, nursing home residency at the time of death and length of nursing home residency prior to death (regardless of site of death), number of ADL dependencies (0‐6 scale), diagnosis of arthritis (pain models only), depression reported in interview wave prior to death, highest level of education, household wealth, household income, proxy relationship to decedent, time elapsed between decedent’s death and proxy’s interview, average hours per week of care given to decedent from all informal carers, highest level of education if proxy was spouse, English/Spanish preference for interview, and year that the decedent entered the HRS sample

Appendix Table 5. Adjusted Total Percent Change in Prevalence of Symptoms between 1998 and 2010 (Fully Adjusted Models Excluding 2002 Survey Wave)a

Adjusted % Change (95% CIs)b CHF or Chronic Lung Entire Population Cancer Frailty Sudden Death Outcome Disease (n = 7,204) (n = 1,546) (n = 1,175) (n = 1,161) (n = 2,293) 6.0 6.5 2.6 ‐4.3 1.8 Moderate or severe pain (‐3.9, 18.0) (‐10.3, 28.1) (‐13.7, 22.7) (‐30.9, 31.9) (‐23.4, 40.2) 10.0 11.6 10.3 ‐2.4 7.8 Any pain (1.1, 20.5) (‐4.6, 32.0) (‐4.6, 29.1) (‐25.7, 27.0) (‐15.1, 39.1) 26.4 6.7 30.1 31.5 20.7 Depression (13.7, 40.5) (‐14.1, 31.9) (9.9, 55.4) (3.9, 72.0) (‐9.5, 61.8) 27.7 12.8 21.0 13.7 46.3 Periodic confusion (15.0, 41.6) (‐13.2, 47.6) (2.1, 45.5) (‐0.5, 31.4) (4.2, 113.4) 9.8 7.4 0.6 1.2 35.9 Dyspnea (‐0.4, 21.3) (‐14.6, 32.8) (‐9.1, 12.7) (‐33.8, 46.7) (0.8, 87.1) 9.5 ‐6.9 2.0 4.5 33.3 Incontinence (‐1.9, 21.2) (‐29.6, 22.2) (‐14.5, 21.9) (‐14.1, 26.2) (‐4.0, 95.4) 2.9 4.0 ‐1.6 ‐8.7 13.0 Severe fatigue (‐5.2, 11.5) (‐9.0, 18.3) (‐12.9, 12.4) (‐28.0, 16.3) (‐13.3, 49.4) 4.2 7.2 ‐0.2 ‐12.4 15.3 Anorexia (‐3.6, 12.3) (‐4.1, 20.1) (‐12.2, 13.5) (‐27.6, 5.7) (‐17.4, 55.5) ‐10.1 20.5 ‐32.5 ‐22.1 41.0 Frequent vomiting (‐33.1, 21.7) (‐21.7, 91.9) (‐61.7, 28.6) (‐65.1, 95.0) (‐39.8, 226.9) a Analysis accounts for complex survey design b Percent change is predicted from fully adjusted regression models of each symptom on date of death; confidence intervals are bootstrapped

Appendix Table 6. Adjusted Average Yearly Percent Change in Prevalence of Symptoms between 1998 and 2010 (Fully Adjusted Models)a

Adjusted % Change (95% CIs)b CHF or Chronic Lung Entire Population Cancer Frailty Sudden Death Outcome Disease (n = 7,204) (n = 1,546) (n = 1,175) (n = 1,161) (n = 2,293) Moderate or severe pain 1.4 (‐0.4, 3.2) 1.2 (‐3.0, 5.5) 0.7 (‐2.2, 3.8) 1.4 (‐3.5, 6.5) ‐0.4 (‐4.7, 4.2) Any pain 2.4 (0.6, 4.1) 2.0 (‐2.2, 6.3) 2.5 (‐0.2, 5.3) 2.4 (‐1.5, 6.3) 1.6 (‐3.0, 6.4) Depression 3.8 (2.0, 5.7) 1.4 (‐2.1, 4.9) 4.3 (1.3, 7.4) 5.7 (1.8, 9.8) 2.1 (‐2.1, 6.5) Periodic confusion 5.4 (3.6, 7.2) 3.6 (‐0.2, 7.5) 4.7 (1.7, 7.8) 7.8 (2.6, 13.2) 5.8 (0.6, 11.4) Dyspnea 1.3 (‐0.1, 2.7) 0.3 (‐2.8, 3.5) 0.2 (‐3.1, 3.6) 1.1 (‐3.1, 5.5) 5.0 (1.6, 8.6) Incontinence 2.0 (‐0.2, 4.2) ‐0.8 (‐3.8, 2.4) 1.5 (‐1.5, 4.6) 1.1 (‐2.9, 5.1) 3.4 (‐2.5, 9.7) Severe fatigue 1.2 (‐0.7, 3.2) 3.0 (‐2.1, 8.3) ‐0.8 (‐4.0, 2.5) ‐0.1 (‐4.1, 4.1) 2.9 (‐1.6, 7.5) Anorexia 1.3 (‐0.8, 3.4) 3.0 (‐1.3, 7.5) 0.3 (‐3.0, 3.7) ‐1.5 (‐7.1, 4.4) 2.0 (‐2.8, 7.0) Frequent vomiting ‐1.1 (‐3.6, 1.4) 0.8 (‐3.7, 5.6) ‐3.9 (‐8.6, 1.1) ‐3.0 (‐12.0, 6.9) 5.4 (‐2.7, 14.2) a Analysis accounts for complex survey design b Adjusted percent change is predicted from regression models of each symptom on survey year, controlling for age at death, gender, race/ethnicity, number of nights spent in a hospital in the last two years of life, nursing home residency at the time of death and length of nursing home residency prior to death (regardless of site of death), number of ADL dependencies (0‐6 scale), diagnosis of arthritis (pain models only), depression reported in interview wave prior to death, highest level of education, household wealth, household income, proxy relationship to decedent, time elapsed between decedent’s death and proxy’s interview, average hours per week of care given to decedent from all informal carers, highest level of education if proxy was spouse, English/Spanish preference for interview, and year that the decedent entered the HRS sample

Chapter 3. Populations and Interventions for Palliative and End‐of‐Life Care: A Systematic Review

Abstract

Importance: There is increasing urgency to improve palliative and end‐of‐life care for patients and families living with advanced illness. Payers and providers need guidance on how best to deploy scarce resources.

Objective: To review evidence supporting (1) the identification of individuals most appropriate for palliative and end‐of‐life services and (2) specific elements of health services interventions most effective in achieving better outcomes for patients, caregivers, and the health care system.

Evidence Review: Systematic literature searches were performed of MEDLINE, EMBASE, PsycINFO, Web of Science, and Cochrane Database of Systematic Reviews databases between January 1, 2001, and

January 8, 2015. Studies included randomized controlled trials. 124 articles were selected for review.

Findings: In 124 relevant randomized controlled trials, the most well‐studied conditions were cancer, in which 49% of interventions (38 of 77 studies) demonstrated a majority of significant results, CHF (62%;

13 of 21), COPD (58%; 11 of 19), and dementia (60%; 15 of 25). Most studies that employed explicit prognostic criteria used clinicians’ judgment of “poor prognosis” (73%; 22 of 30). Most interventions included a nurse (70%; 69 of 98), and many of these were nurse‐only (39%; 27 of 69). Social workers were well‐represented in clinically diverse multidisciplinary teams. Home‐based approaches were common (56%; 70 of 124), and home interventions that included visits by intervention personnel were more effective than those without such visits (64%; 28 of 44; vs. 46%; 12 of 26). Interventions were most effective in improving communication and care planning, psychosocial health, and the patient and

caregiver experience. Economic outcomes were poorly characterized overall, but many interventions reduced hospital use (41%; 11 of 27). Palliative care teams did not lower health care costs.

Conclusions and Relevance: Palliative care services improve the support of patients with cancer, CHF,

COPD, and dementia. Effective models include nurses and social workers, home‐based components, and a focus on communication, psychosocial support, and the patient or caregiver experience. Better quality research is needed to understand the impact of palliative care on health care costs.

Introduction

There is increasing focus on palliative and end‐of‐life care because of the high cost and poor quality of care in late life and the growing emphasis on improving health care value.1 Previous systematic reviews have characterized a broad range of effective interventions for aspects of patients’ and caregivers’ HRQOL.2 Given substantial palliative care workforce constraints and often limited resources available to health care systems challenged by end‐of‐life needs, payers and providers are increasingly faced with the decisions of which patients to focus on and what services to prioritize.

The term “palliative care” signifies both an approach to dcare an specific health services, including consultation, hospice, and case management, among others. A recent IOM report defined palliative care as “care that provides relief from pain and other symptoms, supports quality of life, and is focused on patients with serious advanced illness and their families.”1 The report called for efforts to incorporate such principles into education, payment and delivery models, and public awareness campaigns.

To be successful, recommendations require specificity to inform how policymakers, payers, and providers should operationalize “advanced illness” and the specific interventions they should implement. We characterized the evidence base, focusing on RCTs published since 2001, to address these issues and to inform policy, payment, and practice. We focused on the following questions:

 What populations are appropriate for palliative care?

 What health services interventions and intervention elements improve aspects of

patient and/or caregiver quality of life, health care use, and health care costs?

Methods

We systematically reviewed published RCTs that addressed (1) advanced illness populations of patients and/or caregivers for whom palliative and end‐of‐life care should be considered; (2) effective palliative and end‐of‐life care interventions, including specific intervention components; and (3) the impact of palliative and end‐of‐life interventions on quality of life, health care use, and health care costs.

Data Sources and Searches

A research librarian searched MEDLINE, EMBASE, PsycINFO, Cochrane Database of Systematic

Reviews, Web of Sciences Databases (SCI‐EXPANDED, SSCI, A&HCI, CPCI‐S, and CPCI‐SSH), and

CareSearch Palliative Care Knowledge Network Review Collection (www.caresearch.com.au). We derived our primary literature search strategy from a search we had conducted for the NIH State of the

Science Meeting on End of Life Care in 20043 (Appendix Table 1). The original search encompassed terms for (1) specific diseases and debility associated with advanced illness and (2) domains and specific terms for HRQOL; we expanded it to include economic outcomes, which we defined as health care use, health care costs, and site of death.

We implemented this search in 2012 and updated it to March 8, 2013. We identified RCTs from our search effort in four ways: (1) directly from literature searches; (2) from reference lists in systematic reviews identified in searches with AMSTAR ratings of at least six;4 (3) from accepted studies in reviews we had previously conducted for an American College of Physicians guideline;5 (4) from systematic reviews that informed the development of the ACOVE quality measures and the Cancer Quality ASSIST quality indicators.6‐10

To update the review to January 2015, we reviewed title pages to identify all original research articles and systematic reviews published between March 2013 and January 8, 2015, in the New England

Journal of Medicine, Journal of the American Medical Association, Annals of Internal Medicine, British

Medical Journal, Journal of Clinical Oncology, Journal of Pain and Symptom Management, Journal of the 43

American Geriatrics Society, Journal of Palliative Medicine, Palliative Medicine, BioMed Central Palliative

Care, and British Medical Journal Supportive & Palliative Care. We identified RCTs published in these journals and from reference lists in systematic reviews with AMSTAR ratings of at least six published in these journals.4

We did not search gray literature, as we had not found it helpful in our previous review.2 Our evidence synthesis includes all relevant RCTs published between January 1, 2001, and January 8, 2015.

Study Selection

We applied the following inclusion criteria:

 Adults ≥ 18 years old living with or dying with advanced illness, and/or their caregivers

 Health services interventions addressing patient and/or caregiver quality of life‐related

elements in intervention design and/or as outcomes, with or without economic

outcomes

 The specific advanced illnesses of cancer, heart failure and other cardiac conditions,

chronic pulmonary disease, dementia and other neurological conditions, end‐stage liver

disease, or end‐stage renal disease, or any advanced illness populations receiving

palliative care, hospice, or end‐of‐life care

 Randomized controlled trials

 Published after January 1, 2001

We used Cochrane Collaboration definitions of study designs and attributes. We defined health services interventions using the World Health Organization’s definition of a health service.11 We defined quality of life and its related elements consistent with, but expanding upon, our previous work on this topic.3 Here, we considered it to include HRQOL; pain and non‐pain symptoms; functional status; existential or spiritual well‐being; communication with patients or families, including prognostication and care planning; continuity, defined as relationships with providers over time; experience or 44

satisfaction; caregiving, including nonprofessional activities that address emotional, spiritual, practical, or medical aspects of support; and bereavement care for caregivers or other loved ones.

We excluded (1) non‐English publications; (2) studies that were not conducted in the United

States, Western Europe, Israel, Canada, Australia, or New Zealand; (3) studies with only qualitative data;

(4) studies of economic outcomes only; (5) studies of drugs, devices, or technical care if they were not part of a health services intervention; and (6) studies of only support groups or psychological interventions unless they addressed an aspect of health services delivery (e.g., in‐person versus remote support).

Data Extraction and Quality Assessment

Our review team included physicians, nurses, and social workers; all individuals had experience with palliative care clinical delivery, research, or both. The physician reviewers had clinical expertise in palliative care, primary care, pulmonary and critical care, and ethics. The review was led by two senior health services investigators who have extensive literature review experience and who are also palliative care clinicians. Following definition of review goals, procedures, and inclusion and exclusion criteria, teams of paired reviewers conducted title, abstract, and full text reviews on all RCTs and systematic reviews.

Review forms reflected the study’s aims and conceptual framework (i.e., population categories, quality of life‐related elements, and intervention attributes). At all stages, forms were reviewed and piloted. Weekly team discussions resolved conflicts and clarified review procedures, with adjudication by senior team leaders. A single reviewer screened all accepted RCTs for bias using a Cochrane risk of bias tool modified for a previous systematic review of palliative care.12‐15 We used DistillerSR

(http://systematic‐review.ca/) to house review forms and conduct all reviews and data abstraction.

Data Synthesis and Analysis

Because of the heterogeneity of included studies and review outcomes, we determined that a meta‐analysis was inappropriate. Instead, we qualitatively synthesized evidence according to the patient and caregiver populations, intervention elements, and quality of life‐relevant and economic outcomes of included studies. We assessed consistency of evidence through the total numbers of studies that addressed each category of interest and the frequency of positive results within each category. We assessed directness of evidence by evaluating (1) the extent to and manner in which populations were characterized and (2) whether the literature shed light on quality of life‐relevant outcomes.16

Results

Literature Flow

In our original review, we identified 14,961 titles from primary searches and reference mining of prior systematic reviews. We identified 3,342 potentially relevant abstracts and 629 potentially relevant articles, and accepted 99 RCTs. In our update review, we identified 3,647 titles from the title pages of 11 major journals and reference mining of systematic reviews in these journals (Appendix Figure 1). We identified 71 potentially relevant abstracts and 68 potentially relevant articles, and accepted 25 RCTs.

We accepted a total of 124 RCTs in our original and update reviews combined.

Intervention Populations and Characterizing Advanced Illness

Conditions of Study Participants

Table 1 presents the patient populations in the included studies. Ninety‐three percent of all studies (115 of 124) described the conditions of their participants. Cancer was the most‐studied condition, followed by dementia, CHF, and COPD. Of the studies with conditions described, 83% (95 of

115) included only one condition. Cancer and dementia were overwhelmingly studied alone: 84% (65 of

77) and 88% (22 of 25), respectively, of studies that included these conditions did not include any other conditions. By contrast, 76% (16 of 21) and 84% (16 of 19) respectively, of studies that included CHF and

COPD included at least one other condition.

Looking at the majority of significant results among either patient or caregiver quality of life‐ relevant outcomes, we found evidence for effectiveness in cancer, CHF, COPD, and dementia: 62% (13 of

21) and 58% (11 of 19) of interventions including or focused on CHF and COPD, and 49% (38 of 77) and

60% (15 of 25) of interventions for cancer and dementia, had a majority of significant results.

Among cancer studies, interventions for metastatic cancer were more effective than those for non‐metastatic cancer, as were interventions for active treatment as compared with no treatment: 59% of interventions for metastatic cancer (20 of 34), compared with 42% of interventions for non‐ 47

metastatic cancer (18 of 43), demonstrated a majority of significant results; similarly, 58% of interventions for cancer being treated with chemotherapy and/or radiotherapy (15 of 26), compared with 45% of interventions for cancer not being treated with these modalities (23 of 51).

Severity of Illness

Studies assessed patients’ severity of illness in a variety of ways. Forty‐four percent of studies of cancer (34 of 77) included patients with metastatic cancer, and 34% (26 of 77) included patients undergoing chemotherapy and/or radiotherapy. Thirty‐eight percent of studies of CHF (8 of 21) assessed the NYHA class of their participants; of these, 63% (5 of 8) included at least 50% of participants with

NYHA Class III or IV CHF. Studies of other conditions did not report similar measures of disease severity.

Sixty‐two percent of all studies (77 of 124) reported both the number of patients enrolled and the number of decedents. Of these, the median percentage of decedents was 14%, with a range of 0% to 100%.

Sixty percent of all studies (74 of 124) reported the functional status of their patient participants. There were a variety of scales used: common scales included Eastern Cooperative Oncology

Group performance status, Karnofsky Performance Status, Palliative Performance Scale, SF‐12 and SF‐36 physical component summaries and physical functioning subscales, and activities of daily living counts.

The heterogeneity in measuring functional status precluded synthesis of or comparisons among the functional status of participants across studies.

Twenty‐four percent of all studies (30 of 124) used a health services‐derived definition to classify their participants’ prognosis. Of these, 73% (22 of 30) used clinicians’ judgment of “poor prognosis,” and of this latter set, 9% (2 of 22) asked the patient’s primary care or attending physician if they “would not be surprised if the patient died within one year.”17,18 Interventions that used clinicians’ judgment of “poor prognosis” were similarly effective as compared with those that did not. No studies used explicit criteria or administrative data‐derived approaches.

Intervention Elements

Intervention Personnel

Table 2 presents the intervention characteristics of included studies. Clinically trained individuals delivered 86% of all interventions (107 of 124); the remainder were usually delivered by specially trained study staff and/or volunteers. Of the interventions that described the clinical disciplines of their personnel, 48% (47 of 98) were delivered by only one type of clinically trained individual. Nurses, including advanced practice nurses, were the most common clinical discipline to deliver interventions:

70% of interventions (69 of 98) included a nurse, and 39% of these (27 of 69) were delivered only by nurses. Nurse‐only interventions or interventions in which nurses played the primary role often involved palliative case management,19‐26 education in symptom management and symptom monitoring,27‐41 and/or counseling and therapy,28,42,43 and were overwhelmingly delivered at home (85% of nurse‐only interventions [23 of 27] included in‐person nurse home visits and/or telephone calls from nurses to participants at home).

Interventions that included a nurse were equally effective as compared with those that did not:

52% of interventions that included a nurse (36 of 69), compared with 58% of interventions that did not include a nurse (32 of 55), demonstrated a majority of significant results. This pattern was similar for nurse‐only interventions: 52% of interventions that were nurse‐only (14 of 27), compared with 56% of interventions that were not nurse‐only (54 of 97), demonstrated a majority of significant results.

Interventions that included a social worker often involved palliative care teams,17,18,44‐52 palliative case management,44,45,50,53,54 and counseling and therapy,46,47,52,55,56 and 62% (13 of 21) were delivered at home.

Multidisciplinary Teams

Twenty‐six percent of all studies (32 of 124) employed multidisciplinary teams as part of their interventions. Sixty‐six percent of multidisciplinary teams (21 of 32) included three or more different

types of clinically trained individuals, 94% of them (30 of 32) included a nurse, and 50% of them (16 of

32) included a social worker. Social workers were part of clinically diverse teams: 100% of multidisciplinary teams that included a social worker (16 of 16) included three or more different types of clinically trained individuals. The palliative care team itself comprised 53% of all interventions involving palliative care teams (9 of 17);17,18,45,49,51,57‐60 the remainder were a heterogeneous set of interventions delivered alongside palliative consultation or in the context of palliative care.44,46‐48,50,52,61,62

Interventions without multidisciplinary teams often addressed more limited goals than team‐ based interventions (e.g., nurse‐directed pain management versus comprehensive, team‐based patient and family support).

Intervention Settings

Home was the most common intervention setting, followed by outpatient clinics and inpatient hospitals. In addition to the nurse‐led, home‐based interventions described above, interventions delivered at home also involved caregiver and family support and training.48,56,61,63‐72 Sixty‐six percent of home interventions (46 of 70) did not involve any other settings, 63% (44 of 70) involved home visits by intervention personnel, and 59% (41 of 70) involved the use of a telephone. On the whole, the nature of home‐based interventions did not differ between those with and without a telephone component.

Home interventions that included home visits by intervention personnel were more effective than home interventions without such visits: 64% of interventions that included home visits by interventions personnel (28 of 44), compared with 46% of home interventions without such visits (12 of 26), demonstrated a majority of significant results.

Sixty‐one percent of interventions delivered in a hospital (11 of 18) involved at least one other setting, and fifty percent of interventions with a hospital component (9 of 18) were delivered by palliative care teams.17,18,45,46,49,51,52,57,60 Interventions delivered in outpatient clinics were heterogeneous,

ranging from disease and symptom management education40,73‐80 to counseling and therapy.42,43,78,81,82

Outpatient clinics were often one of many settings in multi‐setting, multi‐component interventions.

Supporting Technology

Forty‐eight percent of all interventions (59 of 124) involved a technology component, most often a telephone to enable remote delivery of interventions and/or supportive check‐in calls. Other forms of technology were poorly represented: only three studies employed audio‐ or video‐based telehealth technology,23,65,83 two studies employed EHR‐based tools,84,85 and two studies employed mobile phones or tablets.40,86

Interventions that included a telephone component were more effective than those that did not: 61% of interventions that included a telephone component (31 of 51), compared with 51% of interventions that did not include a telephone component (37 of 73), demonstrated a majority of significant results.

Quality of Life‐Relevant Outcomes

Patient Outcomes

Table 3 presents the quality of life‐relevant outcomes addressed in included studies. Seventy‐ seven percent of all studies (96 of 124) addressed patient quality of life‐relevant domains. The range for the fraction of these studies that addressed individual domains was 11% (11 of 96) for existential or spiritual concerns to 42% (40 of 96) for symptoms other than pain, dyspnea, depressive symptoms, and anxiety. Approximately 20‐30% of studies that addressed pain (10 of 37),29,30,33,43,60,62,73,76,84,87 dyspnea (3 of 14),25,44,88 functional status (12 of 37),22,43,50,53,67,69,75,89‐93 and HRQOL (10 of 38)19,58,64,66,86,90‐92,94,95 had a majority of significant results in each of those domains; these frequencies were approximately 40‐50% for studies that addressed depressive symptoms (12 of 33),19,36,52,58,59,66,69,83,96‐99 anxiety (9 of

22),36,38,44,47,52,69,79,94,100 other symptoms (15 of 40),22,36,38,50,53,58,62,78‐80,91,94,101‐103 existential or spiritual concerns (4 of 11),44,55,88,103 patient experience or satisfaction (8 of 20),17,18,21,24‐26,54,104 and approximately

70% for communication or care planning (12 of 18).18,32,44,53,55,58,104‐109 Of all studies that addressed patient quality of life‐relevant outcomes, 45% (43 of 96) demonstrated a majority of significant results in the majority of quality of life‐relevant domains.

Caregiver Outcomes

Forty‐two percent of all studies (52 of 124) addressed caregiver quality of life‐relevant domains.

The range for the fraction of these studies that addressed individual domains was 6% (3 of 52) for existential or spiritual concerns to 40% (21 of 52) for depressive symptoms. No studies that addressed existential or spiritual concerns (0 of 3) had a majority of significant results in each of those domains; this frequency was approximately 30% for studies that addressed anxiety (2 of 6),110,111 40% for studies that addressed HRQOL (3 of 7),32,37,66 60‐70% for studies that addressed depressive symptoms (12 of

21),56,63,64,66,70,71,81,110‐114 other symptoms (5 of 8),32,56,110,114,115 and caregiver experience or satisfaction (5 of 8),26,55,57,104,116 and 75% for communication or care planning (3 of 4).32,72,117 Of all studies that addressed caregiver quality of life‐relevant outcomes, 60% (31 of 52) demonstrated a majority of significant results in the majority of quality of life‐relevant domains.

Relationships among Key Study Characteristics and Select Quality of Life‐Relevant Outcomes

Sixty‐one percent of interventions that addressed patient symptoms (59 of 96) were delivered only to cancer patients. Patient pain and depressive symptoms in particular were a common focus in cancer studies but less commonly addressed in other conditions.21,25,44,66,83,88,100,102,118,119 Interventions that addressed pain and depressive symptoms in cancer patients were much more effective than those that addressed pain and depressive symptoms in other conditions: no interventions demonstrated significant results for pain, and only two for depressive symptoms,66,83 in CHF, COPD, or dementia.

Interventions that had a significant effect on patient pain (10 studies)29,30,33,43,60,62,73,76,84,87 were usually delivered to cancer patients (80% [8 of10] included only cancer patients)29,30,33,43,73,76,84,87 and specifically designed to address cancer pain, usually involved pain education and management, and

were short‐term (40% [4 of 10] were one‐time interventions,33,62,73,87 and none exceeded six weeks).

Nurse‐only interventions were particularly effective at reducing pain: 50% of all interventions that reduced pain (5 of 10) employed only nurses.29,30,33,43,84

In contrast to pain, interventions that had a significant effect on patient depressive symptoms

(12 studies)19,36,52,58,59,66,69,83,96‐99 were quite heterogeneous in terms of settings, use of multidisciplinary teams, clinical disciplines of intervention personnel, duration, and focus: case management, palliative care teams, hospice, skills training, and other interventions were all effective. Interventions that had a significant effect on caregiver depressive symptoms (12 studies),56,63,64,66,70,71,81,110‐114 by contrast, often involved a home component (83% [10 of 12]),56,63,64,66,70,71,111‐114 did not use multidisciplinary teams, were long‐term (average duration 1.2 years), and often involved caregiver skills training63,64,71,111 or counseling and therapy.56,66,70,112‐114 The use of mental health professionals also differed between effective interventions in patient versus caregiver depressive symptoms: 86% of interventions (6 of 7)64,70,81,112‐114 that employed mental health professionals and addressed caregiver depressive symptoms were effective, as compared with 0% of interventions (0 of 7) that employed mental health professionals and addressed patient depressive symptoms.

Interventions that had a significant effect on patient HRQOL (10 studies)19,58,64,66,86,90‐92,94,95 were generally more comprehensive in focus. They ranged from case management19,90,91 to caregiver training,64 occupational therapy,66 quality of life questionnaire completion,86,95 and a hospital‐to‐home transitional care program.90 Forty percent (4 of 10)58,91,92,94 involved a multidisciplinary team, and all were delivered in outpatient settings and/or at home.

Interventions that had a significant effect on patient and/or caregiver communication or care planning outcomes (14 studies)32,44,53,55,72,104‐109,117 often included decision support,53,104,106,107,117 advance directive completion,44,53,104 education and/or communication training,32,55,72,108 and case management components.44,53,104 Seventy‐one percent of interventions that included a nurse (5 of 7)18,32,44,53,58 and

addressed patient communication or care planning outcomes had a significant effect on these outcomes; 100% of interventions that included a social worker (4 of 4)18,44,53,55 and addressed patient communication or care planning outcomes had a significant effect on these outcomes. Only 7% of interventions (9 of 124) involved decision support, but 56% of these (5 of 9)53,104,106,107,117 had a significant effect on communication or care planning outcomes. This pattern was similar for advance directive completion: only 5% of interventions (6 of 124) involved advance directive completion, but

50% of these (3 of 6)44,53,104 had a significant effect on communication or care planning outcomes.

Interventions with a hospital component focused on patients with a greater burden of illness than those without a hospital component: many were in advanced stages of cancer, CHF, or COPD, and

44% of these interventions (8 of 18)17,18,45,46,52,56,57,110 were delivered to patients determined by clinicians to have a poor prognosis (in one case within days of death).110 These interventions were generally less effective than those without a hospital component.

Longer interventions tended to be more effective for patient depressive symptoms and anxiety

(56% of interventions [11 of 18]19,36,44,47,58,59,69,79,83,96,98 with significant effects on these outcomes lasted at least 12 weeks) and patient and caregiver quality of life (75% of interventions [9 of 12]19,32,58,64,86,90‐92,95 with significant effects on these outcomes lasted at least 12 weeks).

Economic Outcomes

Health Care Use

Forty‐one percent of studies (51 of 124) addressed economic outcomes (Table 4). Health care use was the most frequently studied, followed by costs and drugs. Hospital use was the most well‐ studied utilization outcome: 65% of studies with a significant effect on health care use (11 of

17)17,18,24,25,50,55,75,89,91,102,116 had effects on CHF‐related and all‐cause hospital readmissions and length of stay. Emergency department use was also relatively well‐studied, but only 31% of interventions (4 of 13) had a significant effect on it.17,23,83,90 Fifty‐three percent of interventions with a significant effect on

health care use (9 of 17)23‐25,44,50,75,89‐91 included a case management component, and many included education50,55,75 and disease and symptom management41,83,102 components. Only four studies addressed hospice use.17,18,25,116

Health Care Costs

Of the studies that addressed health care costs, 74% (14 of 19)18,24‐26,54,84,89,91‐93,104,115,120,121 reported intervention costs; however, only 50% of these (7 of 14)26,89,92,93,115,120,121 incorporated intervention costs into statistical tests for intervention efficacy in lowering costs or in cost‐effectiveness analyses. Thirty‐two percent of studies that addressed health care costs (6 of 19)17,18,23,89,92,96 demonstrated significant effects on cost outcomes or cost‐effectiveness. Interventions tended to be more successful in lowering specific costs (36% of interventions that addressed specific costs [4 of

11]18,23,89,96 significantly reduced these costs) than in lowering overall costs of care (20% of interventions that addressed overall costs [3 of 15]17,18,92 significantly reduced these costs). Specific costs reduced included hospital readmission,18,89 ED,23 and pharmacologic89,96 costs.

No single type of intervention was especially effective at lowering health care costs, but most effective interventions included participants from a variety of clinical disciplines (83% [5 of 6])17,18,89,92,96 included at least three clinical disciplines). Nurses and social workers were well‐represented among interventions with significant effects on health care costs: all of these interventions included a nurse, and 67% of them (4 of 6)17,18,92,96 included a social worker. Palliative care teams had little effect on costs: only 22% of interventions in which the intervention itself was the palliative care team (2 of 9)17,18 were effective at lowering health care costs.

Discussion

We identified a large body of RCTs of palliative and end‐of‐life health services interventions for advanced illness. In terms of patient populations, we found the strongest evidence for interventions in cancer, CHF, and COPD. In terms of intervention components, we found the strongest evidence for the roles of nurses and social workers, strong evidence for home‐based components, and moderate evidence for multidisciplinary approaches. In terms of quality of life‐relevant outcomes for both patients and caregivers, we found the strongest evidence for the effectiveness of palliative and end‐of‐life services in improving communication and care planning, moderate evidence for improving psychosocial health and the patient and caregiver experience, and weaker evidence for improving HRQOL, pain, dyspnea, functional status, and existential or spiritual concerns. In terms of economic outcomes, we found moderate evidence for the effectiveness of palliative and end‐of‐life services in reducing hospital use, weaker evidence for their effectiveness in reducing other specific health care use, and moderate evidence for their effectiveness in lowering health care costs. We found weak evidence for palliative care teams specifically in reducing health care costs.

The strategy of identifying patients for palliative care by eliciting clinician‐reported risk was a common and generally effective approach. It is prudent, relatively simple to implement, and may foster buy‐in for implementing palliative services.122 The literature was silent on innovative approaches that contemporary entrepreneurs, payers, and providers are taking to identify high‐risk patients who might be targeted for palliative care. These include data mining that combines health and consumer information and novel sensor technologies that characterize function and high‐risk events (e.g., falls).

Palliative and end‐of‐life services were best‐studied in cancer populations. In general, studies of cancer focused on very sick patients, identified on the basis of severity (e.g., metastatic cancer) or clinician‐identified poor prognosis. Our review found evidence for the effectiveness of palliative and end‐of‐life health services interventions in cancer, supporting a recent American Society of Clinical 56

Oncology guideline’s and other calls for integrating palliative care more routinely in advanced cancer.123

While cancer was the best‐studied condition in the literature, we also found evidence supporting the application of palliative care to other advanced illnesses (CHF, COPD, and dementia) and to mixed populations.

The literature supports various models for delivering palliative care. We found evidence for the role of nurses and social workers both in multidisciplinary teams and working alone (e.g., case management). Our findings underscore the value of delivering services at home, whether in‐person or by telephone, and the role of nurses and social workers in home‐based care. Integrated payment and delivery may herald improved support for nurses and social workers,124 who play crucial roles in communication and coordination; for example, two thirds of Physician Orders for Life Sustaining

Treatment in Oregon are completed in nursing homes by social workers.125

Our review generally supported the ability of palliative services to promote effective communication and deliver symptomatic support. The evidence is strongest for care planning and psychological outcomes, which is consistent with the results of clinical systematic reviews.2 The recent

IOM report on end‐of‐life care called for a special emphasis on communication and advance care planning, which is appropriate given the evidence supporting this practice area.1 We found that decision support and advance directive completion should be considered as a facilitative component of this emphasis.

We found minimal health care use and cost reductions for palliative care in general (aside from hospital use) and palliative care teams specifically. Other reviews have demonstrated an advantage for the tlatter bu have included lower‐quality study designs,126 and a recent high‐quality RCT of a palliative care team that demonstrated clinical benefit and lower health care use failed to demonstrate a cost advantage.127 Our results are reassuring in the sense that they undermine the cynical association of end‐ of‐life care with “death panels.” However, our results most likely reflect a lack of evidence rather than a

true lack of effect, as the included studies had significant shortcomings in characterizing economic outcomes: their measurement of these outcomes was heterogeneous and they often failed to account for various dimensions of cost or use. Our results call for higher quality research on costs and other economic outcomes, and the lack of cost savings for palliative care teams suggests the need to study single provider models that address specific outcomes such as communication or pain management.

We faced several limitations. First, we focused on health services interventions only, so our review does not encompass the full range of effective interventions that palliative care teams and providers are capable of delivering. Second, our definition of what constitutes “palliative care” was not limited to a specialty service and could have increased the heterogeneity of included articles, both because of the subjectivity of the definition and inconsistent indexing in the literature. However, our approach improves the generalizability and applicability of our findings.

In summary, our review underscores the importance of the recent IOM report and its call to more broadly implement palliative and end‐of‐life care. It supports an emphasis on cancer, chronic cardiopulmonary conditions, and dementia; highlights the importance of nurses and social workers, multidisciplinary approaches, and interventions that encompass the home; and reinforces the efficacy of palliative and end‐of‐life services for improving communication and psychosocial support in advanced illness. It also highlights the need to understand better the impact of palliative care on health care costs.

References

1. Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. Washington, DC: Institute of Medicine;2014. 2. Lorenz KA, Lynn J, Dy SM, et al. Evidence for improving palliative care at the end of life: a systematic review. Ann Intern Med. 2008;148(2):147‐159. 3. Lorenz K, Lynn J, Morton SC, et al. End‐of‐Life Care and Outcomes Summary: Evidence Report/Technology Assessment: Number 110. Rockville, MD: Agency for Healthcare Research and Quality;2004. 4. Shea BJ, Bouter LM, Peterson J, et al. External validation of a measurement tool to assess systematic reviews (AMSTAR). PloS One. 2007;2(12):e1350. 5. Qaseem A, Snow V, Shekelle P, et al. Evidence‐based interventions to improve the palliative care of pain, dyspnea, and depression at the end of life: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2008;148(2):141‐146. 6. Lorenz KA, Rosenfeld K, Wenger N. Quality indicators for palliative and end‐of‐life care in vulnerable elders. J Am Geriatr Soc. 2007;55 Suppl 2:S318‐326. 7. Naeim A, Dy SM, Lorenz KA, Sanati H, Walling A, Asch SM. Evidence‐based recommendations for cancer nausea and vomiting. J Clin Oncol. 2008;26(23):3903‐3910. 8. Walling A, Lorenz KA, Dy SM, et al. Evidence‐based recommendations for information and care planning in cancer care. J Clin Oncol. 2008;26(23):3896‐3902. 9. Dy SM, Lorenz KA, Naeim A, Sanati H, Walling A, Asch SM. Evidence‐based recommendations for cancer fatigue, anorexia, depression, and dyspnea. J Clin Oncol. 2008;26(23):3886‐3895. 10. Dy SM, Asch SM, Naeim A, Sanati H, Walling A, Lorenz KA. Evidence‐based standards for cancer pain management. J Clin Oncol. 2008;26(23):3879‐3885. 11. Rubin DB. Multiple imputation after 18+ years. J Am Stat Assoc. 1996;91(434):473‐489. 12. Dy SM, Aslakson R, Wilson RF, et al. Improving Health Care and Palliative Care for Advanced and Serious Illness. Rockville, MD: Agency for Healthcare Research and Quality;2012. 13. Armijo‐Olivo S, Ospina M, da Costa BR, et al. Poor reliability between Cochrane reviewers and blinded external reviewers when applying the Cochrane risk of bias tool in physical therapy trials. PloS One. 2014;9(5):e96920. 14. Graham N, Haines T, Goldsmith CH, et al. Reliability of 3 assessment tools used to evaluate randomized controlled trials for treatment of neck pain. Spine. 2012;37(6):515‐522. 15. Hartling L, Hamm MP, Milne A, et al. Testing the risk of bias tool showed low reliability between individual reviewers and across consensus assessments of reviewer pairs. J Clin Epidemiol. 2013;66(9):973‐981. 16. NHPCO Facts and Figures: Hospice Care in America. Alexandria, VA: National Hospice and Palliative Care Organization;2014. 17. Brumley R, Enguidanos S, Jamison P, et al. Increased satisfaction with care and lower costs: results of a randomized trial of in‐home palliative care. J Am Geriatr Soc. 2007;55(7):993‐1000. 18. Gade G, Venohr I, Conner D, et al. Impact of an inpatient palliative care team: a randomized control trial. J Palliat Med. 2008;11(2):180‐190. 19. Bakitas M, Lyons KD, Hegel MT, et al. Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer: the Project ENABLE II randomized controlled trial. JAMA. 2009;302(7):741‐749. 20. O'Hara RE, Hull JG, Lyons KD, et al. Impact on caregiver burden of a patient‐focused palliative care intervention for patients with advanced cancer. Palliat Support Care. 2010;8(4):395‐404.

21. Egan E, Clavarino A, Burridge L, Teuwen M, White E. A randomized control trial of nursing‐based case management for patients with chronic obstructive pulmonary disease. Lippincotts Case Manag. 2002;7(5):170‐179. 22. Given B, Given CW, McCorkle R, et al. Pain and fatigue management: results of a nursing randomized clinical trial. Oncol Nurs Forum. 2002;29(6):949‐956. 23. Jerant AF, Azari R, Nesbitt TS. Reducing the cost of frequent hospital admissions for congestive heart failure: a randomized trial of a home telecare intervention. Med Care. 2001;39(11):1234‐ 1245. 24. Riegel B, Carlson B, Kopp Z, LePetri B, Glaser D, Unger A. Effect of a standardized nurse case‐ management telephone intervention on resource use in patients with chronic heart failure. Arch Intern Med. 2002;162(6):705‐712. 25. Moore S, Corner J, Haviland J, et al. Nurse led follow up and conventional medical follow up in management of patients with lung cancer: randomised trial. BMJ. 2002;325(7373):1145. 26. Uitdehaag MJ, van Putten PG, van Eijck CH, et al. Nurse‐led follow‐up at home vs. conventional medical outpatient clinic follow‐up in patients with incurable upper gastrointestinal cancer: a randomized study. J Pain Symptom Manage. 2014;47(3):518‐530. 27. Keefe FJ, Ahles TA, Sutton L, et al. Partner‐ guided cancer pain management at the end of life: a preliminary study. J Pain Symptom Manage. 2005;29(3):263‐272. 28. Wright LK, Litaker M, Laraia MT, DeAndrade S. Continuum of care for Alzheimer's disease: a nurse education and counseling program. Issues Ment Health Nurs. 2001;22(3):231‐252. 29. Miaskowski C, Dodd M, West eC, et al. Th use of a responder analysis to identify differences in patient outcomes following a self‐care intervention to improve cancer pain management. Pain. 2007;129(1‐2):55‐63. 30. Miaskowski C, Dodd M, West C, et al. Randomized clinical trial of the effectiveness of a self‐care intervention to improve cancer pain management. J Clin Oncol. 2004;22(9):1713‐1720. 31. Northouse L, Kershaw T, Mood D, Schafenacker A. Effects of a family intervention on the quality of life of women with recurrent breast cancer and their family caregivers. Psychooncology. 2005;14(6):478‐491. 32. Northouse LL, Mood DW, Schafenacker A, et al. Randomized clinical trial of a family intervention for prostate cancer patients and their spouses. Cancer. 2007;110(12):2809‐2818. 33. Ward S, Donovan H, Gunnarsdottir S, Serlin RC, Shapiro GR, Hughes S. A randomized trial of a representational intervention to decrease cancer pain (RIDcancerPain). Health Psychol. 2008;27(1):59‐67. 34. Wells N, Hepworth JT, Murphy BA, Wujcik D, Johnson R. Improving Cancer Pain Management Through Patient and Family Education. J Pain Symptom Manage. 2003;25(4):344‐356. 35. Anderson KO, Mendoza TR, Payne R, et al. Pain education for underserved minority cancer patients: a randomized controlled trial. J Clin Oncol. 2004;22(24):4918‐4925. 36. Ream E, Richardson A, Alexander‐Dann C. Supportive intervention for fatigue in patients undergoing chemotherapy: a randomized controlled trial. J Pain Symptom Manage. 2006;31(2):148‐161. 37. McMillan SC, Small BJ, Weitzner M, et al. Impact of coping skills intervention with family caregivers of hospice patients with cancer: a randomized clinical trial. Cancer. 2006;106(1):214‐ 222. 38. Bruera E, Yennurajalingam S, Palmer JL, et al. Methylphenidate and/or a nursing telephone intervention for fatigue in patients with advanced cancer: a randomized, placebo‐controlled, phase II trial. J Clin Oncol. 2013;31(19):2421‐2427.

39. Donovan HS, Ward SE, Sereika SM, et al. Web‐based symptom management for women with recurrent ovarian cancer: a pilot randomized controlled trial of the WRITE Symptoms intervention. J Pain Symptom Manage. 2014;47(2):218‐230. 40. Nguyen HQ, Donesky D, Reinke LF, et al. Internet‐based dyspnea self‐management support for patients with chronic obstructive pulmonary disease. J Pain Symptom Manage. 2013;46(1):43‐ 55. 41. Yount SE, Rothrock N, Bass M, et al. A randomized trial of weekly symptom telemonitoring in advanced lung cancer. J Pain Symptom Manage. 2014;47(6):973‐989. 42. Given B, Given CW, Sikorskii A, Jeon S, Sherwood P, Rahbar eM. Th impact of providing symptom management assistance on caregiver reaction: results of a randomized trial. J Pain Symptom Manage. 2006;32(5):433‐443. 43. Dalton JA, Keefe FJ, Carlson J, Youngblood R. Tailoring cognitive‐behavioral treatment for cancer pain. Pain Manag Nurs. 2004;5(1):3‐18. 44. Rabow MW, Dibble SL, Pantilat SZ, McPhee. SJ The comprehensive care team: a controlled trial of outpatient palliative medicine consultation. Arch Intern Med. 2004;164(1):83‐91. 45. Jordhoy MS, Fayers P, Loge JH, Ahlner‐Elmqvist M, Kaasa S. Quality of life in palliative cancer care: results from a cluster randomized trial. J Clin Oncol. 2001;19(18):3884‐3894. 46. Chochinov HM, Kristjanson LJ, Breitbart W, et al. Effect of dignity therapy on distress and end‐of‐ life experience in terminally ill patients: a randomised controlled trial. Lancet Oncol. 2011;12(8):753‐762. 47. Moorey S, Cort E, Kapari M, et al. A cluster randomized controlled trial of cognitive behaviour therapy for common mental disorders in patients with advanced cancer. Psychol Med. 2009;39(5):713‐723. 48. Walsh K, Jones L, Tookman A, et al. Reducing emotional distress in people caring for patients receiving specialist palliative care. Randomised trial. Br J Psychiatry. 2007;190:142‐147. 49. Hanks GW, Robbins M, Sharp D, et al. The imPaCT study: a randomised controlled trial to evaluate a hospital palliative care team. Br J Cancer. 2002;87(7):733‐739. 50. Abernethy AP, Currow DC, Shelby‐James T, et al. Delivery strategies to optimize resource utilization and performance status for patients with advanced life‐limiting illness: results from the "palliative care trial" [ISRCTN 81117481]. J Pain Symptom Manage. 2013;45(3):488‐505. 51. Wallen GR, Baker K, Stolar M, et al. Palliative care outcomes in surgical oncology patients with advanced malignancies: a mixed methods approach. Qual Life Res. 2012;21(3):405‐415. 52. Juliao M, Oliveira F, Nunes B, Vaz Carneiro A, Barbosa A. Efficacy of dignity therapy on depression and anxiety in Portuguese terminally ill patients: a phase II randomized controlled trial. J Palliat Med. 2014;17(6):688‐695. 53. Aiken LS, Butner J, Lockhart CA, Volk‐Craft BE, Hamilton G, Williams FG. Outcome evaluation of a randomized trial of the PhoenixCare intervention: program of case management and coordinated care for the seriously chronically ill. J Palliat Med. 2006;9(1):111‐126. 54. Laramee AS, Levinsky SK, Sargent J, Ross R, Callas P. Case management in a heterogeneous congestive heart failure population: a randomized controlled trial. Arch Intern Med. 2003;163(7):809‐817. 55. Engelhardt JB, Rizzo VM, Della Penna RD, et al. Effectiveness of care coordination and health counseling in advancing illness. Am J Manag Care. 2009;15(11):817‐825. 56. Kissane DW, McKenzie M, Bloch S, Moskowitz C, McKenzie DP, O'Neill I. Family focused grief therapy: a randomized, controlled trial in palliative care and bereavement. Am J Psychiatry. 2006;163(7):1208‐1218.

57. Ringdal GI, Jordhoy MS, Kaasa S. Family satisfaction with end‐of‐life care for cancer patients in a cluster randomized trial. J Pain Symptom Manage. 2002;24(1):53‐63. 58. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non‐ small‐cell lung cancer. N Engl. J Med 2010;363(8):733‐742. 59. Pirl WF, Greer JA, Traeger L, et al. Depression and survival in metastatic non‐small‐cell lung cancer: effects of early palliative care. J Clin Oncol. 2012;30(12):1310‐1315. 60. Higginson IJ, McCrone P, Hart SR, Burman R, Silber E, Edmonds PM. Is short‐term palliative care cost‐effective in multiple sclerosis? A randomized phase II trial. J Pain Symptom Manage. 2009;38(6):816‐826. 61. Hudson PL, Aranda S, Hayman‐White K. A psycho‐educational intervention for family caregivers of patients receiving palliative care: a randomized controlled trial. J Pain Symptom Manage. 2005;30(4):329‐341. 62. Mitchell GK, Del M, C. B., O'Rourke PK, Clavarino AM. Do case conferences between general practitioners and specialist palliative care services improve quality of life? A randomised controlled trial (ISRCTN 52269003). Palliat Med. 2008;22(8):904‐912. 63. Bourgeois MS, Shulz R, Burgio LD, Beach S. Skills training for spouses of patients with Alzheimer's disease: outcomes of an intervention study. J Clin Geropsychol. 2002;8(1):53‐73. 64. Teri L, McCurry SM, Logsdon R, Gibbons LE. Training community consultants to help family members improve dementia care: a randomized controlled trial. Gerontologist. 2005;45(6):802‐ 811. 65. Gant JR, Steffen AM, Lauderdale SA. Comparative outcomes of two distance‐based interventions for male caregivers of family members with dementia. Am J Alzheimers Dis Other Demen. 2007;22(2):120‐128. 66. Graff MJ, Vernooij‐Dassen MJ, Thijssen M, Dekker J, Hoefnagels WH, Olderikkert MG. Effects of community occupational therapy on quality of life, mood, and health status in dementia patients and their caregivers: a randomized controlled trial. J Gerontol A Biol Sci Med Sci. 2007;62(9):1002‐1009. 67. Graff MJ, Vernooij‐Dassen MJ, Thijssen M, Dekker J, Hoefnagels WH, Rikkert MG. Community based occupational therapy for patients with dementia and their care givers: randomised controlled trial. BMJ. 2006;333(7580):1196. 68. Nobili A, Riva E, Tettamanti M, et al.e Th effect of a structured intervention on caregivers of patients with dementia and problem behaviors: a randomized controlled pilot study. Alzheimer Dis Assoc Disord. 2004;18(2):75‐82. 69. Porter LS, Keefe FJ, Garst J, et al. Caregiver‐assisted coping skills training for lung cancer: results of a randomized clinical trial. J Pain Symptom Manage. 2011;41(1):1‐13. 70. Eisdorfer C, Czaja SJ, Loewenstein DA, et al. The effect of a family therapy and technology‐based intervention on caregiver depression. Gerontologist. 2003;43(4):521‐531. 71. Belle SH, Burgio L, Burns R, et al. Enhancing the quality of life of dementia caregivers from different ethnic or racial groups: a randomized, controlled trial. Ann Intern Med. 2006;145(10):727‐738. 72. Done DJ, Thomas JA. Training in communication skills for informal carers of people suffering from dementia: a cluster randomized clinical trial comparing a therapist led workshop and a booklet. Int J Geriatr Psychiatry. 2001;16(8):816‐821. 73. Kalauokalani D, Franks P, Oliver JW, Meyers FJ, Kravitz RL. Can patient coaching reduce racial/ethnic disparities in cancer pain control? Secondary analysis of a randomized controlled trial. Pain Med. 2007;8(1):17‐24.

74. Oliver JW, Kravitz RL, Kaplan SH, Meyers FJ. Individualized patient education and coaching to improve pain control among cancer outpatients. J Clin Oncol. 2001;19(8):2206‐2212. 75. Doughty RN, Wright SP, Pearl A, et al. Randomized, controlled trial of integrated heart failure management: The Auckland Heart Failure Management Study. Eur Heart J. 2002;23(2):149‐146. 76. Vallieres I, Aubin M, Blondeau L, Simard S, Giguere A. Effectiveness of a clinical intervention in improving pain control in outpatients with cancer treated by radiation therapy. Int J Radiat Oncol Biol Phys. 2006;66(1):234‐237. 77. Ward SE, Serlin RC, Donovan HS, et al. A randomized trial of a representational intervention for cancer pain: does targeting the dyad make a difference? Health Psychol. 2009;28(5):588‐597. 78. Boesen EH, Ross L, Frederiksen K, et al. Psychoeducational intervention for patients with cutaneous malignant melanoma: a replication study. J Clin Oncol. 2005;23(6):1270‐1277. 79. de Raaf PJ, de Klerk C, Timman R, Busschbach JJ, Oldenmenger WH, van der Rijt CC. Systematic monitoring and treatment of physical symptoms to alleviate fatigue in patients with advanced cancer: a randomized controlled trial. J Clin Oncol. 2013;31(6):716‐723. 80. Berry DL, Hong F, Halpenny B, et al. Electronic self‐report assessment for cancer and eself‐car support: results of a multicenter randomized trial. J Clin Oncol. 2014;32(3):199‐205. 81. Haley WE, Bergman EJ, Roth DL, McVie T, Gaugler JE, Mittelman MS. Long‐term effects of bereavement and caregiver intervention on dementia caregiver depressive symptoms. Gerontologist. 2008;48(6):732‐740. 82. McLean LM, Walton T, Rodin G, Esplen MJ, Jones JM. A couple‐based intervention for patients and caregivers facing end‐stage cancer: outcomes of a randomized controlled trial. Psychooncology. 2013;22(1):28‐38. 83. Gellis ZD, Kenaley BL, Ten Have T. Integrated telehealth care for chronic illness and depression in geriatric home care patients: the Integrated Telehealth Education and Activation of Mood (I‐ TEAM) study. J Am Geriatr Soc. 2014;62(5):889‐895. 84. McDonald MV, Pezzin LE, Feldman PH, Murtaugh CM, Peng TR. Can just‐in‐time, evidence‐based "reminders" improve pain management among home health care nurses and their patients? J Pain Symptom Manage. 2005;29(5):474‐488. 85. Tierney WM, Dexter, PR Gramelspacher GP, Perkins AJ, Zhou XH, Wolinsky FD. The effect of discussions about advance directives on patients' satisfaction with primary care. J Gen Intern Med. 2001;16(1):32‐40. 86. Velikova G, Booth L, Smith AB, et al. Measuring quality of life in routine oncology practice improves communication and patient well‐being: a randomized controlled trial. J Clin Oncol. 2004;22(4):714‐724. 87. Syrjala KL, Abrams JR, Polissar NL, et al. Patient training in cancer pain management using integrated print and video materials: a multisite randomized controlled trial. Pain. 2008;135(1‐ 2):175‐186. 88. Allen RS, Hilgeman MM, Ege MA, Shuster JL, Jr., Burgio LD. Legacy activities as interventions approaching the end of life. J Palliat Med. 2008;11(7):1029‐1038. 89. Capomolla S, Febo O, Ceresa M, et al. Cost/utility ratio in chronic heart failure: comparison between heart failure management program delivered by day‐hospital and usual care. J Am Coll Cardiol. 2002;40(7):1259‐1266. 90. Harrison MB, Browne GB, Roberts J, Tugwell P, Gafni A, Graham ID. Quality of life of individuals with heart failure: a randomized trial of the effectiveness of two models of hospital‐to‐home transition. Med Care. 2002;40(4):271‐282.

91. Kasper EK, Gerstenblith G, Hefter G, et al. A randomized trial of the efficacy of multidisciplinary care in heart failure outpatients at high risk of hospital readmission. J Am Coll Cardiol. 2002;39(3):471‐480. 92. Jones L, Fitzgerald G, Leurent B, et al. Rehabilitation in advanced, progressive, recurrent cancer: a randomizedd controlle trial. J Pain Symptom Manage. 2013;46(3):315‐325.e313. 93. Hollingworth W, Metcalfe C, Mancero S, et al. Are needs assessments cost effective in reducing distress among patients with cancer? A randomized controlled trial using the Distress Thermometer and Problem List. J Clin Oncol. 2013;31(29):3631‐3638. 94. Rummans TA, Clark MM, Sloan JA, et al. Impacting quality of life for patients with advanced cancer with a structured multidisciplinary intervention: a randomized controlled trial. J Clin Oncol. 2006;24(4):635‐642. 95. Mills ME, Murray LJ, Johnston BT, Cardwell C, Donnelly M. Does a patient‐held quality‐of‐life diary benefit patients with inoperable lung cancer? J Clin Oncol. 2009;27(1):70‐77. 96. Markle‐Reid M, Weir R, Browne G, Roberts J, Gafni A, Henderson S. Health promotion for frail older home care clients. J Adv Nurs. 2006;54(3):381‐395. 97. McMillan SC, Small BJ, Haley WE. Improving hospice outcomes through systematic assessment: a clinical trial. Cancer Nurs. 2011;34(2):89‐97. 98. McLachlan SA, Allenby A, Matthews J, et al. Randomized trial of coordinated psychosocial interventions based on patient self‐assessments versus standard care to improve the psychosocial functioning of patients with cancer. J Clin Oncol. 2001;19(21):4117‐4125. 99. Stanton AL, Thompson EH, Crespi CM, Link JS, Waisman JR. Project connect online: randomized trial of an internet‐based program to chronicle the cancer experience and facilitate communication. J Clin Oncol. 2013;31(27):3411‐3417. 100. Chapman DG, Toseland RW. Effectiveness of advanced illness care teams for nursing home residents with dementia. Soc Work. 2007;52(4):321‐329. 101. McMillan SC, Small BJ. Using the COPE intervention for family caregivers to improve symptoms of hospice homecare patients: a clinical trial. Oncol Nurs Forum. 2007;34(2):313‐321. 102. Rea H, McAuley S, Stewart A, Lamont C, Roseman P, Didsbury P. A chronic disease management programme can reduce days in hospital for patients with chronic obstructive pulmonary disease. Intern Med J. 2004;34(11):608‐614. 103. Allen RS, Harris GM, Burgio LD, et al. Can senior volunteers deliver reminiscence and creative activity interventions? Results of the legacy intervention family enactment randomized controlled trial. J Pain Symptom Manage. 2014;48(4):590‐601. 104. Engelhardt JB, McClive‐Reed KP, Toseland RW, Smith TL, Larson DG, Tobin DR. Effects of a program for coordinated care of advanced illness on patients, surrogates, and healthcare costs: a randomized trial. Am J Manag Care. 2006;12(2):93‐100. 105. Clayton JM, Butow PN, Tattersall MH, et al. Randomized controlled trial of a prompt list to help advanced cancer patients and their caregivers to ask questions about prognosis and end‐of‐life care. J Clin Oncol. 2007;25(6):715‐723. 106. El‐Jawahri A, Podgurski LM, Eichler AF, et al. Use of video to facilitate end‐of‐life discussions with patients with cancer: a randomized controlled trial. J Clin Oncol. 2010;28(2):305‐310. 107. Volandes AE, Ferguson LA, Davis AD, et al. Assessing end‐of‐life preferences for advanced dementia in rural patients using an educational video: a randomized controlled trial. J Palliat Med. 2011;14(2):169‐177. 108. Au DH, Udris EM, Engelberg RA, et al. A randomized trial to improve communication about end‐ of‐life care among patients with COPD. Chest. 2012;141(3):726‐735.

109. Stein RA, Sharpe L, Bell ML, Boyle FM, Dunn SM, Clarke SJ. Randomized controlled trial of a structured intervention to facilitate end‐of‐life decision making in patients with advanced cancer. J Clin Oncol. 2013;31(27):3403‐3410. 110. Lautrette A, Darmon M, Megarbane B, et al. A communication strategy and brochure for relatives of patients dying in the ICU. N Engl J Med. 2007;356(5):469‐478. 111. Beauchamp N, Irvine AB, Seeley J, Johnson B. Worksite‐based internet multimedia program for family caregivers of persons with dementia. Gerontologist. 2005;45(6):793‐801. 112. Mittelman MS, Brodaty H, Wallen AS, Burns A. A three‐country randomized controlled trial of a psychosocial intervention for caregivers combined with pharmacological treatment for patients with Alzheimer disease: effects on caregiver depression. Am J Geriatr Psychiatry. 2008;16(11):893‐904. 113. Mittelman MS, Roth DL, Coon DW, Haley WE. Sustained benefit of supportive intervention for depressive symptoms in caregivers of patients with Alzheimer's disease. Am J Psychiatry. 2004;161(5):850‐856. 114. Livingston G, Barber J, Rapaport P, et al. Clinical effectiveness of a manual based coping strategy programme (START, STrAtegies for RelaTives) in promoting the mental health of carers of family members with dementia: pragmatic randomised controlled trial. BMJ. 2013;347:f6276. 115. Knapp M, King D, Romeo R, et al. Cost effectiveness of a manual based coping strategy programme in promoting the mental health of family carers of people with dementia (the START (STrAtegies for RelaTives) study): a pragmatic randomised controlled trial. BMJ. 2013;347:f6342. 116. Casarett D, Karlawish J, Morales K, Crowley R, Mirsch T, Asch DA. Improving the use of hospice services in nursing homes: a randomized controlled trial. JAMA. 2005;294(2):211‐217. 117. Volandes AE, Mitchell SL, Gillick MR, Chang Y, Paasche‐Orlow MK. Using video images to improve the accuracy of surrogate decision‐making: a randomized controlled trial. J Am Med Dir Assoc. 2009;10(8):575‐580. 118. Pantilat SZ, O'Riordan DL, Dibble SL, Landefeld CS. Hospital‐based palliative medicine consultation: a randomized controlled trial. Arch Intern Med. 2010;170(22):2038‐2040. 119. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271‐2281. 120. Ward SE, Wang KK, Serlin RC, Peterson SL, Murray ME. A randomized trial of a tailored barriers intervention for Cancer Information Service (CIS) callers in pain. Pain. 2009;144(1‐2):49‐56. 121. Spiro SG, Rudd RM, Souhami RL, et al. Chemotherapy versus supportive care in advanced non‐ small cell lung cancer: improved survival without detriment to quality of life. Thorax. 2004;59(10):828‐836. 122. Walling AM, Schreibeis‐Baum H, Pimstone N, et al. Proactive case finding to concurrently improve curative and palliative care in patients with end‐stage liver disease. J Palliat Med. 2014. 123., Smith TJ Temin S, Alesi ER, et al. American Society of Clinical Oncology provisional clinical opinion: the integration of palliative care into standard oncology care. J Clin Oncol. 2012;30(8):880‐887. 124. Lynn J. Reliable and sustainable comprehensive care for frail elderly people. JAMA. 2013;310(18):1935‐1936. 125. Hickman SE, Tolle SW, Brummel‐Smith K, Carley MM. Use of the Physician Orders for Life‐ Sustaining Treatment program in Oregon nursing facilities: beyond resuscitation status. J Am Geriatr Soc. 2004;52(9):1424‐1429.

126. May P, Normand C, Morrison RS. Economic impact of hospital inpatient palliative care consultation: review of current evidence and directions for future research. J Palliat Med. 2014;17(9):1054‐1063. 127. Greer JA, Tramontano A, McMahon PM, et al. Cost analysis of a randomized trial of early palliative care (PC) in patients with metastatic non‐small cell lung cancer (NSCLC). Boston, MA: Palliative Care in Oncology Symposium;2014. 128. Goldberg LR, Piette JD, Walsh MN, et al. Randomized trial of a daily electronic home monitoring system in patients with advanced heart failure: the Weight Monitoring in Heart Failure (WHARF) trial. Am Heart J. 2003;146(4):705‐712. 129. Melis RJ, van Eijken MI, van Achterberg T, et al. The effect on caregiver burden of a problem‐ based home visiting programme for frail older people. Age Ageing. 2009;38(5):542‐547. 130. Dyar S, Lesperance M, Shannon R, Sloan J, Colon‐Otero G. A nurse practitioner directed intervention improves the quality of life of patients with metastatic cancer: results of a randomized pilot study. J Palliat Med. 2012;15(8):890‐895. 131. Hoekstra J, de Vos R, van Duijn NP, Schade E, Bindels PJ. Using the symptom monitor in a randomized controlled trial: the effect on symptom prevalence and severity. J Pain Symptom Manage. 2006;31(1):22‐30. 132. Brown RF, Butow PN, Sharrock MA, et al. Education and role modelling for clinical decisions with female cancer patients. Health Expect. 2004;7(4):303‐316. 133. Curtis JR, Nielsen EL, Treece PD, et al. Effect of a quality‐improvement intervention on end‐of‐ life care in the intensive care unit: a randomized trial. Am J Respir Crit Care Med. 2011;183(3):348‐355. 134. Fortinsky RH, Kulldorff M, Kleppinger A, Kenyon‐Pesce L. Dementia care consultation for family caregivers: collaborative model linking an Alzheimer's association chapter with primary care physicians. Aging Ment Health. 2009;13(2):162‐170. 135. Grande GE, Farquhar MC, Barclay SI, Todd CJ. Caregiver bereavement outcome: relationship with hospice at home, satisfaction with care, and home death. J Palliat Care. 2004;20(2):69‐77. 136. Cornbleet MA, Campbell P, Murray S, Stevenson M, Bond S. Patient‐held records in cancer and palliative care: a randomized, prospective trial. Palliat Med. 2002;16(3):205‐212. 137. Hepburn KW, Lewis M, Narayan S, et al. Partners in caregiving: a psychoeducation program affecting dementia family caregivers' distress and caregiving outlook. Clin Gerontologist. 2005;29(1):53‐69. 138. Magai C, Cohen CI, Gomberg D. Impact of training dementia caregivers in sensitivity to nonverbal emotion signals. Int Psychogeriatr. 2002;14(1):25‐38. 139. Pillemer K, Suitor JJ. Peer support for Alzheimer's caregivers: is it enough to make a difference? Res Aging. 2002;24(2):171‐192. 140. Wilkie D, Berry D, Cain K, et al. Effects of coaching patients with lung cancer to report cancer pain. West J Nurs Res. 2010;32(1):23‐46. 141. Williams J, Cheung W, Chetwynd N, et al. Pragmatic randomised trial to evaluate the use of patient held records for the continuing care of patients with cancer. Qual Health Care. 2001;10(3):159‐165. 142. Steinhauser KE, Alexander SC, Byock IR, George LK, Olsen MK, Tulsky JA. Do preparation and life completion discussions improve functioning and quality of life in seriously ill patients? Pilot randomized control trial. J Palliat Med. 2008;11(9):1234‐1240. 143. Azoulay E, Pochard F, Chevret S, et al. Impact of a family information leaflet on effectiveness of information provided to family members of intensive care unit patients: a multicenter, prospective, randomized, controlled trial. Am J Respir Crit Care Med. 2002;165(4):438‐442.

144. Epstein AS, Volandes AE, Chen LY, et al. A randomized controlled trial of a cardiopulmonary resuscitation video in advance care planning for progressive pancreas and hepatobiliary cancer patients. J Palliat Med. 2013;16(6):623‐631. 145. Kinley J, Stone L, Dewey M, et al. The effect of using high facilitation when implementing the Gold Standards Framework in Care Homes programme: a cluster randomised controlled trial. Palliat Med. 2014;28(9):1099‐1109. 146. Metzelthin SF, van Rossum E, de Witte LP, et al. Effectiveness of interdisciplinary primary care approach to reduce disability in community dwelling frail older people: cluster randomised controlled trial. BMJ. 2013;347:f5264.

Tables and Figures

Table 4. Patient Populations

Participant Conditions No. Studies Pct. of Studies Conditions described 115 93% Cancer 77 Mixed 60 Metastatic 34 Chemotherapy and/or radiotherapy 26 CHF 21 COPD 19 Dementia 25

Health Services‐Defined Prognosis Participants satisfy health services‐defined prognosis 30 24% Clinician‐defined “poor prognosis” 22 Referred to or receiving palliative care 8 Referred to or receiving hospice 6 Other explicit criteria 0

Table 5. Intervention Components

Multidisciplinary Teams No. Studies Pct. of Studies Any 32 26% Palliative care team 17 Hospice care team 4 Other 11

Clinical Disciplines of Intervention Personnel Delivered by clinically trained individuals 107 86% Disciplines specified 98 Physicians 45 Nurses 61 Advanced practice nurses 12 Social workers 21 Chaplains 14 Mental health professionals 22

Intervention Settings Outpatient clinics 47 38% Home 70 56% Hospital 18 15% Nursing home 6 5% Residential care facilities 4 3% Not specified 11 9%

Technology Any technology 59 48% Telephone support 51 Other (telehealth, EHR‐based tools, mobile computing, other) 16

Intervention Elements Palliative care team 17 14% Inpatient consultation 8 Outpatient consultation 13 Hospice 5 4% Case management 22 18% Decision support 9 7% Family meetings 2 2% Advance directive completion 6 5%

Intervention Duration (Days) Duration described 115 93% One‐time intervention 19 Sustained intervention Minimum 3 Maximum 1826 Median (IQR) 91 (35‐183)

Table 6. Patient and Caregiver Quality of Life‐Relevant Outcomes

No. Studies Addressed Majority Results Pct. of Patient QOL‐Relevant Domains Outcome Significant Studies Patient outcomes described 96 77% Pain 37 10 27% Dyspnea 14 3 21% Depressive symptoms 33 12 36% Anxiety 22 9 41% Other symptoms 40 15 38% Existential or spiritual concerns 11 4 36% Communication or care planning 18 12 67% Experience or satisfaction 20 8 40% Functional status 37 12 32% HRQOL 38 10 26%

Total No. Majority Results Pct. of

Studies Significant Studies Majority of results significant in majority of QOL‐ 96 43 45% relevant domains No. domains addressed 1 24 15 63% 2 12 11 92% 3‐10 60 17 28%

Majority of total number of results significant in total 96 34 35% number of QOL‐relevant outcomes No. outcomes addressed 1 11 9 82% 2‐4 27 11 41% 5‐13 58 14 24%

Addressed Majority Results Pct. of

Outcome Significant Studies Caregiver QOL‐Relevant Domains Caregiver outcomes described 52 42% Depressive symptoms 21 12 57% Anxiety 6 2 33% Other symptoms 8 5 63% Existential or spiritual concerns 3 0 0% Communication or care planning 4 3 75%

Experience or satisfaction 8 5 63% HRQOL 7 3 43%

Total No. Majority Results Pct. of

Studies Significant Studies Majority of results significant in majority of QOL‐ 52 31 60% relevant domains No. domains addressed 1 26 18 69% 2 14 6 43% 3‐5 12 7 58%

Majority of total number of results significant in total 52 30 58% number of QOL‐relevant outcomes No. outcomes addressed 1 18 13 72% 2‐3 13 6 46% 4‐10 21 11 52%

Table 7. Economic Outcomes

No. Studies Addressed Significant Pct. of Economic outcomes Outcome Results Studies Any 51 41% Site of death 9 3 33% Drugs 16 5 31% Medical devices 0 0 0% Procedures 2 0 0% Health care use 36 17 47% Inpatient hospital 27 11 41% ICU 6 1 17% Outpatient 11 1 9% Emergency department 13 4 31% Nursing home 2 0 0% Home health care 4 0 0% Hospice 4 2 50% Referral 2 1 50% Use 3 2 67% Life‐extending devices or procedures 3 1 33% Invasive ventilation 3 1 33% Dialysis 1 0 0% CPR 1 0 0% Costs 19 6 32% Overall 15 3 20% Specific 11 4 36%

Appendix Figure 1. Literature Flow

Appendix Table 7. Literature Search Strategy

SEARCH #1: Quality of life‐relevant outcomes search strategy

DATABASE SEARCHED AND TIME PERIOD COVERED: MEDLINE; 1/1/2004 to 9/23/2011

LANGUAGE: English

SEARCH STRATEGY: death[ti] OR death[mh:noexp] OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR “last year of life” OR “end of life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality of life OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome[ti] OR outcomes[ti] OR consumer satisfaction OR patient satisfaction OR personal satisfaction AND pain/th OR pain/psychology OR “pain management” OR “pain assessment” OR “relieve suffering” OR “relieve symptoms” OR palliative care[mh] OR pain[ti] OR “pain relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology[sh] OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders[mh] OR depression OR depressive disorder[mh] OR depressed OR “attitude to death” OR neoplasms/psychology OR “emotional health” OR spiritual OR emotions OR support[ti] OR supportive OR communication OR relationships OR religion OR religiosity OR “treatment decision” OR decisionmaking OR “decision making” OR home care services/standards OR home nursing/st OR hospice care/st OR “nursing assistance” OR nursing homes/st OR residential facilities/st OR intensive care units/st OR life support care/st OR “home care” OR hospice* OR “nursing homes”[tiab] OR “nursing home”[tiab] OR “intensive care”[tiab] OR icu[tiab] OR icus[tiab] OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family[mh] OR family[tiab] OR families[tiab] OR “social services” OR “social support” AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure” OR “end stage liver disease” OR cirrhosis NOT letter[pt] OR news[pt] OR editorial[pt] OR case reports[pt] NOT ethics[mh] OR euthanasia[mh] OR suicide, assisted[mh] OR pregnancy[mh] OR pregnancy complications[mh] OR fetal death[mh])

NUMBER OF RESULTS:

3,096 SEARCH #2: Variants of quality of life‐relevant outcomes search strategy (variant terms indicated in bold)

DATABASE SEARCHED AND TIME PERIOD COVERED: MEDLINE; 1/1/2004 to 12/28/2011

LANGUAGE: English

SEARCH #2A STRATEGY: death[ti] OR death[mh:noexp] OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR “last year of life” OR “end of life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly AND pain/th OR pain/psychology OR “pain management” OR “pain assessment” OR “relieve suffering” OR “relieve symptoms” OR palliative care[mh] OR pain[ti] OR “pain relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology[sh] OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders[mh] OR depression OR depressive disorder[mh] OR depressed OR “attitude to death” OR neoplasms/psychology OR “emotional health” OR spiritual OR emotions OR support[ti] OR supportive OR communication OR relationships OR religion OR religiosity OR “treatment decision” OR decisionmaking OR “decision making” OR home care services/standards OR home nursing/st OR hospice care/st OR “nursing assistance” OR nursing homes/st OR residential facilities/st OR intensive care units/st OR life support care/st OR “home care” OR hospice* OR “nursing homes”[tiab] OR “nursing home”[tiab] OR “intensive care”[tiab] OR icu[tiab] OR icus[tiab] OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family[mh] OR family[tiab] OR families[tiab] OR “social services” OR “social support” AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality of life OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome[ti] OR outcomes[ti] OR consumer satisfaction OR patient satisfaction OR personal satisfaction AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure” OR “end stage liver disease” OR cirrhosis NOT letter[pt] OR news[pt] OR editorial[pt] OR case reports[pt] OR ethics[mh] OR euthanasia[mh] OR suicide, assisted[mh] OR pregnancy[mh] OR pregnancy complications[mh] OR fetal death[mh]

NUMBER OF RESULTS: 3,229

SEARCH #2B STRATEGY: death[ti] OR “death”[MeSH Terms:noexp] OR dying[tiab] OR “last year of life”[All Fields] OR “end of life”[All Fields] OR “terminal illness”[All Fields] OR “terminal illnesses”[All Fields] OR (“terminal care”[MeSH Terms] OR (“terminal”[All Fields] AND “care”[All Fields]) OR “terminal care”[All Fields]) OR “death and dying”[All Fields] OR “limited life expectancies”[All Fields] OR “limited life expectancy”[All Fields] OR “limited life span”[All Fields] OR “limited lifespan”[All Fields] OR “limited life spans”[All Fields] OR (“terminally ill”[MeSH Terms] OR (“terminally”[All Fields] AND “ill”[All Fields]) OR “terminally ill”[All Fields]) OR (“critical illness”[MeSH Terms] OR (“critical”[All Fields] AND “illness”[All Fields]) OR “critical illness”[All Fields]) OR (“frail elderly”[MeSH Terms] OR (“frail”[All Fields] AND “elderly”[All Fields]) OR “frail elderly”[All Fields]) AND pain/th OR pain/psychology OR “pain management” OR “pain assessment” OR “relieve suffering” OR “relieve symptoms” OR palliative care[mh] OR pain[ti] OR “pain relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology[sh] OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders[mh] OR depression OR depressive disorder[mh] OR depressed OR “attitude to death” OR neoplasms/psychology OR “emotional health” OR spiritual OR emotions OR support[ti] OR supportive OR communication OR relationships OR religion OR religiosity OR “treatment decision” OR decisionmaking OR “decision making” OR home care services/standards OR home nursing/st OR hospice care/st OR “nursing assistance” OR nursing homes/st OR residential facilities/st OR intensive care units/st OR life support care/st OR “home care” OR hospice* OR “nursing homes”[tiab] OR “nursing home”[tiab] OR “intensive care”[tiab] OR icu[tiab] OR icus[tiab] OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family[mh] OR family[tiab] OR families[tiab] OR “social services” OR “social support” AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality of life OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome[ti] OR outcomes[ti] OR consumer satisfaction OR patient satisfaction OR personal satisfaction AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure” OR “end stage liver disease” OR cirrhosis NOT letter[pt] OR news[pt] OR editorial[pt] OR case reports[pt] OR ethics[mh] OR euthanasia[mh] OR suicide, assisted[mh] OR pregnancy[mh] OR pregnancy complications[mh] OR fetal death[mh] NOT [RESULTS OF SEARCH 2A]

NUMBER OF RESULTS: 179

SEARCH #2C STRATEGY: death[ti] OR death[mh:noexp] OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR dying[tiab] OR “last year of life” OR “end of life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly OR palliative care OR hospice AND pain/th OR pain/psychology OR “pain management” OR “pain assessment” OR “relieve suffering” OR “relieve symptoms” OR pain[ti] OR “pain relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology[sh] OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders[mh] OR depression OR depressive disorder[mh] OR depressed OR “attitude to death” OR neoplasms/psychology OR “emotional health” OR spiritual OR emotions OR support[ti] OR supportive OR communication OR relationships OR religion OR religiosity OR “treatment decision” OR decisionmaking OR “decision making” OR home care services/standards OR home nursing/st OR “nursing assistance” OR nursing homes/st OR residential facilities/st OR intensive care units/st OR life support care/st OR “home care” OR “nursing homes”[tiab] OR “nursing home”[tiab] OR “intensive care”[tiab] OR icu[tiab] OR icus[tiab] OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family[mh] OR family[tiab] OR families[tiab] OR “social services” OR “social support” AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality of life OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome[ti] OR outcomes[ti] OR consumer satisfaction OR patient satisfaction OR personal satisfaction AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure” OR “end stage liver disease” OR cirrhosis NOT letter[pt] OR news[pt] OR editorial[pt] OR case reports[pt] OR ethics[mh] OR euthanasia[mh] OR suicide, assisted[mh] OR pregnancy[mh] OR pregnancy complications[mh] OR fetal death[mh] NOT [RESULTS OF SEARCHES 2A OR 2B]

NUMBER OF RESULTS: 2,155

SEARCH #3 Sub‐search A: quality of life‐relevant outcomes search strategy with original disease terms Sub‐search B: quality of life‐relevant outcomes search strategy with new disease terms

SEARCH #3A1:

DATABASE SEARCHED AND TIME PERIOD COVERED: MEDLINE; 1/1/2007 to 2/2/2012

SEARCH STRATEGY: death[ti] OR death[mh:noexp] OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR dying[tiab] OR “last year of life” OR “end of life” OR “end‐of‐life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly OR hospice OR “palliative care” OR palliat* AND pain/th OR pain/psychology OR “pain management” OR “pain assessment” OR “relieve suffering” OR “relieve symptoms” OR pain[ti] OR “pain relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology[sh] OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders[mh] OR depression OR depressive disorder[mh] OR depressed OR “attitude to death” OR neoplasms/psychology OR “emotional health” OR emotions OR support[ti] OR supportive OR communication OR relationships OR “treatment decision” OR decisionmaking OR “decision making” OR home care services/standards OR home nursing/st OR “nursing assistance” OR nursing homes/st OR residential facilities/st OR intensive care units/st OR life support care/st OR “home care” OR “nursing homes”[tiab] OR “nursing home”[tiab] OR “intensive care”[tiab] OR icu[tiab] OR icus[tiab] OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family[mh] OR family[tiab] OR families[tiab] OR “social services” OR “social support” OR bereav* OR grief OR spiritual* OR religio* OR “quality of life” OR quality of life[mh] AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome[ti] OR outcomes[ti] OR consumer satisfaction OR patient satisfaction OR personal satisfaction OR cost OR costs OR cost benefit OR econom* OR cost analys* OR financial OR utiliz* AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure”

NUMBER OF RESULTS: 4761

SEARCH #3A2

DATABASE SEARCHED AND TIME PERIOD COVERED: Cochrane; 1/1/2007 to 2/2/2012

SEARCH STRATEGY: “death” OR death OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR dying OR “last year of life” OR “end of life” OR “end‐of‐life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly OR hospice OR “palliative care” OR palliat* in Title, Abstract or Keywords AND pain OR “relieve suffering” OR “relieve symptoms” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders OR depression OR depressive disorder OR depressed OR “attitude to death” OR psychology OR “emotional health” OR emotions OR support OR supportive OR communication OR relationships OR “treatment decision” OR decisionmaking OR “decision making” OR home care services OR home nursing OR “nursing assistance” OR nursing homes OR residential facilities OR intensive care units OR life support care OR “home care” OR “nursing homes” OR “nursing home” OR “intensive care” OR icu OR icus OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family OR families[tiab] OR “social services” OR “social support” OR bereav* OR grief OR spiritual* OR religio* OR “quality of life” in Title, Abstract or Keywords AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome* OR consumer satisfaction OR patient satisfaction OR personal satisfaction OR cost OR costs OR econom* OR financial OR utiliz* in Title, Abstract or Keywords AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure” in Title, Abstract or Keywords

NUMBER OF RESULTS: Cochrane Reviews: 44 Other Reviews: 4 Clinical Trials: 192 Methods Studies: 1 Technology Assessments: 1 Economic Evaluations: 7 Cochrane Groups: 0

SEARCH #3B1

DATABASE SEARCHED AND TIME PERIOD COVERED: MEDLINE; 1/1/2001 to 2/2/2012

SEARCH STRATEGY: death[ti] OR death[mh:noexp] OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR dying[tiab] OR “last year of life” OR “end of life” OR “end‐of‐life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly OR hospice OR “palliative care” OR palliat* AND pain/th OR pain/psychology OR “pain management” OR “pain assessment” OR “relieve suffering” OR “relieve symptoms” OR pain[ti] OR “pain relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology[sh] OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders[mh] OR depression OR depressive disorder[mh] OR depressed OR “attitude to death” OR neoplasms/psychology OR “emotional health” OR emotions OR support[ti] OR supportive OR communication OR relationships OR “treatment decision” OR decisionmaking OR “decision making” OR home care services/standards OR home nursing/st OR “nursing assistance” OR nursing homes/st OR residential facilities/st OR intensive care units/st OR life support care/st OR “home care” OR “nursing homes”[tiab] OR “nursing home”[tiab] OR “intensive care”[tiab] OR icu[tiab] OR icus[tiab] OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family[mh] OR family[tiab] OR families[tiab] OR “social services” OR “social support” OR bereav* OR grief OR spiritual* OR religio* OR “quality of life” OR quality of life[mh] AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome[ti] OR outcomes[ti] OR consumer satisfaction OR patient satisfaction OR personal satisfaction OR cost OR costs OR cost benefit OR econom* OR cost analys* OR financial OR utiliz* AND “end stage liver disease” OR cirrhosis OR “chronic lung disease” OR copd OR “chronic obstructive pulmonary”

NUMBER OF RESULTS: 607

SEARCH #3B2

DATABASE SEARCHED AND TIME PERIOD COVERED: Cochrane; 1/1/2001 to 2/2/2012

SEARCH STRATEGY: “death” OR death OR “dying loved one” OR “dying patient” OR “dying patients” OR “dying people” OR “dying person” OR dying OR “last year of life” OR “end of life” OR “end‐of‐life” OR “terminal illness” OR “terminal illnesses” OR terminal care OR “death and dying” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR terminally ill OR critical illness OR frail elderly OR hospice OR “palliative care” OR palliat* in Title, Abstract or Keywords AND pain OR “relieve suffering” OR “relieve symptoms” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “symptom relief” OR “pain distress” OR “pain easing” OR “pain free” OR “psychological distress” OR psychology OR wellbeing OR “well being” OR anxiety OR anxious OR anxiety disorders OR depression OR depressive disorder OR depressed OR “attitude to death” OR psychology OR “emotional health” OR emotions OR support OR supportive OR communication OR relationships OR “treatment decision” OR decisionmaking OR “decision making” OR home care services OR home nursing OR “nursing assistance” OR nursing homes OR residential facilities OR intensive care units OR life support care OR “home care” OR “nursing homes” OR “nursing home” OR “intensive care” OR icu OR icus OR “place of death” OR health care facilities, manpower and services OR caregiver* OR caregivers OR “care giving” OR family OR families[tiab] OR “social services” OR “social support” OR bereav* OR grief OR spiritual* OR religio* OR “quality of life” in Title, Abstract or Keywords AND delivery of health care OR quality assurance, health care OR “outcome and process assessment (health care)” OR quality indicators OR quality of health care OR patient care management OR continuity of care OR outcome* OR consumer satisfaction OR patient satisfaction OR personal satisfaction OR cost OR costs OR econom* OR financial OR utiliz* in Title, Abstract or Keywords AND “end stage liver disease” OR cirrhosis OR “chronic lung disease” OR copd OR “chronic obstructive pulmonary” in Title, Abstract or Keywords

NUMBER OF RESULTS: Cochrane Reviews: 9 Other Reviews: 0 Clinical Trials: 39 Methods Studies: 1 Technology Assessments: 0 Economic Evaluations: 0 Cochrane Groups: 0

SEARCH #4 Economic outcomes search strategy

SEARCH #4A

DATABASE SEARCHED AND TIME PERIOD COVERED: MEDLINE; 1/1/2001 to 2/1/2012

LANGUAGE: English

NUMBER OF RESULTS: 6270

NUMBER OF RESULTS AFTER REMOVAL OF DUPLICATES: 4580

SEARCH #4B

DATABASE SEARCHED AND TIME PERIOD COVERED: Cochrane; 1/1/2001 to 2/1/2012

LANGUAGE: English

NUMBER OF RESULTS: Cochrane Reviews: 43 Other Reviews: 11 Clinical Trials: 269 Methods Studies: 4 Technology Assessments: 1 Economic Evaluations: 43 Cochrane Groups: 0

SEARCH #5 Sub‐search A: quality of life‐relevant outcomes search strategy (systematic reviews) Sub‐search B: costs search strategy (systematic reviews)

SEARCH #5A1

DATABASE SEARCHED AND TIME PERIOD COVERED: EMBASE; 1/1/2001 to 5/14/2012

LANGUAGE: English

OTHER LIMITERS: Human

SEARCH STRATEGY: ‘death’/exp OR death OR ‘dying’/exp OR dying OR ‘last year of life’ OR ‘end of life’ OR ‘end‐of‐life’ OR ‘terminal illness’/exp OR ‘terminal illness’ OR ‘terminal illnesses’ OR terminal AND care OR ‘death and dying’ OR ‘limited life expectancies’ OR ‘limited life expectancy’ OR ‘limited life span’ OR ‘limited lifespan’ OR ‘limited life spans’ OR terminally AND ill OR critical AND (‘illness’/exp OR illness) OR frail AND (‘elderly’/exp OR elderly) OR ‘hospice’/exp OR hospice OR ‘palliative care’/exp OR ‘palliative care’ OR palliat* AND ‘pain’/exp OR pain OR ‘relieve suffering’ OR ‘relieve symptoms’ OR discomfort OR ‘physical comfort’ OR ‘comfort care’ OR ‘symptom distress’ OR ‘symptom burden’ OR ‘symptom control’ OR ‘symptom intensity’ OR ‘symptom management’ OR ‘symptom relief’ OR ‘pain distress’ OR ‘pain easing’ OR ‘pain free’ OR ‘psychological distress’/exp OR ‘psychological distress’ OR ‘psychology’/exp OR psychology OR ‘wellbeing’/exp OR wellbeing OR ‘well being’/exp OR ‘well being’ OR anxious OR ‘anxiety’/exp OR anxiety AND disorders OR ‘depression’/exp OR depression OR depressive AND (‘disorder’/exp OR disorder) OR depressed OR ‘attitude to death’/exp OR ‘attitude to death’ OR ‘psychology’/exp OR psychology OR ‘emotional health’/exp OR ‘emotional health’ OR ‘emotions’/exp OR emotions OR support OR supportive OR ‘communication’/exp OR communication OR relationships OR ‘treatment decision’ OR decisionmaking OR ‘decision making’/exp OR ‘decision making’ OR ‘home’/exp OR home AND care AND services OR ‘home’/exp OR home AND (‘nursing’/exp OR nursing) OR ‘nursing assistance’/exp OR ‘nursing assistance’ OR ‘nursing’/exp OR nursing AND homes OR residential AND facilities OR intensive AND care AND units OR ‘life’/exp OR life AND support AND care OR ‘home care’/exp OR ‘home care’ OR ‘nursing homes’/exp OR ‘nursing homes’ OR ‘nursing home’/exp OR ‘nursing home’ OR ‘intensive care’/exp OR ‘intensive care’ OR icu OR icus OR ‘place of death’ OR ‘health’/exp OR health AND care AND facilities, AND (‘manpower’/exp OR manpower) AND services OR caregiver* OR ‘caregivers’/exp OR caregivers OR ‘care giving’ OR ‘family’/exp OR family OR families OR ‘social services’ OR ‘social support’/exp OR ‘social support’ OR bereav* OR ‘grief’/exp OR grief OR spiritual* OR religio* OR ‘quality of life’/exp OR ‘quality of life’ AND ‘cancer’/exp OR cancer OR ‘neoplasms’/exp OR neoplasms OR ‘dementia’/exp OR dementia OR alzheimer* OR ‘heart failure’/exp OR ‘heart failure’ OR ‘end stage liver disease’/exp OR ‘end stage liver disease’ OR ‘cirrhosis’/exp OR cirrhosis OR ‘chronic lung disease’/exp OR ‘chronic lung disease’ OR ‘copd’/exp OR copd OR ‘chronic obstructive pulmonary’ AND 84

‘systematic review’/exp OR ‘systematic review’ OR ‘systematic reviews’/exp OR ‘systematic reviews’ OR metaanaly* OR ‘meta analysis’/exp OR ‘meta analysis’ OR ‘meta analyses’ OR ‘meta analytic’ OR ‘meta analytical’

NUMBER OF RESULTS: 456

SEARCH #5A2

DATABASE SEARCHED AND TIME PERIOD COVERED: Web of Science Databases (SCI‐EXPANDED, SSCI, A&HCI, CPCI‐S, and CPCI‐SSH); 1/1/2001 to 5/18/2012

LANGUAGE: English

OTHER LIMITERS: Human

SEARCH STRATEGY: Topic=(dying OR “last year of life” OR “end of life” OR “end‐of‐life” OR “terminal illness” OR “terminal illnesses” OR “terminal care”) OR Topic=(“terminally ill” OR “critically ill” OR “critical illness” OR “frail elderly” OR “vulnerable elderly” OR hospice OR palliat*) OR Topic=(“limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans”) AND TOPIC=(pain OR “relieve suffering” OR “relieve symptoms’“ OR “symptom relief” OR discomfort OR “physical comfort” OR “comfort care” OR “symptom distress” OR “symptom burden” OR “symptom control” OR “symptom intensity” OR “symptom management” OR “pain‐free” OR “psychological distress” OR psychology OR wellbeing OR “well being” OR anxious OR anxiety OR depression OR depressive OR depressed OR “attitude to death” OR psychology OR “emotional health” OR emotions OR support OR supportive OR communication OR relationships OR “treatment decision” OR decisionmaking OR “decision making” OR “home care” OR “nursing assistance” OR “nursing homes” OR “nursing home” OR “residential facilities” OR “intensive care unit” OR “intensive care units” OR “life support” OR “home care” OR “nursing homes” OR “nursing home” OR “intensive care” OR icu OR icus OR “place of death” OR (health care AND facilit*) OR (healthcare AND facilit*) OR (manpower AND services) OR caregiver* OR caregivers OR “care giving” OR family OR families OR “social services” OR “social support” OR bereav* OR grief OR spiritual* OR religio* OR “quality of life”) AND Topic=(cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure’“ OR “end stage liver disease’“ OR cirrhosis OR “chronic lung disease” OR “chronic lung disease” OR copd OR “chronic obstructive pulmonary”) AND Topic=(“systematic review” OR “systematic reviews” OR metaanaly* OR “meta analysis” OR “meta analyses” OR “‘meta analytic” OR “‘meta analytical”)

NUMBER OF RESULTS: 607

SEARCH #5A3

DATABASE SEARCHED AND TIME PERIOD COVERED: PsycINFO; 1/1/2001 to 5/30/2012

LANGUAGE: English

SEARCH STRATEGY: dying OR “last year of life” OR “limited life expectancies” OR “limited life expectancy” OR “limited life span” OR “limited lifespan” OR “limited life spans” OR “end of life” OR “end‐of‐life” OR “terminal illness” OR “terminal illnesses” OR “terminal care” OR “terminally ill” OR “critically ill” OR “critical illness” OR “frail elderly” OR “vulnerable elderly” OR hospice OR palliat* AND cancer OR neoplasms OR dementia OR alzheimer* OR “heart failure’“ OR “end stage liver disease’“ OR cirrhosis OR “chronic lung disease” OR “chronic lung disease” OR copd OR “chronic obstructive pulmonary” AND “systematic review” OR “systematic reviews” OR metaanaly* OR “meta analysis” OR “meta analyses” OR “‘meta analytic” OR “‘meta analytical” ) in all fields

Boolean phrase search

NUMBER OF RESULTS: 183

NUMBER OF RESULTS AFTER REMOVAL OF DUPLICATES FROM OTHER SYSTEMATIC REVIEWES: SEARCH 149

SEARCH #5B1

DATABASE SEARCHED AND TIME PERIOD COVERED: EMBASE; 1/1/2001 to 5/14/2012

LANGUAGE: English

OTHER LIMITERS: Human

NUMBER OF RESULTS: 157

SEARCH #5B2

DATABASE SEARCHED AND TIME PERIOD COVERED: Web of Science Databases (SCI‐EXPANDED, SSCI, A&HCI, CPCI‐S, and CPCI‐SSH); 1/1/2001 to 5/18/2012

LANGUAGE: English

OTHER LIMITERS: Human

NUMBER OF RESULTS: 119

SEARCH #6

DATABASE SEARCHED AND TIME PERIOD COVERED: CareSearch Palliative Care Knowledge Network, Review Collection; Inception to 5/21/2012 http://www.caresearch.com.au/Caresearch/Default.aspx “This collection provides an easy way to find reviews on topics relevant to palliative care.”

Care Issues End‐of‐life Pain

Disease Groups Cancer COPD General Heart failure Liver Failure Non‐malignant

NUMBER OF RESULTS AFTER REMOVAL OF DUPLICATES FROM ALL PREVIOUS SEARCHES: 17

Appendix Table 8. Acronyms Used in Evidence Tables

Acronym Meaning AACS Assessment of Awareness about Communication Strategies ABIM American Board of Internal Medicine ACE Angiotensin‐converting enzyme ADL Activities of daily living ADRDA Alzheimer’s Disease and Related Disorders Association AHRQ Agency for Healthcare Research and Quality AICCP Advanced Illness Coordinated Care Program AKPS Australia‐modified Karnofsky Performance Status ALS Amyotrophic lateral sclerosis AMPS Assessment of Motor and Process Skills APS American Pain Society AQEL Assessment of Quality of life at the End of Life ARB Angiotensin receptor blocker BACS Beliefs About Caregiving Scale BDI Beck Depression Inventory BDI‐II Beck Depression Inventory‐II BDI‐SF Beck Depression Inventory‐Short Form BEHAVE‐AD Behavioral Pathology in Alzheimer’s Disease Rating Scale BFI‐I Brief Fatigue Inventory‐Interference BHS Beck Hopelessness Scale BMMRS Brief Multidimensional Measure of Religion and Spirituality BPI Brief Pain Inventory BPI‐SF Brief Pain Inventory‐Short Form BPP Biobehavioral Pain Profile BQ‐II Barriers Questionnaire‐II BQ‐r Barriers Questionnaire‐revised BSI Brief Symptom Inventory CAD Coronary artery disease CAS Constipation Assessment Scale CASE Cancer Self‐Efficacy Scale CBT Cognitive Behavioral Therapy CCFNI Critical Care Family Needs Inventory CCI Charlson Comorbidity Index CDS Cornell Scale for Depression in Dementia CES‐D Center for Epidemiologic Studies Depression Scale CHF Congestive heart failure CIS Cancer Information Service CMAI Cohen‐Mansfield Agitation Inventory

CNQ Cancer Needs Questionnaire COPD Chronic obstructive pulmonary disease COPE Creativity, Optimism, Planning and Expert Information CPR Cardiopulmonary resuscitation CQOLC Caregiver Quality of Life Index‐Cancer CRA Caregiver Reaction Assessment CRQ Chronic Respiratory Questionnaire CSDD Cornell Scale for Depression in Dementia CSI Caregiver Strain Index CSQ Coping Strategies Questionnaire CT Computed tomography CTIS Computer‐telephone integrated system DNR/DNI Do not resuscitate / do not intubate DPOA‐HC Durable Power of Attorney for Health Care Dqol Dementia Quality of Life Instrument DSM Diagnostic and Statistical Manual of Mental Disorders DWI‐R Dealing With Illness Inventory‐Revised ECOG Eastern Cooperative Oncology Group ED Emergency department eDSMP Internet‐based dyspnea self‐management program EFT Emotionally Focused Therapy EORTC QLQ‐ European Organization for Research Treatment for Cancer Quality of Life C30 Questionnaire Core 30 EPIC Expanded Prostate Cancer Index Composite EQ‐5D EuroQol‐5 Dimensions EQ‐VAS EuroQol‐Visual Analogue Scale ESAS Edmonton Symptom Assessment Scale ESAS Edmonton Symptom Assessment System ESLD End‐stage liver disease ESRD End‐stage renal disease EU European Union FACIT‐F Functional Assessment of Chronic Illness Therapy – Palliative Care Scale FACIT‐Pal Functional Assessment of Chronic Illness Therapy – Palliative Care Scale FACIT‐Sp Functional Assessment of Chronic Illness Therapy – Spiritual Well‐Being Scale FACIT‐TS‐PS Functional Assessment of Chronic Illness Therapy – Treatment Satisfaction – Patient Satisfaction FACT‐B Functional Assessment of Cancer Therapy – Breast FACT‐G Functional Assessment of Cancer Therapy – General FACT‐L Functional Assessment of Cancer Therapy – Lung fDSMP Face‐to‐face dyspnea self‐management program FES‐I Falls Efficacy Scale‐International

FEV1 Forced expiratory volume in one second FLACC Faces Legs Activity Cry Consolability Behavioral Pain Scale FLSI Functional Assessment of Cancer Therapy Lung Symptom Index FOCUS Family involvement, Optimistic attitude, Coping effectiveness, Uncertainty reduction, and Symptom management FPQ Family Pain Questionnaire FS‐ICU Family Satisfaction in the ICU FVC Forced vital capacity GARS Groningen Activity Restriction Scale GDS Geriatric Depression Scale GHQ‐12 General Health Questionnaire‐12 GHQ‐28 General Health Questionnaire‐28 GOLD Global Initiative for Chronic Obstructive Lung Disease GP General practitioner GPS Gracely Pain Scale GSFCH Gold Standards Framework for Care Homes HADS Hospital Anxiety and Depression Scale HAM‐D Hamilton Depression Rating Scale HQLI Hospice Quality‐of‐Life Index HQLI‐14 Hospice Quality of Life Index‐14 HRQOL Health‐related quality of life IADL Instrumental activities of daily living ICP Integrated Care Plan for the last days of life ICU Intensive care unit IDDD Interview of deterioration in daily activities in dementia IES Impact of Event Scale IQR Interquartile range KPS Karnofsky Performance Status LASA Linear Analog Scales of Assessment LCP Liverpool Care Pathway LCS Lung Cancer Subscale MAI Multilevel Assessment Inventory MAOI Monoamine oxidase inhibitor MCOHPQ Modified City of Hope Patient Questionnaires MCS Mental Component Summary MCTS Modified Conflict Tactics Scale MFI Multidimensional Fatigue Inventory MISS Mutuality and Interpersonal Sensitivity Scale MLHFQ Minnesota Living with Heart Failure Questionnaire MMSE Mini‐Mental State Examination MOS Medical Outcomes Study

MOS‐PAQ Medical Outcomes Study Patient Assessment Questionnaire MOS‐VSQ Medical Outcomes Study Visit‐Specific Questionnaire MP Minimum Protocol MPQ McGill Pain Questionnaire MRC Medical Research Council MRI Magnetic resonance imaging MS Multiple sclerosis MSAS Memorial Symptom Assessment Scale MSAS‐SF Memorial Symptom Assessment Scale – Short Form MSPP Maastricht Social Participation Profile NCI National Cancer Institute NHS National Health Service NINCDS National Institute of Neurological and Communicative Disorders and Stroke NRS Numeric Rating Scale NSAID Nonsteroidal anti‐inflammatory drug NSCLC Non‐small cell lung cancer NYU New York University OARS Older Americans Resource and Services Multidimensional Functional Assessment Questionnaire OASIS Outcome Assessment and Information Set OSQ Omega Screening Questionnaire PAINAD Pain in Advanced Dementia scale PANAS Positive and Negative Affect Schedule PCS Physical Component Summary PHQ‐2 Patient Health Questionnaire‐2 PHQ‐8 Patient Health Questionnaire‐8 PHQ‐9 Patient Health Questionnaire‐9 PMI Pain Management Index POMS Profile of Mood States POMS‐B Profile of Mood States‐B POMS‐SF Profile of Mood States Short Form POS‐8 Palliative Care Outcome Scale PPS Palliative Performance Scale PQLI Palliative Care Quality of Life Index PRI Pain Rating Index PRQ85 Personal Resource Questionnaire 85 PTGI Posttraumatic Growth Inventory PTSD Posttraumatic stress disorder QLI‐CV Quality of Life Index‐Cancer Version QOC Quality of communication QODD Quality of Dying and Death

QOL Quality of life QOL‐AD Quality of Life in Alzheimer’s Disease QOLS Quality of Life Scale QUAL‐E Quality of Life at the End of Life RCT Randomized controlled trial RDAS Revised Dyadic Adjustment Scale RDRS‐2 Rapid Disability Rating Scale‐2 REACH Resources for Enhancing Alzheimer’s Caregiver Health RIDcancerPain Representational Intervention to Decrease Cancer Pain RMBPC Revised Memory and Behavior Problems Checklist RSS Relative Stress Scale SAS Specific Activity Scale SBI‐C Spontaneous Behavior Interview Section C SCB Screen for Caregiver Burden SCQ Sense of Competence Questionnaire SD Standard deviation SDMT Symbol Digit Modalities Test SDS Symptom Distress Scale SDS‐15 Symptom Distress Scale 15 SET Structural Ecosystems Therapy SF‐12 12‐item Short Form Health Survey SF‐36 36‐item Short Form Health Survey SGRQ St. Georges Respiratory Questionnaire SISC Structured Interview for Symptoms and Concerns SMBQ Symptom Management Barriers Questionnaire SPMSQ Short Portable Mental Status Questionnaire SPMSQ 18 Short Portable Mental Status Questionnaire 18 SRQ Symptom Representation Questionnaire SSL‐I12 Social Support List‐Interaction version STAI State‐Trait Anxiety Inventory SWD Satisfaction with Decision TACI Thomas Assessment of Communication Inadequacy TBI Tailored Barriers Intervention TCA Tricyclic antidepressant TICS‐m Telephone Interview for Cognitive Status‐modified TOI Trial Outcome Index TPQM Total Pain Quality Management TRIG Texas Revised Inventory of Grief UCD University of California, Davis VAS Visual Analog Scale VES‐13 Vulnerable Elders Survey‐13

WONCA World Organization of National Colleges, Academies, and Academic Associations of General Practitioners/Family Physicians WRITE Written Representational Intervention To Ease Symptoms Symptoms ZBI Zarit Burden Interview ZBI‐12 Zarit Burden Interview 12‐item version

Appendix Table 9. Evidence Tables

Evidence Table 1: Populations and Interventions

Control and Inclusion and Setting(s) of Care, Intervention First Exclusion Criteria, Patient Population Symptom(s) Multidisciplinarity, and Description, Author, Number Participants Condition(s) and and Functional Status Clinical Disciplines of Intervention Year Study Elements and Decedents Prognosis at Baseline Intervention Participants Duration Overall Study Result

Rabow et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 44 Outcomes 2004 US Criteria: Cancer (type not Dyspnea: control, 85%; Control Arm - Outpatient clinics Usual care Study Arms - 2 Patients were eligible for described), CHF, COPD intervention, 84%; p=.90. Comparing a palliative care Intervention Arm - Outpatient consultation intervention with Level of inclusion if they had a Pain: control, 77%; clinics, home Intervention Arm: diagnosis of cancer, advanced usual care, the study Randomization - intervention, 90%; p=.18. Palliative care team, COPD, or advanced CHF, and Prognosis: demonstrated significant Sites Anxiety: control, 85%; palliative case improvements in the intervention life expectancy of 1 to 5 years. Identified by clinicians as Multidisciplinary Team and management, decision Study Duration - 12 intervention, 80%; p=.54. group in anxiety, the frequency They were excluded if they poor prognosis Disciplines: support, family meetings, with which dyspnea limits months had a diagnosis of advance directive Depression: CES-D greater Control Arm - No multidisciplinary activities, spiritual well-being, nonmelanoma skin cancer, completion, family Risk of Bias - 5 than or equal to 16, control, team; clinical disciplines not and completion of funeral dementia, or psychosis, were caregiver training and 74%; intervention, 48%; specified. arrangements. The study did not enrolled in hospice, or were support, medical chart p=.83. review of patient demonstrate significant changes unable to complete a written Intervention Arm - Palliative care in the degree to which dyspnea medications, spiritual survey in English or Spanish. team; physicians, nurses, social and psychological limits activities, pain, depression, Patient Functional Status: workers, chaplains, pharmacists, sleep quality and duration, support, support groups, functional status, overall RDRS-2 score (18-72 scale, medical and pharmacy students. art projects Number of Patient HRQOL, satisfaction with care, lower is better): control, mean Participants and Decedents: or durable power of attorney 25.1 (SD 7.8); intervention, Palliative care team completion. Screened - 330 mean 27.5 (SD 6.2); p=.11. elements:

Enrolled - 90 Outpatient consultation Health Care Use Completed - 66 Caregiver Symptoms: The study demonstrated a Decedents - 15 None described Palliative case management elements: significant reduction in the intervention group in clinic visits Nurse-led case and urgent care visits. The study management, social did not demonstrate significant worker-led case changes in specialist visits, ED management visits, hospital admissions, or total hospital days. Decision support elements: Site of Death Patient and family education The study did not demonstrate a significant change among the

likelihoods of dying at home or in Advance directive a nursing home, in the hospital, completion elements: or in hospice care. DPOA-HC, funeral arrangements Health Care Costs

The study did not demonstrate Intervention Duration: significant changes in overall 12 months medical center services costs or in costs for clinic visits, urgent care visits, ED visits, or inpatient services. 97

Evidence Table 2: Outcomes

First Clinical Significant Clinical Significant Economic Author, Quality of Life Measures and Economic Measures and Year Study Elements Outcome(s) and Measures Result(s) Outcome(s) Result(s) Overall Study Result

Rabow et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 44 Outcomes 2004 US Pain - BPI hospital, outpatient visits, ED Number of clinic visits: control, Study Arms - 2 Site of Death - Died at home or mean 10.6 (SD 7.5); Comparing a palliative care Dyspnea - University of California, Dyspnea: consultation intervention with Level of San Diego Shortness of Breath nursing home vs. hospital vs. intervention, mean 7.5 (SD 4.9); Frequency dyspnea limits activities (0- hospice p=.03. usual care, the study Randomization - Questionnaire (whole scale and 18 scale, lower is better): control, six demonstrated significant Sites degree dyspnea interferes with daily months, mean 6.5; 12 months, mean Health Care Costs - Charges for Number of urgent care visits: improvements in the intervention Study Duration - 12 activities subscale) 7.1; intervention, six months, mean all medical center services, clinic control, mean 0.6 (SD 0.9); group in anxiety, the frequency months Depression - CES-D 5.8; 12 months, mean 3.6; group by visits, urgent care visits, ED intervention, mean 0.3 (SD 0.5); with which dyspnea limits time p=.01. visits, inpatient services, and p=.04. activities, spiritual well-being, Risk of Bias - 5 Anxiety - POMS other and completion of funeral Existential or Spiritual Well-Being - arrangements. The study did not Spiritual Well-Being Scale (whole Anxiety: demonstrate significant changes in the degree to which dyspnea scale and existential and religious Anxiety (0-24 scale, lower is better): limits activities, pain, depression, subscales) control, six months, mean 5.5; 12 sleep quality and duration, Communication or Care Planning - months, mean 5.9; intervention, six functional status, overall DPOA-HC completion, funeral months, mean 6.8; 12 months, mean HRQOL, satisfaction with care, arrangements 5.3; group by time p=.05. or durable power of attorney completion. Experience or Satisfaction - Group

Health Association of America Existential or Spiritual Well-Being: Consumer Satisfaction Survey Health Care Use Overall spiritual well-being (20-120 The study demonstrated a Functional Status - RDRS-2 scale, higher is better): control, six significant reduction in the HRQOL - Multidimensional QOL months, mean 91.2; 12 months, mean intervention group in clinic visits Scale - Cancer Version 92.4; intervention, six months, mean and urgent care visits. The study Other QOL-Relevant Outcomes - 98.0; 12 months, mean 105.5; group did not demonstrate significant Sleep quality and duration (MOS by time p=.05. changes in specialist visits, ED sleep items) visits, hospital admissions, or total hospital days. Communication or Care Planning:

Caregiver Outcomes: Completion of funeral arrangements Site of Death None described among participants who had not completed them at baseline: 12 The study did not demonstrate a months, control, 5%; intervention, 35%; significant change among the p=.03. likelihoods of dying at home or in a nursing home, in the hospital, or in hospice care. Caregiver Outcomes: N/A Health Care Costs The study did not demonstrate significant changes in overall medical center services costs or in costs for clinic visits, urgent care visits, ED visits, or inpatient services.

Evidence Table 1: Populations and Interventions

Jordhoy et Study Location - Inclusion Exclusion Criteria: Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 45 al., 2001 EU Patients were eligible for Cancer (mixed cancer) Fatigue: control, mean 64 (SD Control Arm - Outpatient clinics, Usual care Outcomes Study Arms - 2 inclusion if they were at least 26); intervention, mean 63 hospital Comparing a palliative care 18 years old, had incurable, (SD 28). program intervention in a Level of Prognosis: Intervention Arm - Outpatient Intervention Arm: Randomization - malignant cancer, and had life Nausea vomiting: control, clinics, home, hospital palliative medicine unit with Sites expectancy between 2 and 9 Identified by clinicians as mean 26 (SD 30); Palliative care team, usual care, the study did not months. poor prognosis palliative case demonstrate any significant Study Duration - intervention, mean 23 (SD 29). Multidisciplinary and Clinical management, family changes in any QOL Two years Disciplines: meetings measures, including physical, Number of Patient Pain: control, mean 48 (SD role, emotional, cognitive, Risk of Bias - 3 Control Arm - No multidisciplinary Participants and Decedents: 34); intervention, mean 47 social, or global health team; physicians, nurses. Screened - Not provided (SD 37). Palliative care team functioning; or fatigue, nausea Intervention Arm - palliative care elements: and vomiting, pain, dyspnea, Enrolled - 434 Dyspnea: control, mean 42 (SD 37); intervention, mean team; physicians, nurses, social Inpatient consultation, anorexia, diarrhea, Completed - 108 37 (SD 36). workers, chaplains, physical outpatient constipation, or sleep therapists, nutritionists. disturbances. Decedents - 395 Anorexia: control, mean 53 consultation, home (SD 39); intervention, mean care and GP visits 46 (SD 40). Diarrhea: control, mean 27 Palliative case (SD 34); intervention, mean management 23 (SD 33). elements: Constipation: control, mean 41 Nurse-led case (SD 40); intervention, mean management 44 (SD 38).

Sleep disturbances: control, Intervention mean 41 (SD 36); Duration: intervention, mean 39 (SD 35). Six months

Patient Functional Status: EORTC QLQ-C30 physical functioning scale (0-100 scale, higher is better): control, mean 48 (SD 29); intervention, mean 46 (SD 32). KPS score less than or equal to 70 (0-100 scale, higher is better): control, 41%; intervention, 37%.

Caregiver Symptoms: None described

Evidence Table 2: Outcomes

Jordhoy et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 45 al., 2001 EU Pain - EORTC QLQ-C30 pain scale Outcomes Study Arms - 2 Dyspnea - EORTC QLQ-C30 Comparing a palliative care Level of dyspnea single item program intervention in a palliative medicine unit with Randomization - Other Symptoms - EORTC QLQ-C30 Sites usual care, the study did not diarrhea, constipation, appetite loss, demonstrate any significant Study Duration - fatigue, and sleep disturbances changes in any QOL Two years single items measures, including physical, Risk of Bias - 3 Functional Status - EORTC QLQ- role, emotional, cognitive, C30 physical functioning scale social, or global health functioning; or fatigue, nausea

and vomiting, pain, dyspnea, Caregiver Outcomes: anorexia, diarrhea, None described constipation, or sleep disturbances.

100

Evidence Table 1: Populations and Interventions

McDonald et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 84 al., 2005 US Criteria: Cancer (mixed cancer) Constant pain: control, 13.2%; Control Arm - Home Usual care Outcomes Study Arms - 3 Patients were eligible for basic intervention, 16.5%; Intervention Arm 1 - Home Comparing a home health inclusion if they were at least augmented intervention, nurse-targeted email Level of Prognosis: Intervention Arm 2 - Home Intervention Arm 1: Randomization - 18 years old, admitted with a 10.7%. intervention that included Providers primary cancer diagnosis, and Not described Pain intensity (0-10 scale, Email reminder to evidence-based reminders for had self-reported daily or pain management (basic lower is better): control, mean Multidisciplinary and Clinical home health nurses Study Duration - 45 constant pain at admission. intervention), an intervention days 5.3 (SD 2.2); basic Disciplines: They were excluded if they intervention, mean 5.4 (SD that included these reminders Control Arm - No multidisciplinary Intervention Arm 2: Risk of Bias - 3 were not cognitively able to 2.1); augmented intervention, as well as cards for better pain team; nurses give informed consent and did mean 5.4 (SD 2.2). Email reminder to assessment, nurse physician not speak English or Spanish. Intervention Arm 1 - No home health nurses, communication, and a self- multidisciplinary team; nurses. which included care guide to help patients Patient Functional Status: open a dialogue on pain Intervention Arm 2 - Home health provider prompts, Number of Patient OASIS combined ADLADL management (augmented nurses, oncology clinical nurse patient education Participants and Decedents: score (scale not described, intervention), and usual care specialist; advanced practice materials, and clinical Screened - 1729 lower is better): control, mean nurse specialist provided by home health nurses, nurses. nurses, the basic intervention Enrolled - 673 5.5 (SD 2.7); basic outreach intervention, mean 5.5 (SD demonstrated a significant

Completed - 673 2.7); augmented intervention, improvement in average pain, Decedents - Not provided mean 5.5 (SD 2.5). Intervention while the augmented Duration: intervention demonstrated

significant improvements in 45 days Caregiver Symptoms: worst pain and use of None described alternative pain treatments. The study did not demonstrate significant changes in insomnia, constipation, or overall HRQOL.

Health Care Use The study did not demonstrate a significant change in hospitalization rates.

Health Care Costs The study did not demonstrate significant changes in overall costs or home care-related costs.

101

Evidence Table 2: Outcomes

McDonald et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient None Patient QOL-Relevant 84 al., 2005 US Pain - BPI (whole scale and pain hospital, ED Outcomes Study Arms - 3 interference scale), EORTC QLQ- Pain: Health Care Costs - Overall Comparing a home health C30 pain scale, use of alternative costs, home care-related costs nurse-targeted email Level of Worst pain (0-10 scale, lower is better): Randomization - pain treatments intervention that included control, mean 4.5; augmented evidence-based reminders for Providers Other Symptoms - EORTC QLQ-C30 intervention, mean 3.3; p=.05. insomnia and constipation scales pain management (basic Study Duration - 45 Average pain (0-10 scale, lower is intervention), an intervention days HRQOL - EORTC QLQ-C30 QOL better): control, mean 3.7; basic that included these reminders Risk of Bias - 3 scale intervention, mean 2.2; p=.03. as well as cards for better pain assessment, nurse physician Use of alternative pain treatments: communication, and a self- Caregiver Outcomes: control, 26.9%; augmented care guide to help patients intervention, 15.9%; p=.02. None described open a dialogue on pain management (augmented Caregiver Outcomes: intervention), and usual care provided by home health N/A nurses, the basic intervention demonstrated a significant improvement in average pain, while the augmented intervention demonstrated significant improvements in worst pain and use of alternative pain treatments. The study did not demonstrate significant changes in insomnia, constipation, or overall HRQOL.

Health Care Use The study did not demonstrate a significant change in hospitalization rates.

Health Care Costs The study did not demonstrate significant changes in overall costs or home care-related costs.

102

Evidence Table 1: Populations and Interventions

Casarett et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 116 al., 2005 US Criteria: Cancer (type not None described Control Arm - Nursing home Usual care Outcomes Study Arms - 2 Patients were eligible for described), CHF, COPD, Intervention Arm - Nursing home Comparing a structured inclusion if they were nursing dementia interview that identified Level of Patient Functional Status: Intervention Arm: Randomization - home residents. They were patients who would be Patients and excluded if they were admitted ADL dependencies (1-5 scale, Multidisciplinary and Clinical Other care planning appropriate for hospice care to a nursing home for a respite Prognosis: with usual efforts to identify families/caregivers lower is better): control, mean Disciplines: stay, were already receiving such patients, the study only Not described 4.5; intervention, mean 4.4; Control Arm - No multidisciplinary hospice care, or were too Intervention demonstrated a significant p=.06. team; physicians. Study Duration - cognitively impaired to Duration: improvement in the Six months complete the interview and did Intervention Arm - No One-time intervention intervention group in families’ Risk of Bias - 6 not have a surrogate. Caregiver Symptoms: multidisciplinary team; physicians. overall assessment of quality None described of care provided in the last month of life. Number of Patients Participants and Decedents: Screened - 400 Health Care Use Enrolled - 205 The study demonstrated a significant reduction in the Completed - 205 intervention group in hospital Decedents - 0 days and acute care admissions.

Hospice Referral or Use The study demonstrated significant improvements in the intervention group in referrals to hospice within 30 days of the interview and mean days of hospice care. The study did not demonstrate a significant change in hospice enrollment at time of death.

Site of Death The study did not demonstrate a significant change in the likelihood of death in a nursing home.

103

Evidence Table 2: Outcomes

Casarett et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Caregiver QOL-Relevant 116 al., 2005 US None described N/A hospital Hospital days: control, mean 3.0 Outcomes Study Arms - 2 Hospice Referral or Use - (range 0-29); intervention, mean Comparing a structured Referral to hospice within 30 1.2 (range 0-18); p=.03. interview that identified Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - days; days of hospice care Acute care admissions: control, patients who would be Patients and Experience or Satisfaction - Toolkit Site of Death - Death in nursing mean 0.49 (range 0-4); appropriate for hospice care with usual efforts to identify families/caregivers Afterdeath Survey Experience or Satisfaction: home intervention, mean 0.28 (range only 0-4); p=.04. such patients, the study Families’ overall assessment of quality demonstrated a significant Study Duration - of care in last month of life: control improvement in the Six months group, mean 2.2 (range 1-5); Hospice Referral or Use: intervention group in families’ intervention group, mean 4.3 (range 2- overall assessment of quality Risk of Bias - 6 Referral to hospice within 30 5); p=.01. of care provided in the last days: control, 1%; intervention, month of life. 20%; p<.001.

Days of hospice care: control, mean 14; intervention, mean 64; Health Care Use p=.01. The study demonstrated a significant reduction in the intervention group in hospital days and acute care admissions.

Site of Death The study did not demonstrate a significant change in the likelihood of death in a nursing home.

104

Evidence Table 1: Populations and Interventions

Engelhardt Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant et al., US Criteria: Cancer (mixed cancer), None described Control Arm - Outpatient clinics Usual care Outcomes 2006104 Study Arms - 2 Patients were eligible for CHF, COPD Intervention Arm - Outpatient clinics Comparing a care Level of inclusion if they had COPD or coordination and support CHF with at least one ICU Patient Functional Status: Intervention Arm: program that included Randomization - Prognosis: Patients and admission or at least two Not described Multidisciplinary and Clinical Palliative case attention to provider support, acute care admissions in the health literacy, and prevention, families/caregivers Not described Disciplines: management, last six months. They were with usual care, the study only Control Arm - No multidisciplinary advance directive also eligible for inclusion if Caregiver Symptoms: demonstrated significant team; clinical disciplines not completion, other Study Duration - they had cancer of the improvements in the None described specified. care planning Six months esophagus, trachea, colon, intervention group in the time

Risk of Bias - 2 liver, pancreas, lung, or Intervention Arm - No to completion of first advance uterus; cancers of the prostate multidisciplinary team; clinical Palliative case directive, completion of at or breast with metastasis; and disciplines not specified. management least one advance directive, melanoma, leukemia, lymph elements: number of advance directives sarcoma, Hodgkin’s disease, AICCP, delivered by completed, and patient or multiple myeloma. allied health satisfaction. The study did not personnel demonstrate a significant change in patient attitudes Number of Patient about participation in Participants and Decedents: Intervention treatment planning. Screened - Not provided Duration: Enrolled - 275 Not described Caregiver QOL-Relevant Completed - 186 Outcomes Decedents - Not provided The study demonstrated a significant improvement in the

intervention group in problems Number of Caregiver reported by surrogates with Participants and Decedents: providers’ emotional and Screened - Not provided spiritual support. Enrolled - 168 Completed - 143 Health Care Costs The study did not demonstrate significant changes in total cost per patient from six months pre-baseline to six months post-baseline.

105

Evidence Table 2: Outcomes

Engelhardt Study Location - Patient Outcomes: Patient Outcomes: Health Care Costs - Total cost None Patient QOL-Relevant et al., US Communication or Care Planning - per patient from six months pre- Outcomes 2006104 baseline to six months post- Study Arms - 2 Time to completion of first advance Communication or Care Planning: Comparing a care directive, completion of at least one baseline coordination and support Level of Time to completion of first advance Randomization - advance directive, number of program that included advance directives completed directive: control, median 238 days; attention to provider support, Patients and intervention, median 46 days; p=.02. families/caregivers Experience or Satisfaction - Study- health literacy, and prevention, only specific patient satisfaction scale, Completion of at least one advance with usual care, the study study-specific attitudes about directive: three months, control, 43.2%; demonstrated significant Study Duration - intervention, 64.7%; p=.005; six improvements in the Six months participation in treatment planning item months, control, 48.4%; intervention, intervention group in the time Risk of Bias - 2 69.4%; p=.006. to completion of first advance directive, completion of at Number of advance directives least one advance directive, Caregiver Outcomes: completed: three months, control, number of advance directives Experience or Satisfaction - Modified mean 0.82 (SD 1.06); intervention, completed, and patient EOL Family Interview mean 1.22 (SD 1.00); p=.01; six satisfaction. The study did not months, control, mean 0.93 (SD 1.07); demonstrate a significant intervention, mean 1.33 (SD 1.07); change in patient attitudes p=.01. about participation in treatment planning. Experience or Satisfaction: Patient satisfaction (1-5 scale, higher is Caregiver QOL-Relevant better): control, pretest mean 3.83 (SD Outcomes 0.76), posttest mean 3.98 (SD 0.67); The study demonstrated a intervention, pretest mean 3.70 (SD significant improvement in the 0.74), posttest mean 4.07 (SD 0.68); intervention group in problems effect size 0.18, p=.03. reported by surrogates with providers’ emotional and Caregiver Outcomes: spiritual support.

Experience or Satisfaction: Health Care Costs Problems reported by surrogates with The study did not demonstrate providers’ emotional and spiritual significant changes in total support (scale not described, lower is cost per patient from six better): control, posttest mean 0.53 months pre-baseline to six (SD 0.32); intervention, posttest mean months post-baseline. 0.41 (SD 0.30); effect size 0.39, p=.03.

106

Evidence Table 1: Populations and Interventions

Goldberg et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 128 al., 2003 US Criteria: CHF, hypertension, None described Control Arm - Not specified Usual care Outcomes Study Arms - 2 Patients were eligible for myocardial infarction, Intervention Arm - Home Comparing a technology- inclusion if they had NYHA diabetes based heart failure monitoring Level of Patient Functional Status: Intervention Arm: Randomization - class III or IV heart failure with system with usual care, the Patients a left ventricular ejection SF-12 PCS (0-100 scale, Multidisciplinary and Clinical Technology-based study did not demonstrate any fraction of less than or equal significant changes in overall Prognosis: higher is better): control, mean Disciplines: heart failure Study Duration - to 35% as measured within six HRQOL, overall physical or 32.1 (SD 8.7); intervention, Control Arm - No multidisciplinary monitoring system 180 days months of enrollment, were Not described mental health, or satisfaction mean 31.7 (SD 7.3); p=.64. team; clinical disciplines not being treated with a diuretic with care. Risk of Bias - 7 NYHA class (I-IV scale, lower specified. and vasodilator, weighed less Intervention is better): III: control, 75.2%; than 400 pounds, had the Intervention Arm - No Duration: intervention, 75.8%; IV: ability to stand for at least 20 multidisciplinary team; nurses. Site of Death control, 24.8%; intervention, 180 days seconds without holding the 24.2%; p=.86. The study did not demonstrate wall, and could speak English a significant change in the or Spanish. They were likelihood of dying at home excluded if they had no phone Caregiver Symptoms: versus in the hospital. line in their home, had an

anticipated life expectancy of None described less than six months, or had a Health Care Use heart transplant, unstable The study did not demonstrate coronary syndromes (unstable significant changes in all- angina, angina-limited cause or heart failure hospital exercise, or myocardial readmission rates or in ED infarction within the eight visits. weeks before enrollment), primary valvular heart disease of either a native or prosthetic valve, primary myocardial disease (obstructive cardiomyopathy, amyloidosis, or active myocarditis), pericardial disease, uncorrected thyroid disease, advanced renal disease (dialysis or creatinine greater than 4.0 mg/dL), or requirement for chronic inotropic therapy.

Number of Patient Participants and Decedents: Screened - Not provided Enrolled - 280 Completed - 241 Decedents - 37

107

Evidence Table 2: Outcomes

Goldberg et Study Location - Patient Outcomes: None Health Care Use - Inpatient None Patient QOL-Relevant 128 al., 2003 US Experience or Satisfaction - Study- hospital, ICU, ED Outcomes Study Arms - 2 specific patient satisfaction item Site of Death - Home death vs. Comparing a technology- Level of Functional Status - SF-12 PCS hospital death based heart failure monitoring system with usual care, the Randomization - HRQOL - MLHFQ Patients study did not demonstrate any Other QOL-Relevant Outcomes - significant changes in overall Study Duration - MOS health distress scale, SF-12 HRQOL, overall physical or 180 days MCS mental health, or satisfaction with care. Risk of Bias - 7

Caregiver Outcomes: Site of Death None described The study did not demonstrate a significant change in the likelihood of dying at home versus in the hospital.

Health Care Use The study did not demonstrate significant changes in all- cause or heart failure hospital readmission rates or in ED visits.

108

Evidence Table 1: Populations and Interventions

Hanks et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 49 2002 EU Criteria: Cancer (mixed cancer), Pain: control, 57%; Control Arm - Hospital Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for non-cancer chronic intervention, 53%. Intervention Arm - Outpatient care team Comparing a full palliative Level of inclusion if they were newly disease Dyspnea: control, 15%; clinics, hospital care team with limited referred to the palliative care telephone advice provided by Randomization - intervention, 8%. Palliative care team Patients team. They were excluded if a palliative care team for they were unable to give Prognosis: Weakness: control, 7%; Multidisciplinary and Clinical elements: newly referred patients, the Study Duration - informed consent, were not Not described intervention, 11%. Disciplines: Inpatient consultation study did not demonstrate any Four weeks well enough to undertake the significant changes in Gastrointestinal problems Control Arm - Palliative care team; baseline assessment, were HRQOL, mood, degree of Risk of Bias - 6 (constipation, sickness, physicians, advanced practice, Intervention Arm: not aware of their diagnosis, bother by emotional problems, anorexia, lack of appetite, nurses. were likely to die or be mouth problems): control, Palliative care team or satisfaction with hospital discharged within 24 hours, 10%; intervention, 8%. Intervention Arm - Palliative care care. needed advice very urgently, team; physicians, nurses, social Worry/anxiety/depression expressed a strong preference workers, chaplains, psychologist. Palliative care team anger/frustration: control, 5%; to see the palliative care team elements: Caregiver QOL-Relevant intervention, 6%. or had a referring consultant Inpatient consultation Outcomes who insisted that they be The study did not demonstrate seen, or were in extreme Patient Functional Status: a significant change in distress. Intervention caregiver satisfaction with None described Duration: hospital care. Four weeks Number of Patient Caregiver Symptoms: Participants and Decedents: Health Care Use None described Screened - 684 The study did not demonstrate Enrolled - 261 significant changes in length of hospital stay, hospital Completed - 109 readmission rates, or number Decedents - Not provided of visits from other hospital therapists.

Other Economic Outcomes The study did not demonstrate significant changes in the number of diagnostic images or diagnostic tests.

109

Evidence Table 2: Outcomes

Hanks et al., Study Location - Patient Outcomes: None Health Care Use - Inpatient None Patient QOL-Relevant 49 2002 EU Experience or Satisfaction - hospital, number of visits from Outcomes Study Arms - 2 Satisfaction with hospital care other hospital therapists Comparing a full palliative Level of (Adapted MacAdam’s Assessment of Other Economic Outcomes - care team with limited Randomization - Suffering Questionnaire) Number of diagnostic images, telephone advice provided by Patients HRQOL - EORTC QLQ-C30 global number of diagnostic tests a palliative care team for health status/QOL subscale newly referred patients, the Study Duration - study did not demonstrate any Four weeks Other QOL-Relevant Outcomes - significant changes in Risk of Bias - 6 Mood (Memorial Pain Assessment HRQOL, mood, degree of Card), bothered by emotional bother by emotional problems, problems (WONCA scale) or satisfaction with hospital care. Caregiver Outcomes: Experience or Satisfaction - Caregiver QOL-Relevant Satisfaction with hospital care Outcomes (study-specific satisfaction with care The study did not demonstrate questionnaire, which included the a significant change in FAMCARE and HADS scales) caregiver satisfaction with hospital care.

Health Care Use The study did not demonstrate significant changes in length of hospital stay, hospital readmission rates, or number of visits from other hospital therapists.

Other Economic Outcomes The study did not demonstrate significant changes in the number of diagnostic images or diagnostic tests.

110

Evidence Table 1: Populations and Interventions

Ringdal et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 57 al., 2002 EU Criteria: Cancer (mixed cancer) None described Control Arm - Outpatient clinics, Usual care Outcomes Study Arms - 2 Patients were eligible for hospital Comparing a palliative care inclusion if they were at least program intervention in a Level of Prognosis: Patient Functional Status: Intervention Arm - Outpatient Intervention Arm: Randomization - 18 years old, had incurable, clinics, home, hospital palliative medicine unit with Sites malignant cancer, and had life Identified by clinicians as None described Palliative care team usual care, the study expectancy between 2 and 9 demonstrated a significant poor prognosis Study Duration - months. Multidisciplinary and Clinical improvement in the One month after Caregiver Symptoms: Disciplines: Palliative care team intervention group in overall patient death None described elements: family satisfaction with care. Number of Patient Control Arm - No multidisciplinary Risk of Bias - 0 Inpatient consultation, Participants and Decedents: team; physicians, nurses. outpatient Site of Death Screened - Not provided Intervention Arm - Palliative care consultation, home team; physicians, nurses. Enrolled - 434 care and GP visits The study did not demonstrate a significant change in home

Completed - Not provided deaths as compared with Decedents - 183 Intervention deaths in a hospital or nursing Duration: home.

One month after Number of Caregiver patient death Participants: Screened - 426 Enrolled - 312 Completed - 183

111

Evidence Table 2: Outcomes

Ringdal et Study Location - Patient Outcomes: Patient Outcomes: Site of Death - Home death vs. None Caregiver QOL-Relevant 57 al., 2002 EU None described N/A hospital or nursing home death Outcomes Study Arms - 2 Comparing a palliative care program intervention in a Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - palliative medicine unit with Sites Experience or Satisfaction - Experience or Satisfaction: usual care, the study demonstrated a significant FAMCARE Scale Satisfaction with care (0-100 scale, Study Duration - improvement in the higher is better): control group, mean One month after intervention group in overall 63.1; intervention group, mean 71.7; patient death family satisfaction with care. p<.01. Risk of Bias - 0 Site of Death The study did not demonstrate a significant change in home deaths as compared with deaths in a hospital or nursing home.

112

Evidence Table 1: Populations and Interventions

Aiken et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 53 2006 US Criteria: CHF, COPD Bodily pain (0-100 scale, Control Arm - Not described Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for higher better): all participants, Intervention Arm - Home case management Comparing a comprehensive Level of inclusion if they were at least mean 48.5. case management and 18 years old; had CHF with Prognosis: coordinated care program with Randomization - Palliative case Patients class IIIb or IV heart failure, or Identified by clinicians as Multidisciplinarity and Clinical usual care, the study had COPD with oxygen poor prognosis Patient Functional Status: Disciplines: management demonstrated significant Study Duration - saturation of less than 88% on elements: improvements in the SF-36 physical functioning Control Arm - No multidisciplinary Nine months room air or baseline pO2 less intervention group in physical subscale (0-100 scale, higher team; clinical disciplines not Nurse-led case than 55 on room air and were functioning, general health Risk of Bias - 3 is better): all participants, specified. management on continuous oxygen; had perceptions, vitality (in COPD mean 13.0. fatigue, palpitation, dyspnea, Intervention Arm - Care patients only), communication

or angina after engaging in coordination team; nurses, social Intervention Arm: of information to handle illness any activity; were treated for Caregiver Symptoms: workers, chaplains, medical Palliative case and education about exacerbations of their None described director. management, community resources, most illnesses in an ED, urgent care decision support, troublesome symptom, facility, or hospital within three advance directive symptom distress (in COPD months prior to enrollment; completion patients only), and advance had a telephone in their home; directive completion rates. The and spoke English or had a study also demonstrated a translator in their home. Palliative case significant increase in management symptom distress (CHF elements: patients only). Number of Patient Participants and Decedents: Nurse-led case management Screened - 965 Health Care Use

Enrolled - 190 The study did not demonstrate Decision support a significant change in ED Completed - 66 elements: visits. Decedents - 38 Printed decision aid

Intervention Duration: Nine months

113

Evidence Table 2: Outcomes

Aiken et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - ED None Patient QOL-Relevant 53 2006 US Other Symptoms - MSAS Outcomes Other Symptoms: Study Arms - 2 Communication or Care Planning - Comparing a comprehensive Level of Study-specific received sufficient Most troublesome symptom (1-5 scale, lower is case management and better): three months, control, mean 4.29; Randomization - information to handle illness item, coordinated care program with intervention, mean 3.41; p<.05; six months, usual care, the study Patients study-specific received education control, mean 3.80; intervention, mean 2.85; about community resources item, p<.05. demonstrated significant Study Duration - improvements in the Nine months living will or advance directive Symptom distress, COPD patients only: three completion months, lower in intervention vs. control, p<.05. intervention group in physical Risk of Bias - 3 functioning, general health Functional Status - SF-36 physical Symptom distress, CHF patients only: six months, higher in intervention vs. control, p<.05. perceptions, vitality (in COPD functioning subscale patients only), communication Other QOL-Relevant Outcomes - of information to handle illness Communication or Care Planning: SF-36 general health perceptions and education about and vitality subscales Received sufficient information to handle illness community resources, most (1-4 scale, higher is better): six months, control, troublesome symptom, mean 3.51; intervention, mean 3.90; p<.05. symptom distress (in COPD Caregiver Outcomes: Received education about community resources (1-4 scale, higher is better): three months, patients only), and advance None described control, mean 2.02; intervention, mean 2.57; directive completion rates. The p<.05. study also demonstrated a Rate of living will or advance directive significant increase in completion: univariate: three months, control, symptom distress (CHF 65%; intervention, 71%; p<.05; multivariate: patients only). three months, intervention vs. control, OR 4.4 (95% CI 1.10-18.18). Health Care Use Functional Status: The study did not demonstrate Physical functioning (0-100 scale, higher is a significant change in ED better), COPD patients only: baseline, all participants, mean 13.0; nine months, control, visits. mean -0.95; intervention, mean +1.00; p<.05. Physical functioning (0-100 scale, higher is better), CHF patients only: baseline, all participants, mean 13.0; nine months, control, mean -1.39; intervention, mean +0.18; p<.05.

Other QOL-Relevant Outcomes: General health perceptions (0-100 scale, higher is better), COPD patients only: baseline, all participants, mean 17.4; nine months, control, mean -1.67; intervention, mean +0.54; p<.05. General health perceptions (0-100 scale, higher is better), CHF patients only: baseline, all participants, mean 17.4; nine months, control, mean -0.17; intervention, mean +0.16; p<.05. Vitality (0-100 scale, higher is better), COPD patients only: three months, higher in intervention vs. control, p<.05.

Caregiver Outcomes: N/A

114

Evidence Table 1: Populations and Interventions

Bakitas et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 19 al., 2009 US Criteria: Cancer (mixed cancer) ESAS score (0-900 scale, Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for lower is better): control, mean Intervention Arm - Home Comparing a patient inclusion if they were within 8 286.3 (SD 154.0); psychoeducational and case Level of Prognosis: Intervention Arm: Randomization - to 12 weeks of a new intervention, mean 282.5 (SD management intervention with Patients and diagnosis of life-limiting Identified by clinicians as 148.8); p=.83. Multidisciplinarity and Clinical Palliative case usual care, the study families/caregivers cancer (unresectable stage III poor prognosis Depression (0-60 scale, lower Disciplines: management, demonstrated a significant or IV gastrointestinal, stage improvement in overall only is better): control, mean 13.8 Control Arm - No multidisciplinary advance directive IIIb or IV non-small cell or HRQOL and depressed mood (SD 8.9); intervention, mean team; clinical disciplines not completion Study Duration - 13 extensive small cell lung, or but demonstrated no 12.1 (SD 8.5); p=.11. specified. months stage IV and visceral crisis, significant change in overall Risk of Bias - 5 lung or liver metastases, Palliative case symptom intensity. management estrogen receptor negative, Patient Functional Status: Intervention Arm - No HER2/neu positive breast elements: KPS score (0-100 scale, multidisciplinary team; Advanced cancer). They were excluded Health Care Use higher is better): control, mean practice nurses. Nurse-led case if they scored less than 17 on 77.4 (SD 12.8); intervention, management The study did not demonstrate a modified MMSE, had an significant changes in hospital mean 78.4 (SD 11.1); p=.50. Axis I psychiatric disorder, or days, ICU days, or ED visits. had active substance abuse. Intervention Caregiver Symptoms: Duration: Number of Patient None described 13 months Participants and Decedents: Screened - 1222 Enrolled - 322 Completed - 279 Decedents - 47

115

Evidence Table 2: Outcomes

Bakitas et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient None Patient QOL-Relevant 19 al., 2009 US Depression - CES-D hospital, ICU, ED Outcomes Study Arms - 2 Other Symptoms - Overall symptom Depression: Comparing a patient psychoeducational and case Level of intensity (ESAS) Depressed mood (0-60 scale, lower is Randomization - management intervention with HRQOL - FACIT-Pal better): 13 months vs. baseline, usual care, the study Patients and intervention vs. control, mean -1.8 (SE families/caregivers demonstrated a significant 0.81), p=.02. improvement in overall only Caregiver Outcomes: HRQOL and depressed mood Study Duration - 13 None described but demonstrated no HRQOL: months significant change in overall Risk of Bias - 5 Overall QOL (0-184 scale, higher is symptom intensity. better): 13 months vs. baseline,

intervention vs. control, mean +4.6 (SE 2.0), p=.02. Health Care Use Overall QOL, patients who died during The study did not demonstrate the study only (0-184 scale, higher is significant changes in hospital better): 13 months vs. baseline, days, ICU days, or ED visits. intervention vs. control, mean +8.6 (SE 3.6), p=.02.

Caregiver Outcomes: N/A

116

Evidence Table 1: Populations and Interventions

Brumley et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 17 al., 2007 US Criteria: Cancer (mixed cancer), None described Control Arm - Outpatient clinics, Usual care Outcomes Study Arms - 2 Patients were eligible for CHF, COPD home, hospital Comparing an in-home Level of inclusion if they had a primary Intervention Arm - Outpatient palliative care intervention diagnosis of CHF, COPD, or Patient Functional Status: Intervention Arm: provided by a multidisciplinary Randomization - Prognosis: clinics, home, hospital Patients cancer; had a life expectancy PPS score (0-100 scale, Palliative care team team with usual care, the one year or less; had at least study demonstrated a Identified by clinicians as higher is better): control, mean Study Duration - one hospital or ED visit in the poor prognosis 58.5 (SD 12.0); intervention, Multidisciplinarity and Clinical significant improvement in the 120 days past year; scored 70% or less mean 57.8 (SD 13.1); p>.05. Disciplines: Palliative care team intervention group in on the PPS; and had a elements: satisfaction with care. Risk of Bias - 6 Control Arm - No multidisciplinary primary care physician who team; clinical disciplines not Outpatient answered no to the question Caregiver Symptoms: specified. consultation “Would you be surprised if this Health Care Use None described patient died in the next year?” Intervention Arm - Palliative care The study demonstrated team; physicians, nurses, social Intervention significant decreases in the workers, chaplains, bereavement Duration: intervention group in number Number of Patient coordinator, home health aide, 120 days of ED visits, number of Participants and Decedents: pharmacist, dietician, volunteer, hospitalizations, and hospital Screened - 718 physical therapist, occupational days. therapist, speech therapist. Enrolled - 310 Completed - 297 Hospice Referral or Use Decedents - 225 The study did not demonstrate a significant change in hospice enrollment.

Site of Death The study demonstrated a significant increase in the intervention group in the rate of home deaths.

Health Care Costs The study demonstrated significant decreases in the intervention group in overall costs of care per patient and cost per patient per day.

117

Evidence Table 2: Outcomes

Brumley et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 17 al., 2007 US Experience or Satisfaction - Reid- hospital, ED ED visits: intervention vs. Outcomes Study Arms - 2 Gundlach Satisfaction with Services Experience or Satisfaction: Hospice Referral or Use - control, linear regression Comparing an in-home Level of instrument Hospice enrollment coefficient -0.35, p=.02. palliative care intervention Satisfaction with health care services provided by a multidisciplinary Randomization - (0-48 scale, higher is better): 30 days Site of Death - Home death One or more ED visits: control, Patients 33%; intervention, 20%; p=.01. team with usual care, the Caregiver Outcomes: post-enrollment, intervention vs. Health Care Costs - Overall study demonstrated a Study Duration - None described control, OR 3.37 (95% CI 1.42-8.10); costs of care per patient, cost One or more hospitalizations: significant improvement in the 120 days p=.006; 90 days post-enrollment, per patient per day control, 59%; intervention, 36%; intervention group in Risk of Bias - 6 intervention vs. control, OR 3.37 (95% p<.001. satisfaction with care. CI 0.65-4.96); p=.03. Hospital days: intervention vs. control, linear regression Health Care Use Caregiver Outcomes: coefficient -4.36, p<.001. The study demonstrated N/A significant decreases in the Site of Death: intervention group in number Home death: control, 51%; of ED visits, number of intervention, 71%; p<.001. hospitalizations, and hospital days.

Hospice Referral or Use The study did not demonstrate a significant change in hospice enrollment.

Site of Death The study demonstrated a significant increase in the intervention group in the rate of home deaths.

Health Care Costs The study demonstrated significant decreases in the intervention group in overall costs of care per patient and cost per patient per day.

118

Evidence Table 1: Populations and Interventions

Chapman et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 100 al., 2007 US Criteria: Dementia Pain (scale not described, Control Arm - Nursing home Usual care Outcomes Study Arms - 2 Patients were eligible for lower is better): control, mean Intervention Arm - Nursing home Comparing a care planning inclusion if they had 0.38 (SD 0.29); intervention, intervention administered by Level of Prognosis: Intervention Arm: Randomization - Alzheimer’s disease or other mean 0.31 (SD 0.29). advanced care teams with Patients dementia, had four or more Not described Pain (scale not described, Multidisciplinarity and Clinical Other care planning usual care for nursing home ADL dependencies, scored 23 residents, the study lower is better): control, mean Disciplines: Study Duration - 10 or less on the MMSE, and demonstrated a significant 1.76 (SD 1.38); intervention, Control Arm - Nursing home care weeks scored four or more on the Intervention improvement in the mean 1.64 (SD 1.46). team; nurses, social workers, Global Deterioration Scale. Duration: intervention group in Risk of Bias - 4 chaplains, occupational and They were excluded if they Depression (scale not physically nonaggressive physical therapists. Eight weeks had serious medical described, lower is better): behavior. The study did not conditions requiring immediate control, mean 0.08 (SD 0.08); Intervention Arm - Advanced illness demonstrate significant treatment (e.g., aspiration intervention, mean 0.11 (SD care team; physicians, nurses, changes in pain, depression, pneumonia, difficulty 0.14). social workers, team members from aggressive behavior, or swallowing, dehydration, or the disciplines of psychology, verbally agitated behavior. decubitus ulcers). physical and occupational therapy, Patient Functional Status: and nutrition.

All participants had four or Number of Patient more ADL dependencies. Participants and Decedents:

Screened - Not described Caregiver Symptoms: Enrolled - 118 None described Completed - 118 Decedents - Not described

119

Evidence Table 2: Outcomes

Chapman et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 100 al., 2007 US Pain - FLACC, PAINAD Outcomes Study Arms - 2 Depression - CSDD Anxiety: Comparing a care planning Level of intervention administered by Anxiety - CMAI physically Physically nonaggressive behavior: advanced care teams with Randomization - nonaggressive behavior, aggressive control, baseline, mean 1.36 (SD 0.52); Patients usual care for nursing home behavior, and verbally agitated eight weeks, mean 1.29 (SD 0.49); residents, the study Study Duration - 10 behavior subscales intervention, baseline, mean 1.64 (SD demonstrated a significant weeks 1.10); eight weeks, mean 1.30 (SD improvement in the 0.60); group by time p<.05. intervention group in Risk of Bias - 4 Caregiver Outcomes: physically nonaggressive None described behavior. The study did not Caregiver Outcomes: demonstrate significant N/A changes in pain, depression, aggressive behavior, or verbally agitated behavior.

120

Evidence Table 1: Populations and Interventions

Chochinov et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 46 al., 2011 Canada Criteria: Cancer (mixed cancer) Physically distressing Control Arm - Home, hospital, Usual care, palliative Outcomes Study Arms - 3 Patients were eligible for symptoms (0-5 scale, lower is residential care facilities care team Comparing standard palliative Level of inclusion if they were at least better): standard palliative Intervention Arm 1 - Home, care plus dignity therapy, 18 years old, had a terminal Prognosis: care, mean 2.28 (SD 1.21); standard palliative care plus Randomization - hospital, residential care facilities Palliative care team Patients prognosis with a life Identified by clinicians as standard palliative care plus client-centered care, and expectancy of six months or poor prognosis, Referred client-centered care, mean Intervention Arm 2 - Home, elements: standard palliative care alone, Study Duration - 10 less according to their treating to or receiving palliative 2.57 (SD 1.33); standard hospital, residential care facilities Inpatient consultation, the study did not demonstrate days physician, were receiving care palliative care plus dignity outpatient any significant changes in palliative care in a hospital or therapy, 2.32 (SD 1.29). HRQOL, spiritual well-being, Risk of Bias - 2 Multidisciplinarity and Clinical consultation community setting (hospice or patient assessment of Anxiety (0-5 scale, lower is Disciplines: home), were willing to commit better): standard palliative common symptoms and end- Control Arm - Palliative care team; Intervention Arm 1: 3-4 contacts over the study care, mean 1.73 (SD 1.01); of-life concerns, pain, physicians, nurses, social workers, period, and were able and standard palliative care plus Palliative care team, dyspnea, depression, anxiety, chaplains, psychologists, willing to provide written client-centered care, mean client-centered care nausea, drowsiness, or informed consent. They were psychiatrists. anorexia. 1.93 (SD 1.11); standard excluded if they were delirious palliative care plus dignity Intervention Arm 1 - Palliative care or cognitively impaired, too ill therapy, 1.93 (SD 1.09). team; physicians, nurses, social Palliative care team to complete the requirements workers, chaplains, psychologists, elements: of the study protocol, or Depression (0-5 scale, lower psychiatrists. Inpatient consultation, unable to read and speak is better): standard palliative outpatient English. care, mean 1.69 (SD 0.95); Intervention Arm 2 - Palliative care standard palliative care plus team; physicians, nurses, social consultation

client-centered care, mean workers, chaplains, psychologists, Number of Patient 1.76 (SD 1.00); standard psychiatrists. Intervention Arm 2: Participants and Decedents: palliative care plus dignity Palliative care team, Screened - 1513 therapy, 1.84 (SD 1.12). dignity therapy Enrolled - 441 Depression (0-6 scale, lower is better): standard palliative Completed - 326 care, mean 1.14 (SD 1.31); Palliative care team Decedents - 213 standard palliative care plus elements: client-centered care, mean 1.25 (SD 1.19); standard Inpatient consultation, palliative care plus dignity outpatient therapy, mean 1.22 (SD 1.26). consultation Pain (1-10 scale, lower is better): standard palliative Intervention care Duration: 10 days Patient Functional Status: PPS score (0-100 scale, higher is better): all participants, mean 44.3 (SD 15.1).

Caregiver Symptoms: None described

121

Evidence Table 2: Outcomes

Chochinov et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 46 al., 2011 Canada Pain - ESAS pain item Outcomes Study Arms - 3 Dyspnea - ESAS dyspnea item Comparing standard palliative Level of care plus dignity therapy, Depression - HADS depression standard palliative care plus Randomization - subscale Patients client-centered care, and Anxiety - HADS anxiety subscale standard palliative care alone, Study Duration - 10 the study did not demonstrate Other Symptoms - ESAS nausea, days any significant changes in drowsiness, and anorexia items Risk of Bias - 2 HRQOL, spiritual well-being, Existential or Spiritual Well-being - patient assessment of FACIT-Sp common symptoms and end- HRQOL - QOLS of-life concerns, pain, dyspnea, depression, anxiety, Other QOL-Relevant Outcomes - nausea, drowsiness, or Common symptoms and end-of-life anorexia. concerns (SISC)

Caregiver Outcomes: None described

122

Evidence Table 1: Populations and Interventions

Clayton et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 105 al., 2007 Australia Criteria: Cancer None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient Comparing a question prompt Level of inclusion if they were at least clinics, home list given to patients during 18 years old, spoke English, Prognosis: Patient Functional Status: Intervention Arm: palliative consultation with Randomization - Patients had a diagnosis of advanced Referred to or receiving None described Question prompt list usual care, the study progressive life-limiting demonstrated significant palliative care Multidisciplinarity and Clinical given to patients Study Duration - cancer, and were able and Disciplines: during palliative improvements in the Immediately after well enough to read question Caregiver Symptoms: intervention group in the Control Arm - No multidisciplinary consultation intervention prompt list and complete number of questions asked None described team; clinical disciplines not Risk of Bias - 7 questionnaires. during consultation and in specified. Intervention discussions of what was Intervention Arm - No Duration: happening with the patient’s Number of Patient multidisciplinary team; physicians. One-time intervention illness and what to expect in Participants and Decedents: the future. The study did not Screened - 196 demonstrate significant changes in most other Enrolled - 174 measures of communication Completed - 170 or in patient anxiety or Decedents - Not described satisfaction.

123

Evidence Table 2: Outcomes

Clayton et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 105 al., 2007 Australia Anxiety - STAI Outcomes Study Arms - 2 Communication or Care Planning - Communication or Care Planning: Comparing a question prompt Level of Single item from the Cassileth list given to patients during Unmet need regarding discussion of palliative consultation with Randomization - Information Styles Questionnaire “what to expect in the future”: control, Patients measuring amount of detail usual care, the study 28%; 24 hours, 14%; p=.04. demonstrated significant Study Duration - preferred; single item from the Cassileth Information Styles Discussed “what is happening with my improvements in the Immediately after illness”: control, 84%; three weeks, intervention group in the intervention Questionnaire measuring amount of detail preferred pertaining to each of 96%; p=.05. number of questions asked during consultation and in Risk of Bias - 7 (1) patients’ perception regarding Number of questions asked: discussions of what was whether the topics were discussed, intervention vs. control, OR 2.31 (95% happening with the patient’s (2) unmet information need, or (3) CI 1.68-3.18); p<.001. too much information; number of illness and what to expect in

questions asked the future. The study did not Caregiver Outcomes: demonstrate significant Experience or Satisfaction - changes in most other Undescribed patient satisfaction N/A measures of communication scale or in patient anxiety or satisfaction. Caregiver Outcomes: None described

124

Evidence Table 1: Populations and Interventions

Jawahri et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 106 al., 2010 US Criteria: Cancer (malignant None described Control Arm - Outpatient clinics Decision support Outcomes Study Arms - 2 Patients were eligible for glioma) Intervention Arm - Outpatient Comparing a goals-of-care Level of inclusion if they could clinics, home video and verbal narrative with communicate in English, could Patient Functional Status: Decision support the verbal narrative alone, the Randomization - Prognosis: elements: Patients provide informed consent, and None described study demonstrated significant had an MMSE score of at improvements in the Identified by clinicians as Multidisciplinarity and Clinical Verbal narrative Study Duration - least 24. intervention group in poor prognosis Disciplines: Immediately after Caregiver Symptoms: knowledge of goals-of-care intervention Control Arm - No multidisciplinary None described Intervention Arm: options and uncertainty in Number of Patient team; physicians. Risk of Bias - 6 Decision support end-of-life decision-making, a Participants and Decedents: Intervention Arm - No significant increase in

Screened - 73 multidisciplinary team; physicians. preference for comfort care Decision support over life-prolonging and basic Enrolled - 50 elements: medical care, and a significant Completed - 50 decrease in preference for Video-based decision CPR. Decedents - Not described aid, verbal narrative

Intervention Duration: One-time intervention

125

Evidence Table 2: Outcomes

Jawahri et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 106 al., 2010 US Communication or Care Planning - Outcomes Study Arms - 2 Study-specific knowledge of goals- Communication or Care Planning: Comparing a goals-of-care Level of of-care options scale, Decisional video and verbal narrative with Conflict Scale uncertainty subset, Increase in knowledge of goals-of-care the verbal narrative alone, the Randomization - options (0-6 scale, higher is better): Patients study-specific preference for types of study demonstrated significant care measure control, mean 0.9; intervention, mean improvements in the Study Duration - 1.9; p=.004. Other QOL-Relevant Outcomes - intervention group in Immediately after Uncertainty in end-of-life decision- knowledge of goals-of-care intervention Study-specific preference for CPR measure making (0-15 scale, higher is better): options and uncertainty in Risk of Bias - 6 control, mean 13.5; intervention, mean end-of-life decision-making, a 13.7; p=.002. significant increase in preference for comfort care Caregiver Outcomes: Preference for types of care: post- over life-prolonging and basic None described intervention, control, life-prolonging medical care, and a significant care, 25.9%; basic medical care, decrease in preference for 51.9%; comfort care, 22.2%; uncertain, CPR. 0%; intervention, life-prolonging care, 0%; basic medical care, 4.4%; comfort care, 91.3%; uncertain, 4.4%; p<.001.

Other QOL-Relevant Outcomes: Preference for CPR: control, baseline, 29.6%; post-intervention, 40.7%; intervention, baseline, 34.8%; post- intervention, 8.7%; change in intervention group from baseline to post-intervention, p=.04; intervention vs. control groups at post-intervention, p=.02.

Caregiver Outcomes: N/A

126

Evidence Table 1: Populations and Interventions

Engelhardt Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant Outcomes 55 Criteria: et al., 2009 US Cancer (mixed cancer), None described Control Arm - Not described Usual care Comparing a care coordination, Patients were eligible for education, and health counseling Study Arms - 2 CHF, COPD, ESRD intervention with usual care, the study inclusion if they had solid-tumor Intervention Arm - Not described Level of demonstrated significant cancers with metastasis; had Patient Functional Status: Intervention Arm: improvements in the intervention Randomization - CHF with persistent NYHA class Prognosis: group in symptom communication and Patients and III or IV symptoms and an SF-12 physical functioning Multidisciplinarity and Clinical Other care planning, care for physical and emotional families/caregivers ejection fraction of 30% or lower, Identified by clinicians as subscale (0-100 scale, higher Disciplines: health counseling, discomfort, help in coordinating needs for spiritual support, support from only severe aortic stenosis or mitral poor prognosis is better): control, mean 33.45 Control Arm - No multidisciplinary education (SD 9.49); intervention, mean providers in understanding and coping Study Duration - regurgitation, or neither of these; team; clinical disciplines not with illness, support from providers in had end-stage pulmonary 33.26 (SD 11.06). managing family decision-making and Nine months specified. Intervention disease, as defined by (1) problem resolution, acquisition of Risk of Bias - 5 severe chronic lung disease with Intervention Arm 1 - No Duration: information about and implementation of advance planning, completion of disabling dyspnea at rest or Caregiver Symptoms: multidisciplinary team; social Four months during minimal exertion, poor or workers, health educator. advance directives, and agreement to None described DNR/DNI, as well as a significant no response to bronchodilators, decrease in the time to completion of or decreased functional advance directives. The study did not capacity, or (2) hypoxemia or demonstrate significant changes in hypercapnia on room air or physical functioning, mental health, continuous oxygen; or were on anxiety about death, spiritual well- dialysis or had a being, or family communication and support. recommendation to begin dialysis within three months, and also had at least one Caregiver QOL-Relevant Outcomes hospitalization or ED visit within The study demonstrated a significant the past year for complications improvement in caregivers’ of renal disease. They were assessment that the intervention excluded if they did not speak attended to the emotional and spiritual English, were cognitively needs of the family. The study did not impaired, resided in intensive demonstrate significant changes in caregivers’ assessment that the care or skilled nursing settings, intervention encouraged advance care or were enrolled in an end-of-life planning, informed and promoted care service. shared decision-making, or provided physical comfort and emotional support. Number of Patient Participants and Decedents: Health Care Use Screened - Not described The study demonstrated a significant Enrolled - 532 decrease in inpatient admissions. The study did not demonstrate significant Completed - 403 changes in ED visits, outpatient visits, home health visits, radiology tests, or Decedents - 88 laboratory tests.

Number of Caregiver Drugs Participants: The study did not demonstrate a Screened - Not described significant change in pharmacy prescriptions. Enrolled - Not described Completed - 185

127

Evidence Table 2: Outcomes

Engelhardt Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant Outcomes 55 et al., 2009 US Anxiety - Revised Death Anxiety hospital, ED, outpatient visits, Inpatient admissions: control, Comparing a care coordination, education, and health counseling Study Arms - 2 Scale Existential or Spiritual Well-Being: home health care, radiology baseline, mean 2.19 (SD 5.95); tests, laboratory tests intervention with usual care, the study Level of Existential or Spiritual Well-Being - Help in coordinating needs for spiritual support post-intervention, mean 4.33 demonstrated significant (scale not described, higher is better): control, improvements in the intervention Randomization - FACIT-Sp, AICCP Questionnaire Drugs - Pharmacy prescriptions (SD 16.26); intervention, baseline, mean 1.79 (SD 1.22); post- baseline, mean 1.97 (SD 3.35); group in symptom communication and Patients and Communication or Care Planning - intervention, mean 1.93 (SD 1.38); intervention, care for physical and emotional baseline, mean 1.91 (SD 1.25); post- post-intervention, mean 2.44 discomfort, help in coordinating needs families/caregivers AICCP Questionnaire; study-specific only intervention, mean 2.48 (SD 1.49); p<.001. (SD 5.11); p=.045. for spiritual support, support from completion of advance directive, providers in understanding and coping

Study Duration - proxy, or living will measure; study- with illness, support from providers in Nine months specific agreement to DNR/DNI Communication or Care Planning: managing family decision-making and Support from providers in understanding and problem resolution, acquisition of Risk of Bias - 5 measure; study-specific time to information about and implementation completion of advance directive coping with illness (scale not described, higher is better): control, baseline, mean 3.95 (SD 0.87); of advance planning, completion of measure post-intervention, mean 4.00 (SD 0.94); advance directives, and agreement to DNR/DNI, as well as a significant Functional Status - SF-12 physical intervention, baseline, mean 3.98 (SD 0.77); post-intervention, mean 4.19 (SD 0.83); p=.01. decrease in the time to completion of functioning subscale advance directives. The study did not Support from providers in managing family demonstrate significant changes in HRQOL - McGill QOL Questionnaire decision-making and problem resolution (scale physical functioning, mental health, not described, higher is better): control, baseline, anxiety about death, spiritual well- Other QOL-Relevant Outcomes - mean 2.65 (SD 1.37); post-intervention, mean being, or family communication and SF-12 mental health subscale 2.80 (SD 1.36); intervention, baseline, mean support. 2.35 (SD 1.17); post-intervention, mean 3.05 (SD 1.42); p=.002. Caregiver Outcomes: Symptom communication and care for physical Caregiver QOL-Relevant Outcomes and emotional discomfort (0-6 scale, higher is The study demonstrated a significant Experience or Satisfaction - Revised better): control, baseline, mean 3.60 (SD 0.88); improvement in caregivers’ Surrogate Integrated Care Systems post-intervention, mean 3.58 (SD 1.02); assessment that the intervention After Death Interview from the intervention, baseline, mean 3.45 (SD 0.91); attended to the emotional and spiritual Toolkit of Instruments to Measure post-intervention, mean 3.76 (SD 0.93); p=.02. needs of the family. The study did not Acquisition of more information about and demonstrate significant changes in End-of-Life Care (whole scale and caregivers’ assessment that the domains of “encourage advance implementation of advance planning (scale not described, higher is better): control, baseline, intervention encouraged advance care care planning,” “inform and promote mean 2.59 (SD 1.41); post-intervention, mean planning, informed and promoted shared decision-making,” and 2.76 (SD 1.47); intervention, baseline, mean shared decision-making, or provided 2.67 (SD 1.27); post-intervention, mean 3.50 (SD physical comfort and emotional “physical comfort and emotional support. support”) 1.25); p<.001. Completion of advance directive, proxy, or living will: control, 21.1%; intervention, 47.0%; p<.001. Health Care Use Agreement to DNR/DNI: control, 31.8%; The study demonstrated a significant intervention, 40.2%; p=.04. decrease in inpatient admissions. The Time to completion of advance directive: control, study did not demonstrate significant 21.0% by 237 days; intervention, 21.0% by 73 changes in ED visits, outpatient visits, days; p<.001. home health visits, radiology tests, or laboratory tests.

Caregiver Outcomes: Drugs Experience or Satisfaction: The study did not demonstrate a Intervention attended to emotional and spiritual significant change in pharmacy needs of the family (scale not described, lower is prescriptions. better): control, mean 0.64 (SD 0.28); intervention, mean 0.54 (SD 0.32); difference, p=.04.

128

Evidence Table 1: Populations and Interventions

Gade et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 18 2008 US Criteria: Cancer (type not Symptom severity (composite Control Arm - Hospital Usual care Outcomes Study Arms - 2 Patients were eligible for described), CHF, of pain, fatigue, sleep Intervention Arm - Hospital Comparing an inpatient inclusion if they were at least myocardial infarction, changes, nausea, palliative care team with usual Level of Intervention Arm: Randomization - 18 years old, were other heart disease, constipation, diarrhea, dry care for hospitalized patients, Patients hospitalized with at least one COPD, other pulmonary mouth, change in appetite, Multidisciplinarity and Clinical Palliative care team the study demonstrated a life-limiting diagnosis, and had disease, ESRD, dyspnea) (scale not significant improvement in the Disciplines: Study Duration - an attending physician who dementia, stroke, organ described, lower is better): intervention group in patient Control Arm - No multidisciplinary Six months indicated they “would not be failure control, mean 5.2 (SD 1.8); Palliative care team satisfaction with hospital care team; clinical disciplines not surprised if the patient died intervention, 5.1 (SD 1.8); elements: and providers and a significant Risk of Bias - 6 specified. within 1 year.” They were p=.72. Inpatient consultation increase in the rate of excluded if they had impaired Prognosis: Intervention Arm - Palliative care advance directive completions

cognitive status and no Identified by clinicians as team; physicians, nurses, social at hospital discharge. The surrogate, or if they were poor prognosis Patient Functional Status: workers, chaplains. Intervention study did not demonstrate currently enrolled hospice or ECOG performance status (0- Duration: significant changes in other palliative care studies. symptom burden, emotional 5 scale, lower is better): Six months control, median 2 (IQR 2-3); and spiritual support, or intervention, median 2 (IQR 2- HRQOL. Number of Patient 3); p=.68. Participants and Decedents:

Screened - 1168 Health Care Use Caregiver Symptoms: Enrolled - 517 The study demonstrated a None described significant decrease in the Completed - 512 intervention group in the Decedents - 305 number of ICU admissions. The study did not demonstrate a significant change in the number of hospital readmissions.

Hospice Referral or Use The study demonstrated a significant increase in the intervention group in hospice days. The study did not demonstrate significant changes in time from study enrollment to hospice admission or percent of patients admitted to hospice.

Health Care Costs The study demonstrated significant decreases in the intervention group in total health costs per patient and hospital readmission costs per patient. 129

Evidence Table 2: Outcomes

Gade et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 18 2008 US Other Symptoms - MCOHPQ hospital, ICU Number of ICU admissions: Outcomes Study Arms - 2 Physical Area scale Communication or Care Planning: Hospice Referral or Use - control, 21; intervention, 12; Comparing an inpatient Level of Existential or Spiritual Well-Being - Hospice days, time from study p=.04. palliative care team with usual Completion of advance directive at enrollment to hospice care for hospitalized patients, Randomization - MCOHPQ Spiritual Area scale hospital discharge: control, 77.8%; Patients admission, percent of patients the study demonstrated a Communication or Care Planning - intervention, 91.1%; p<.001. admitted to hospice Hospice Referral or Use: significant improvement in the Study Duration - Study-specific completion of Health Care Costs - Total health Hospice days: control, median intervention group in patient Six months advance directive at hospital 12; intervention, median 24; satisfaction with hospital care discharge measure Experience or Satisfaction: costs per patient, hospital Risk of Bias - 6 readmission costs per patient p=.04. and providers and a significant Satisfaction with hospital care and increase in the rate of Experience or Satisfaction - MCOHPQ Place of Care providers (0-10 scale, higher is better): advance directive completions Environment scale; Doctors, Nurses control, mean 6.4; intervention, mean Health Care Costs: at hospital discharge. The Other Care Providers 6.8; p<.001. Total health costs per patient: study did not demonstrate Communication scale Satisfaction with hospital care and control, mean $21,252; significant changes in symptom burden, emotional HRQOL - MCOHPQ QOL scale providers (0-10 scale, higher is better): intervention, mean $14,486;, control, mean 7.2; intervention, mean p=.001. and spiritual support, or Other QOL-Relevant Outcomes - 8.3; p<.001. HRQOL. MCOHPQ emotional relationship Hospital readmission costs per

area scale patient: control, mean $13,275; Caregiver Outcomes: intervention, mean $6,421; Health Care Use p=.009. The study demonstrated a Caregiver Outcomes: N/A significant decrease in the None described intervention group in the number of ICU admissions. The study did not demonstrate a significant change in the number of hospital readmissions.

Health Care Costs The study demonstrated significant decreases in the intervention group in total health costs per patient and hospital readmission costs per patient. 130

Evidence Table 1: Populations and Interventions

Reid et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 96 2006 Canada Criteria: Not described Depression: control, 20.5%; Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for intervention, 30.8%; p=.065. Intervention Arm - Home Comparing a health promotion inclusion if they were at least home care intervention with Level of Prognosis: Bodily pain (0-100 scale, Intervention Arm: Randomization - 75 years old, required higher is better): control, mean usual home care, the study Patients assistance with personal care, Not described 51.63 (SD 34.63); Multidisciplinarity and Clinical Other care planning demonstrated significant and were newly referred to improvements in the intervention, mean 54.03 (SD Disciplines: Study Duration - and eligible for personal intervention group in 32.70). Control Arm - No multidisciplinary Six months support services through the Intervention depression, overall mental team; nurses, social workers, Community Care Access Duration: health, social support, and Risk of Bias - 3 occupational therapists, Centre in Ontario, Canada. emotional role functioning. Patient Functional Status: physiotherapists, speech language Six months They were excluded if they The study did not demonstrate SF-36 physical functioning therapists. were unable to understand significant changes in general subscale (0-100 scale, higher English, refused to give Intervention Arm - No health perceptions, overall is better): control, mean 19.24 informed consent, or were multidisciplinary team; nurses, physical health, physical role (SD 19.02); intervention, deemed ineligible for nursing social workers, occupational functioning, vitality, social role mean 19.57 (SD 18.96). services. therapists, physiotherapists, speech functioning, bodily pain, or SF-36 PCS (0-100 scale, language therapists. coping style. higher is better): control, mean

Number of Patient 37.45 (SD 17.65); Participants and Decedents: intervention, mean 37.94 (SD Health Care Costs Screened - 577 17.83). The study demonstrated a Enrolled - 288 significant decrease in the intervention group in per Caregiver Symptoms: Completed - 242 person costs of prescription Decedents - 17 None described medications. The study did not demonstrate significant changes in total annual per person direct costs of health services or total costs of all types of health and social services.

131

Evidence Table 2: Outcomes

Reid et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Costs - Total Health Care Costs: Patient QOL-Relevant 96 2006 Canada Pain - SF-36 bodily pain subscale annual per person direct costs of Per person cost of prescription Outcomes Study Arms - 2 health services, total costs of all medications: lower in Comparing a health promotion Depression - CES-D Depression: types of health and social Level of intervention vs. control, p=.017. home care intervention with Functional Status - SF-36 PCS and Depression (0-60 scale, lower is services, per person cost of usual home care, the study Randomization - physical role functioning and better): control, baseline, mean 12.94 prescription medication Patients demonstrated significant physical functioning subscales (SD 8.49); six months, mean 11.77 improvements in the Study Duration - Other QOL-Relevant Outcomes - (SD 10.73); intervention, baseline, intervention group in Six months SF-36 MCS and mental health, mean 15.83 (SD 10.00); six months, depression, overall mental Risk of Bias - 3 emotional role functioning, general mean 11.94 (SD 10.18); p=.022. health, social support, and health perceptions, vitality, and emotional role functioning. The study did not demonstrate social role functioning subscales; Other QOL-Relevant Outcomes: PRQ85 part two; coping style significant changes in general (Coping Questionnaire) Overall mental health (0-100 scale, health perceptions, overall higher is better): control, baseline, physical health, physical role

mean 60.69 (SD 18.68); six months, functioning, vitality, social role Caregiver Outcomes: mean 65.18 (SD 22.05); intervention, functioning, bodily pain, or None described baseline, mean 54.32 (SD 19.45); six coping style. months, mean 65.11 (SD 22.44);

p=.009. Health Care Costs Mental health (0-100 scale, higher is better): control, baseline, mean 74.74 The study demonstrated a (SD 20.62); six months, mean 73.75 significant decrease in the (SD 22.83); intervention, baseline, intervention group in per mean 64.59 (SD 22.28); six months, person costs of prescription mean 71.06 (SD 23.88); p=.005. medications. The study did not demonstrate significant Emotional role functioning (0-100 changes in total annual per scale, higher is better): control, person direct costs of health baseline, mean 83.27 (SD 27.41); six services or total costs of all months, mean 82.86 (SD 30.83); types of health and social intervention, baseline, mean 73.19 (SD services. 31.98); six months, mean 83.26 (SD 28.55); p=.014. Perceived social support (25-175 scale, higher is better): control, baseline, mean 127.55 (SD 21.67); six months, mean 125.20 (SD 22.29); intervention, baseline, mean 121.92 (SD 21.49); six months, mean 124.88 (SD 21.87); p=.009.

Caregiver Outcomes: N/A

132

Evidence Table 1: Populations and Interventions

McMillan et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 101 al., 2007 US Criteria: Cancer (mixed cancer) All participants had at least Control Arm - Home Usual care, hospice Outcomes Study Arms - 3 Patients were eligible for two of the following three Intervention Arm 1 - Home Comparing standard home inclusion if they were at least symptoms: pain, dyspnea, and hospice care plus COPE Level of Prognosis: Intervention Arm 2 - Home Intervention Arm 1: Randomization - 18 years old, had a diagnosis constipation. intervention, standard home Patients and of cancer, were receiving Referred to or receiving Hospice, supportive hospice care plus supportive hospice care, had a minimum visits by nurses, and standard families/caregivers hospice Multidisciplinarity and Clinical visits in which nurses PPS score of 40, and had an Patient Functional Status: home hospice care alone, the only Disciplines: provided generalized identified family caregiver. PPS score (0-100, higher is individual support to study demonstrated a Study Duration - 30 Both patients and family better): standard care, mean Control Arm - Hospice care team; caregivers significant improvement in the days caregivers had to provide clinical disciplines not specified. standard home hospice care 51.42 (SD 9.96); standard Risk of Bias - 4 consent, had at least a sixth care plus supportive visits, Intervention Arm 1 - Hospice care plus COPE intervention group grade education, were able to mean 52.57 (SD 11.09); team; nurses, home health aide. Intervention Arm 2: in symptom distress. The read and understand English, standard care plus COPE study did not demonstrate Intervention Arm 2 - Hospice care Hospice; decision and had a minimum score of intervention, mean 54.50 (SD significant changes in pain, team; nurses, home health aide. support; COPE seven on the SPMSQ. 7.88). intervention, which dyspnea, constipation, or Patients were excluded if they HRQOL. teaches and supports did not have at least two of the family caregiver three following symptoms: Caregiver Symptoms: problem solving pain, dyspnea, and None described constipation. Decision support elements: Number of Patient Participants and Decedents: Printed decision aid Screened - 354 Enrolled - 329 Intervention Duration: Completed - 103 Nine days Decedents - Not described

Number of Caregiver Participants: Screened - 354 Enrolled - 329 Completed - 103

133

Evidence Table 2: Outcomes

McMillan et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 101 al., 2007 US Pain - NRS Outcomes Study Arms - 3 Dyspnea - Dyspnea Intensity Scale Other Symptoms: Comparing standard home Level of hospice care plus COPE Other Symptoms - MSAS, CAS Symptom distress (0-96 scale, lower is intervention, standard home Randomization - better): random-effects model Patients and HRQOL - HQLI hospice care plus supportive estimate, baseline, 21.243 (SE 1.899); visits by nurses, and standard families/caregivers 30 days, -0.101 (SE -0.038); group by only home hospice care alone, the Caregiver Outcomes: time significance for standard home study demonstrated a Study Duration - 30 hospice care plus COPE intervention None described significant improvement in the days showing greater change than standard standard home hospice care Risk of Bias - 4 home hospice care plus supportive plus COPE intervention group visits and standard home hospice care in symptom distress. The alone, p=.009. study did not demonstrate significant changes in pain, dyspnea, constipation, or Caregiver Outcomes: HRQOL. N/A

134

Evidence Table 1: Populations and Interventions

McMillan et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 97 al., 2011 US Criteria: Cancer (type not Depression (0-10 scale, lower Control Arm - Home Usual care, hospice Outcomes Study Arms - 2 Patients were eligible for described) is better): all participants, Intervention Arm - Home Comparing an intervention Level of inclusion if they were at least random-effects model that employed systematic 18 years old, had a diagnosis estimate 4.51 (SE 0.11). Intervention Arm: assessment of patient and Randomization - Prognosis: Patients and of cancer, were newly Symptom distress (0-4 scale, Multidisciplinarity and Clinical Hospice caregiver outcomes in home admitted to home hospice hospice care with usual home families/caregivers Referred to or receiving lower is better): all Disciplines: care, were able to read and hospice care alone, the study only hospice participants, random-effects Control Arm - Hospice care team; understand English, and were Intervention demonstrated a significant model estimate 1.99 (SE physicians, nurses, social workers, Study Duration - able to pass mental status Duration: improvement in the 0.06). chaplains, bereavement Two weeks screening. intervention group in counselors, home health aides, Two weeks depression. The study did not Risk of Bias - 2 volunteer coordinators. Caregivers were eligible for Patient Functional Status: demonstrate significant inclusion if they were at least Intervention Arm - Hospice care changes in overall HRQOL, PPS score (0-100 scale, 18 years old, were identified team; physicians, nurses, social symptom distress, or spiritual higher is better): control, mean by the hospice as the primary workers, chaplains, bereavement needs. 57.18 (SD 11.13); caregiver, and provided at counselors, home health aides, intervention, mean 57.06 (SD least four hours of care each volunteer coordinators. 10.95). day. They were excluded if Caregiver QOL-Relevant they were in active treatment Outcomes for cancer. Caregiver Symptoms: The study did not demonstrate significant changes in Depression (0-10 scale, lower caregiver depression, spiritual is better): all participants, Number of Patient needs, or whether or not the random-effects model Participants and Decedents: caregiver received support. estimate 4.48 (SE 0.10). Screened - Not described Enrolled - 716 Completed - 350 Decedents - Not described

Number of Caregiver Participants: Screened - Not described Enrolled - 716 Completed - 350

135

Evidence Table 2: Outcomes

McMillan et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 97 al., 2011 US Depression - CES-D Outcomes Study Arms - 2 Other Symptoms - Modified MSAS Depression: Comparing an intervention Level of distress for each symptom subscales that employed systematic Depression (0-10 scale, lower is assessment of patient and Randomization - Existential or Spiritual Well-Being - better): baseline, random-effects model Patients and caregiver outcomes in home Spiritual Needs Inventory estimate 4.51 (SE 0.11); two weeks, - hospice care with usual home families/caregivers 0.03 (SE 0.01); group by time only HRQOL - HQLI-14 hospice care alone, the study significance for intervention vs. control, demonstrated a significant

Study Duration - p=.027. improvement in the Two weeks Caregiver Outcomes: intervention group in depression. The study did not Risk of Bias - 2 Depression - CES-D Caregiver Outcomes: demonstrate significant Existential or Spiritual Well-Being - None changes in overall HRQOL, Spiritual Needs Inventory symptom distress, or spiritual Other QOL-Relevant Outcomes - needs. Study-specific received support item Caregiver QOL-Relevant Outcomes The study did not demonstrate significant changes in caregiver depression, spiritual needs, or whether or not the caregiver received support.

136

Evidence Table 1: Populations and Interventions

Melis et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 129 2009 EU Criteria: Not described None described Control Arm - Home, nursing home Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home, nursing Comparing an intervention of Level of inclusion if they were at least home in-home nurse visits with frail 70 years old, lived in their own Prognosis: Patient Functional Status: Intervention Arm: elders to discuss problems Randomization - Patients and home or a nursing home, and Not described All participants had one or In-home nurse visits identified by a GP, with usual families/caregivers had one or more limitations in more ADL or IADL Multidisciplinarity and Clinical to discuss problems care, the study demonstrated only cognition, ADLs or IADLs, or dependencies. Disciplines: identified by a GP a significant improvement in mental well-being. They were the intervention group in Control Arm - No multidisciplinary Study Duration - excluded if they were already caregiver burden for team; physicians. Six months receiving forms of Caregiver Symptoms: Intervention caregivers who lived with the Risk of Bias - 3 intermediate care or health None described Intervention Arm - Intervention Duration: patient, but not for all care from a social worker or team; physicians, nurses. Three months caregivers. The study did not community-based geriatrician, demonstrate a significant were on the waiting list for a change in caregiver time spent nursing home, or had a on personal care. terminal illness and life expectancy of less than six months.

Number of Patient Participants and Decedents: Screened - 168 Enrolled - 110 Completed - 99 Decedents - Not described

Number of Caregiver Participants: Screened - 126 Enrolled - 110 Completed - 99

137

Evidence Table 2: Outcomes

Melis et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 129 2009 EU None described N/A Outcomes Study Arms - 2 Comparing an intervention of in-home nurse visits with frail Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - elders to discuss problems Patients and Other QOL-Relevant Outcomes - identified by a GP, with usual care, the study demonstrated families/caregivers ZBI, study-specific time spent on Other QOL-Relevant Outcomes: only personal care item a significant improvement in Caregiver burden (0-88 scale, lower is the intervention group in Study Duration - better), subgroup of caregivers who caregiver burden for Six months lived with the patient: six months, caregivers who lived with the Risk of Bias - 3 intervention vs. control, -6.2; p=.04. patient, but not for all caregivers. The study did not demonstrate a significant change in caregiver time spent on personal care.

138

Evidence Table 1: Populations and Interventions

Moorey et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 47 al., 2009 EU Criteria: Cancer (mixed cancer) Anxiety (0-21 scale, lower is Control Arm - Home Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for better): control, mean 10.71 Intervention Arm - Home care team Comparing palliative care plus inclusion if they were new (SD 0.85); intervention, mean cognitive behavioral therapy Level of Prognosis: Randomization - palliative care patients, had a 10.91 (SD 0.61); p=.85. with palliative care alone, the diagnosis of cancer, and Referred to or receiving Multidisciplinarity and Clinical Palliative care team study demonstrated a Providers Depression (0-21 scale, lower elements: scored eight or higher for palliative care is better): control, mean 10.78 Disciplines: significant improvement in the Study Duration - 16 anxiety or depression on the intervention group in anxiety. (SD 0.62); intervention, mean Control Arm - Palliative care team; Outpatient weeks HADS anxiety and depression The study did not demonstrate 9.91 (SD 0.55); p=.30. advanced practice nurses, social consultation subscales, respectively. They a significant change in Risk of Bias - 5 workers, chaplains, psychiatrists. were excluded if they had Pain (scale not described, depression. significant cognitive, lower is better): control, mean Intervention Arm - Palliative care Intervention Arm: 3.86 (SD 3.28); intervention, communication, or language team; advanced practice nurses, Palliative care team, mean 3.91 (SD 3.44); p=0.94. difficulties that made it difficult social workers, chaplains, CBT for them to participate in the Tiredness (scale not psychiatrists.

research interviews or described, lower is better): psychological therapy; or were control, mean 7.37 (SD 2.51); Palliative care team so severely ill or debilitated intervention, mean 6.98 (SD elements: that they could not participate 2.49); p=0.44. Outpatient in the study. Appetite (scale not described, consultation

higher is better): control, mean Number of Patient 4.74 (3.75); intervention, Intervention Participants and Decedents: mean 4.36 (SD 3.55); p=.09. Duration: Screened - 310 16 weeks Enrolled - 80 Patient Functional Status: Completed - 34 ECOG performance status (0- 5 scale, lower is better): Decedents - 23 control, mean 1.54 (SD 0.89); Screened - Not described intervention, mean 1.52 (SD 0.93).

Caregiver Symptoms: None described

139

Evidence Table 2: Outcomes

Moorey et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 47 al., 2009 EU Depression - HADS depression Outcomes Study Arms - 2 subscale Anxiety: Comparing palliative care plus Level of Anxiety - HADS anxiety subscale cognitive behavioral therapy Anxiety (0-21 scale, lower is better): 16 with palliative care alone, the Randomization - weeks, control, high scores, 56%; Providers study demonstrated a Caregiver Outcomes: intervention, high scores, 19%; p=.04 significant improvement in the Study Duration - 16 intervention group in anxiety. None described weeks The study did not demonstrate Caregiver Outcomes: Risk of Bias - 5 a significant change in N/A depression.

140

Evidence Table 1: Populations and Interventions

O’Hara et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 20 2010 US Criteria: Cancer (mixed cancer) None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a patient inclusion if they were at least psychoeducational and case Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - 18 years old; had received a management intervention with Patients diagnosis of lung, breast, Identified by clinicians as None described Multidisciplinarity and Clinical Palliative case usual care, the study did not gastrointestinal, or demonstrate a significant poor prognosis Disciplines: management, Study Duration - 13 genitourinary cancer within the change in caregiver burden. Control Arm - No multidisciplinary advance directive months past 12 weeks; and had a life Caregiver Symptoms: team; clinical disciplines not completion expectancy of approximately Risk of Bias - 3 None described specified. one year. They were excluded if they had dementia, severe Intervention Arm - No Palliative case confusion, schizophrenia, multidisciplinary team; nurses. management bipolar disorder, or an active elements: substance abuse disorder. Nurse-led case management Number of Patient Participants and Decedents: Advance directive Screened - Not described completion elements: Enrolled - 322 Advance directive and DNR completion Completed - 86

Decedents - 153 Intervention

Duration: Number of Caregiver 13 months Participants: Screened - Not described Enrolled - 220 Completed - 198

141

Evidence Table 2: Outcomes

O’Hara et al., Study Location - Patient Outcomes: None None N/A Caregiver QOL-Relevant 20 2010 US None described Outcomes Study Arms - 2 Comparing a patient psychoeducational and case Level of Caregiver Outcomes: Randomization - management intervention with Patients Other QOL-Relevant Outcomes - usual care, the study did not Montgomery Borgatta Caregiver demonstrate a significant Study Duration - 13 Burden Scale change in caregiver burden. months Risk of Bias - 3

142

Evidence Table 1: Populations and Interventions

Temel et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 58 2010 US Criteria: Cancer (NSCLC) Depression (0-21 scale, lower Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for is better): score greater than Intervention Arm - Outpatient clinics Comparing early palliative inclusion if they had 7, control, 25%; intervention, care with usual care for Level of Prognosis: Intervention Arm: Randomization - pathologically confirmed 22%; p=.70. patients with newly diagnosed Patients NSCLC diagnosed within the Not described Major depression syndrome Multidisciplinarity and Clinical Palliative care team NSCLC, the study previous eight weeks; had an demonstrated significant (determined by PHQ-9 score): Disciplines: Study Duration - 12 ECOG performance status of improvements in the control, 17%; intervention, Control Arm - No multidisciplinary weeks 0 (asymptomatic), 1 Palliative care team intervention group in survival, 12%; p=.48. team; clinical disciplines not Risk of Bias - 3 (symptomatic but fully elements: HRQOL, depression, and lung Anxiety (0-21 scale, lower is specified. ambulatory), or 2 Outpatient cancer symptoms, and a better): score greater than 7, (symptomatic and in bed less Intervention Arm - Palliative care consultation significant increase in the rate than half of the day); and were control, 33%; intervention, team; physicians, advanced of documentation of able to read and respond to 36%; p=.73. practice nurses. resuscitation preferences in questions in English. They Intervention the outpatient electronic were excluded if they were Duration: medical record. The study did Patient Functional Status: already receiving palliative not demonstrate a significant 12 weeks care. ECOG performance status (0- change in anxiety. 5 scale, lower is better): 0,

control, 41%; intervention, Number of Patient 34%; 1, 47%; intervention, Health Care Use Participants and Decedents: 60%; 3, control, 12%; The study demonstrated a Screened - 283 intervention, 6%; p=.24. significant decrease in the Enrolled - 151 intervention group in aggressive care. Completed - 107 Caregiver Symptoms: Decedents - 27 None described

143

Evidence Table 2: Outcomes

Temel et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Aggressive Health Care Use: Patient QOL-Relevant 58 2010 US Depression - HADS depression care (defined as chemotherapy Aggressive care (chemotherapy Outcomes Study Arms - 2 subscale, PHQ-9 within 14 days of death, no within 14 days of death, no Comparing early palliative Depression: hospice care, or admission to Level of Anxiety - HADS anxiety subscale hospice care, or admission to care with usual care for Depression (0-21 scale, lower is hospice within three days of hospice within three days of patients with newly diagnosed Randomization - Other Symptoms - FACT-L LCS better): score greater than 7, control, death) Patients death): control, 54%; NSCLC, the study Communication or Care Planning - baseline, 25%; 12 weeks, 38%; intervention, 33%; p=.05. demonstrated significant Study Duration - 12 Study-specific resuscitation intervention, baseline, 22%; 12 weeks, improvements in the weeks preferences documented in the 16%; p=.01. intervention group in survival, Risk of Bias - 3 outpatient electronic medical record Major depression syndrome HRQOL, depression, and lung measure (determined by PHQ-9 score): control, cancer symptoms, and a significant increase in the rate HRQOL - FACT-L, TOI (sum of baseline, 17%; 12 weeks, 17%; of documentation of FACT-L LCS and physical and intervention, baseline, 12%; 12 weeks, resuscitation preferences in functional well-being subscales) 4%; p=.01. the outpatient electronic

medical record. The study did Caregiver Outcomes: Other Symptoms: not demonstrate a significant change in anxiety. None described Lung cancer symptoms (0-28 scale, higher is better): baseline, control, mean 18.7 (SD 4.4); intervention, Health Care Use mean 20.1 (SD 4.4); 12 weeks, control, mean 19.3 (SD 4.2); intervention, The study demonstrated a mean 21.0 (SD 3.9); p=.04. significant decrease in the intervention group in

aggressive care. Communication or Care Planning: Resuscitation preferences documented in the outpatient electronic medical record: control, 53%; intervention, 28%; p=.05.

HRQOL: QOL (0-136 scale, higher is better): baseline, control, mean 91.7 (SD 16.7); intervention, mean 93.6 (SD 16.5); 12 weeks, control, mean 91.5 (SD 15.8); intervention, mean 98.0 (SD 15.1); p=.03. QOL (0-84 scale, higher is better): baseline, control, mean 55.3 (SD 13.1); intervention, mean 56.2 (SD 13.4); 12 weeks, control, mean 53.0 (SD 11.5); intervention, mean 59.0 (SD 11.6); p=.003.

Caregiver Outcomes: N/A

144

Evidence Table 1: Populations and Interventions

Volandes et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 107 al., 2011 US Criteria: Not described None described Control Arm - Outpatient clinics Decision support, Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient clinics other care planning Comparing a verbal Level of inclusion if they received care description of advanced at a rural primary care clinic in Prognosis: Patient Functional Status: dementia plus a video Randomization - Decision support Patients Greensburg, Louisiana; were Not described None described Multidisciplinarity and Clinical decision aid with a verbal at least 65 years old; could elements: description alone for primary Disciplines: Study Duration - communicate in English; could care patients, the study Control Arm - No multidisciplinary Verbal decision Immediately after provide informed consent; and Caregiver Symptoms: demonstrated a significant team; clinical disciplines not support intervention did not have moderate or increase in the video decision None described specified. Risk of Bias - 3 severe cognitive impairment, aid group in choice of comfort as determined by a score of Intervention Arm - No Intervention Arm: care. less than seven on the multidisciplinary team; clinical Decision support, SPMSQ. disciplines not specified. other care planning

Number of Patient Decision support Participants and Decedents: elements: Screened - 88 Video-based decision Enrolled - 76 aid, verbal decision support Completed - 76

Decedents - 0 Intervention Duration: One-time intervention

145

Evidence Table 2: Outcomes

Volandes et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 107 al., 2011 US Communication or Care Planning - Outcomes Study Arms - 2 Study-specific goals of care Communication or Care Planning: Comparing a verbal Level of preferences item description of advanced Goals of care preferences: verbal dementia plus a video Randomization - description, life-prolonging care, 16%; Patients decision aid with a verbal Caregiver Outcomes: limited care, 12%; comfort care, 72%; description alone for primary Study Duration - None described verbal description plus video, life- care patients, the study Immediately after prolonging care, 0%; limited care, 9%; demonstrated a significant intervention comfort care, 91%; p=.047. increase in the video decision Risk of Bias - 3 aid group in choice of comfort care. Caregiver Outcomes: N/A

146

Evidence Table 1: Populations and Interventions

Volandes et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 117 al., 2009 US Criteria: Not described None described Control Arm - Outpatient clinics Decision support, Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient clinics other care planning Comparing a verbal narrative Level of inclusion if they were at least of advanced dementia and a 65 years old, were Prognosis: Patient Functional Status: video decision aid depicting Randomization - Decision support Patients and community-dwelling, had a Not described None described Multidisciplinarity and Clinical advanced dementia with the SPMSQ 18 score of at least elements: verbal narrative alone, the families/caregivers Disciplines: seven, and were able to study demonstrated a only Control Arm - No multidisciplinary Verbal narrative provide informed consent. Caregiver Symptoms: significant improvement in the team; clinical disciplines not describing advanced Study Duration - video decision aid group in None described specified. dementia One day Caregivers were eligible for patient and surrogate

Risk of Bias - 4 inclusion if they were the Intervention Arm - No concordance of preferences patient’s designated surrogate multidisciplinary team; clinical Intervention Arm: for goals of care. and were able to provide disciplines not specified. Decision support, informed consent. other care planning

Number of Patient Decision support Participants and Decedents: elements: Screened - 14 Video-based decision Enrolled - 14 aid, verbal narrative describing advanced Completed - 14 dementia Decedents - Not described

Intervention Number of Caregiver Duration: Participants: One-time intervention Screened - 14 Enrolled - 14 Completed - 14

147

Evidence Table 2: Outcomes

Volandes et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 117 al., 2009 US None described N/A Outcomes Study Arms - 2 Comparing a verbal narrative of advanced dementia and a Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - video decision aid depicting Patients and Communication or Care Planning - advanced dementia with the verbal narrative alone, the families/caregivers Study-specific preferences for goals Communication or Care Planning: only of care and patient/surrogate study demonstrated a concordance of preferences Preferences for goals of care: control, significant improvement in the Study Duration - measures life-prolonging care, 33%; limited care, video decision aid group in One day 17%; comfort care, 50%; intervention, patient and surrogate Risk of Bias - 4 life-prolonging care, 0%; limited care, concordance of preferences 0%; comfort care, 100%; for goals of care. patient/surrogate concordance of preferences, control, 33%; intervention, 100%; p=.015.

148

Evidence Table 1: Populations and Interventions

Walsh et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 48 2007 EU Criteria: Cancer (mixed cancer) None described Control Arm - Outpatient clinics, Usual care, palliative Outcomes Study Arms - 2 Caregivers were eligible for home care team Comparing outpatient inclusion if they provided specialist palliative care plus a Level of Prognosis: Patient Functional Status: Intervention Arm - Outpatient Randomization - informal care to cancer clinics, home, workplace caregiver support intervention patients newly referred to Referred to or receiving ECOG performance status (0- Palliative care team with outpatient specialist Families or elements: caregivers palliative care and scored palliative care 5 scale, lower is better): palliative care alone for above the threshold of 5 on control, mean 2.8 (SD 0.94); Multidisciplinarity and Clinical Outpatient caregivers of patients with Study Duration - 12 the GHQ-28. They were intervention, mean 2.7 (SD Disciplines: consultation cancer, the study did not weeks excluded if their English skills demonstrate any significant 0.91); p>.05. Control Arm - Palliative care team; would mean they could not changes in caregiver Risk of Bias - 6 physicians, advanced practice gain full benefit from the Intervention Arm: psychological distress, strain, nurses, nurses, social workers. adviser visits or if the patient Caregiver Symptoms: Palliative care team; or HRQOL. were unlikely to survive the None described Intervention Arm - Palliative care intervention to assess time it would take to introduce team; physicians, advanced domains of need and the intervention. practice nurses, nurses, social provide information, workers, career advisors with advice, and emotional experience in community nursing support to caregivers Number of Patient and social work. Participants and Decedents: Screened - Not described Palliative care team elements: Enrolled - Not described Outpatient Completed - Not described consultation Decedents - 109

Intervention Number of Caregiver Duration: Participants: Six weeks Screened - 1557 Enrolled - 271 Completed - 123

149

Evidence Table 2: Outcomes

Walsh et al., Study Location - Patient Outcomes: None None N/A Caregiver QOL-Relevant 48 2007 EU None described Outcomes Study Arms - 2 Comparing outpatient specialist palliative care plus a Level of Caregiver Outcomes Randomization - caregiver support intervention Families or HRQOL - CQOLC with outpatient specialist palliative care alone for caregivers Other QOL-Relevant Outcomes - caregivers of patients with Psychological distress (GHQ-28), Study Duration - 12 cancer, the study did not Carer Strain Index weeks demonstrate any significant Risk of Bias - 6 changes in caregiver psychological distress, strain, or HRQOL.

150

Evidence Table 1: Populations and Interventions

Spiro et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 121 2004 EU Criteria: Cancer (NSCLC) None described Control Arm - Not described Supportive care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Not described Comparing supportive care inclusion if they fulfilled the plus chemotherapy with Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - local criteria for histological or supportive care alone for Patients cytological diagnosis of Not described ECOG performance status (0- Multidisciplinarity and Clinical Supportive care plus patients with lung cancer, the NSCLC, were considered 5 scale, lower is better): 0, Disciplines: chemotherapy study did not demonstrate any Study Duration - 12 unsuitable for or declined significant changes in supportive care alone, 24%; Control Arm - No multidisciplinary weeks radical radiotherapy or HRQOL, physical functioning, supportive care plus team; clinical disciplines not surgery, were considered fit to Intervention emotional functioning, pain, Risk of Bias - 2 chemotherapy, 22%; 1, specified. receive chemotherapy, had no supportive care alone, 53%; Duration: dyspnea, or fatigue. Intervention Arm - No concurrent malignancy or supportive care plus 12 weeks history of malignancy other chemotherapy, 56%; 2, multidisciplinary team; clinical than non-melanomatous skin supportive care alone, 21%; disciplines not specified. Health Care Costs cancer within the last three supportive care plus The study did not demonstrate years, and, along with the chemotherapy, 20%; 3, a significant change in overall doctor, were uncertain about supportive care alone, 2%; costs of care. the value of chemotherapy. supportive care plus chemotherapy, 2%; p>.05. Number of Patient Participants and Decedents: Caregiver Symptoms: Screened - Not described None described Enrolled - 725 Completed - 134 Decedents - 697

151

Evidence Table 2: Outcomes

Spiro et al., Study Location - Patient Outcomes: None Health Care Costs - Overall None Patient QOL-Relevant 121 2004 EU Pain - Adapted MRC Diary Card costs Outcomes Study Arms - 2 Dyspnea - Adapted MRC Diary Card Comparing supportive care Level of plus chemotherapy with Other Symptoms - Fatigue (Adapted supportive care alone for Randomization - MRC Diary Card) Patients patients with lung cancer, the Functional Status - EORTC QLQ- study did not demonstrate any Study Duration - 12 C30 physical functioning subscale significant changes in weeks HRQOL, physical functioning, HRQOL - EORTC QLQ-C30 global emotional functioning, pain, Risk of Bias - 2 health status/QOL subscale dyspnea, or fatigue. - EORTC QLQ-C30 emotional

functioning subscale Health Care Costs

The study did not demonstrate Caregiver Outcomes: a significant change in overall None described costs of care.

152

Evidence Table 1: Populations and Interventions

McMillan et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 37 al., 2005 US Criteria: Cancer (type not None described Control Arm - Home, hospice Usual care, hospice Outcomes Study Arms - 3 Caregivers were eligible for described) Intervention Arm 1 - Home, hospice Comparing hospice care for Level of inclusion if they were patients with cancer plus a providing care for adult Patient Functional Status: Intervention Arm 2 - Home, hospice Intervention Arm 1: caregiver coping skills Randomization - Prognosis: Patients and patients with cancer. Both None described Hospice, supportive intervention, hospice care plus patients and caregivers were supportive visits, and hospice families/caregivers Not described Multidisciplinarity and Clinical visits for caregivers eligible for inclusion if they care alone, the study only Disciplines: consented to participate, had Caregiver Symptoms: demonstrated significant Study Duration - 30 at least a sixth grade None described Control Arm - No multidisciplinary Intervention Arm 2: improvements in the coping days education, were able to read team; clinical disciplines not Hospice; COPE skills group in caregiver Risk of Bias - 7 and understand English, and specified. intervention, which HRQOL, distress caused by achieved a score of at least Intervention Arm 1 - No teaches a problem- patient symptom burden, and seven on the SPMSQ. They multidisciplinary team; nurses, solving method to caregiving task burden. The were excluded if they were in hospice. assist caregivers with study did not demonstrate active treatment for cancer assessing and significant changes in general themselves. Dyads were Intervention Arm 2 - No caregiving mastery, caregiving multidisciplinary team; nurses, managing patient excluded if it was unclear who symptoms task mastery, or problem- or the primary caregiver would hospice. emotion-focused coping. be or if patient performance status suggested that patients Intervention would not survive more than a Duration: few days. 30 days

Number of Patient Participants and Decedents: Screened - 354 Enrolled - 329 Completed - 103 Decedents - 69

Number of Caregiver Participants: Screened - 354 Enrolled - 329 Completed - 103

153

Evidence Table 2: Outcomes

McMillan et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 37 al., 2005 US None described N/A Outcomes Study Arms - 3 Comparing hospice care for patients with cancer plus a Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - caregiver coping skills Patients and HRQOL - CQOLC intervention, hospice care plus supportive visits, and hospice families/caregivers Other QOL-Relevant Outcomes - HRQOL: care alone, the study only Undescribed general caregiving Caregiver QOL (scale not described): demonstrated significant mastery scale, caregiving task Study Duration - 30 30 days vs. baseline, intervention vs. improvements in the coping mastery and task burden (CDS), days control, random effects intercept -0.09 skills group in caregiver problem-and emotion-focused (SE 0.05); group by time p=.05. HRQOL, distress caused by Risk of Bias - 7 coping (Brief COPE Scale), distress patient symptom burden, and caused by patient symptom burden caregiving task burden. The (adapted MSAS) Other QOL-Relevant Outcomes: study did not demonstrate Caregiving task burden (scale not significant changes in general described): 30 days vs. baseline, caregiving mastery, caregiving intervention vs. control, random effects task mastery, or problem- or intercept -0.01 (SE 0.01); group by emotion-focused coping. time p=.021. Distress caused by patient symptom burden (scale not described): 30 days vs. baseline, intervention vs. control, random effects intercept -0.14 (SE 0.04); group by time p=.001.

154

Evidence Table 1: Populations and Interventions

Hudson et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 61 al., 2005 Australia Criteria: Not described None described Control Arm - Home Palliative care team, Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home palliative case Comparing outpatient inclusion if they had advanced management palliative care plus caregiver Level of Prognosis: Patient Functional Status: Randomization - cancer, an ECOG score of 0- support with outpatient 3, and were admitted to a Not described ECOG performance status (0- Multidisciplinarity and Clinical palliative care alone, the study Families or Palliative care team caregivers home-based palliative care 5 scale, lower is better): all Disciplines: demonstrated a significant service within the past week. elements: improvement in the caregiver participants, 1, 0.9%; 2, Control Arm - Palliative care team; Study Duration - support group in caregiving 39.7%; 3, 48.1%; 4, 11.3%. physicians, advanced practice Outpatient Eight weeks after Caregivers were eligible for rewards. The study did not (Only caregivers of patients nurses, nurses, allied health consultation patient death inclusion if they were the demonstrate significant with ECOG scores of 0-3 were professionals. Risk of Bias - 10 primary family caregiver, recruited, but some patients changes in caregiving resided with the patient, were had progressed to an ECOG Intervention Arm - Palliative care Palliative case preparedness, competence, at least 18 years old, were score of 4 by the time the first team; physicians, advanced management self-efficacy, or caregiver able to meet English language data were collected.) practice nurses, nurses, allied elements: anxiety. requirements of participation, health professionals. Nurse-led case and did not have an management intellectual or psychiatric Caregiver Symptoms:

illness that would preclude Anxiety (0-21 scale, lower is completion of study better): control, mean 8.49 Intervention Arm: requirements. (SD 4.99); intervention, mean Palliative care team, 7.68 (SD 3.74). palliative case Number of Caregiver management, Participants: caregiver support Screened - 684 Enrolled - 106 Palliative care team elements: Completed - 27 Outpatient consultation

Palliative case management elements: Nurse-led case management

Intervention Duration: Eight weeks after patient death

155

Evidence Table 2: Outcomes

Hudson et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 61 al., 2005 Australia None described N/A Outcomes Study Arms - 2 Comparing outpatient palliative care plus caregiver Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - support with outpatient Families or Anxiety - HADS anxiety subscale palliative care alone, the study demonstrated a significant caregivers Other QOL-Relevant Outcomes - Other QOL-Relevant Outcomes: improvement in the caregiver Rewards of Caregiving Scale, Study Duration - Caregiving rewards (scale not support group in caregiving Preparedness for Caregiving Scale, Eight weeks after described, higher is better): five weeks rewards. The study did not Caregiver Competence Scale, patient death vs. baseline, control, baseline, mean demonstrate significant unnamed caregiving self-efficacy 3.12 (SD 0.70); five weeks, mean 2.82 changes in caregiving Risk of Bias - 10 scale (SD 0.99); intervention, baseline, mean preparedness, competence, 3.01 (SD 0.83); five weeks, mean 3.09 self-efficacy, or caregiver (SD 0.82); group by time p=.005; eight anxiety. weeks after patient death vs. baseline, control, baseline, mean 3.08 (SD 0.83); eight weeks after patient death, mean 3.04 (SD 0.82); intervention, baseline, mean 3.05 (SD 1.10); eight weeks after patient death, mean 3.50 (SD 0.70); group by time p=.036.

156

Evidence Table 1: Populations and Interventions

Rummans et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 94 al., 2006 US Criteria: Cancer (mixed cancer) None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Not described Comparing a multi-session inclusion if they had been intervention focused on Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosed with cancer in the improving QOL in cancer Patients last 12 months, had an Identified by clinicians as ECOG performance status (0- Multidisciplinarity and Clinical Multi-session patients not yet eligible for expected survival time of at poor prognosis 5 scale, lower is better): Disciplines: intervention focused palliative or hospice services, Study Duration - 27 least six months, had a 5-year with usual care, the study control, 0, 31.5%; 1, 63.0%; 2, Control Arm - No multidisciplinary on improving QOL weeks survival probability of at most demonstrated significant 5.6%; intervention, 0, 32.7%; team; clinical disciplines not with cognitive, 50% as routinely determined improvements in the Risk of Bias - 3 1, 65.3%; 2, 2.0%; p=.65. specified. emotional, physical, by the primary radiation intervention group in HRQOL, social, and spiritual oncologist, and had a Intervention Arm - Multidisciplinary components emotional well-being, treatment recommendation of Caregiver Symptoms: team for patients who did not yet tension/anxiety, and

radiation therapy of at least None described qualify for palliative or hospice confusion/bewilderment at four two weeks. They were services; physicians, advanced Intervention weeks. The study did not excluded if they had previous practice nurses, social workers, Duration: demonstrate significant radiation therapy, recurrent chaplains, physical therapists, Three weeks changes in pain, fatigue, or disease after a disease-free psychiatrists, psychologists. spiritual well-being. At eight period of greater than six and 27 weeks, the study did months, a previous cancer not demonstrate a significant diagnosis within five years, an change in HRQOL and did not MMSE score of less than 20, report the other measures. an ECOG performance score of 3 or more, active alcohol or substance dependence (excluding nicotine), an active thought disorder, suicidal plans, or were participating in another psychosocial research trial.

Number of Patient Participants and Decedents: Screened - 418 Enrolled - 115 Completed - 82 Decedents - Not described

157

Evidence Table 2: Outcomes

Rummans et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 94 al., 2006 US Pain - LASA pain frequency and Outcomes Study Arms - 2 severity subscales Anxiety: Comparing a multi-session Level of Anxiety - POMS-SF tension/anxiety intervention focused on Tension/anxiety (scale not described, improving QOL in cancer Randomization - subscale lower is better): four weeks, lower in Patients patients not yet eligible for Other Symptoms - POMS-SF intervention vs. control, p=.042. palliative or hospice services, Study Duration - 27 confusion/bewilderment subscale, with usual care, the study weeks LASA fatigue subscale demonstrated significant Other Symptoms: Risk of Bias - 3 Existential or Spiritual Well-Being - improvements in the LASA spiritual well-being subscale, Confusion/bewilderment (scale not intervention group in HRQOL, FACIT-Sp described, lower is better): four weeks, emotional well-being, lower in intervention vs. control, tension/anxiety, and HRQOL - LASA QOL subscale, p=.014. confusion/bewilderment at four Spitzer QOL Uniscale weeks. The study did not Other QOL-Relevant Outcomes - demonstrate significant LASA emotional well-being subscale HRQOL: changes in pain, fatigue, or QOL (0-100 scale, higher is better): spiritual well-being. At eight control, baseline, mean 73.0 (SD and 27 weeks, the study did Caregiver Outcomes: 20.80); four weeks, mean 64.1 (SD not demonstrate a significant None described 22.53); intervention, baseline, mean change in HRQOL and did not 70.0 (SD 21.89); four weeks, mean report the other measures. 72.8 (SD 20.62); p=.047.

Other QOL-Relevant Outcomes: Emotional well-being (0-100 scale, higher is better): control scores declined from baseline to four weeks; intervention scores increased from baseline to four weeks; p<.05.

Caregiver Outcomes: N/A

158

Evidence Table 1: Populations and Interventions

McLean et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 82 al., 2013 Canada Criteria: Cancer (mixed cancer) None described Control Arm - Outpatient clinics Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient clinics care team Comparing a couples therapy inclusion if they had intervention focused on Level of Prognosis: Patient Functional Status: Randomization - metastatic cancer, were patterns of interaction with English-speaking, were at Not described KPS score (0-100 scale, Multidisciplinarity and Clinical Palliative care team usual care for patients with Patients and elements: families/caregivers least 18 years old, were in a higher is better): control, mean Disciplines: metastatic cancer, the study romantic partnership for at demonstrated a significant only 69.09 (SD 9.21); intervention, Control Arm - Palliative care team; Outpatient least one year, endorsed improvement in the mean 71.11 (SD 11.82); physicians, social workers, consultation Study Duration - marital distress in at least one intervention arm in marital p=.55. psychiatrist or psychologist. Three months partner, were not in couples functioning. The study did not

Risk of Bias - 5 therapy, and had a KPS score Intervention Arm - No Intervention Arm: demonstrate significant of at least 60. Patients and Caregiver Symptoms: multidisciplinary team; psychologist. EFT, a couples changes in patient depression caregivers were excluded if None described therapy intervention or hopelessness. they had any significant focused on patterns cognitive deficits, the patient of interaction was too ill to participate, or Caregiver QOL-Relevant either partner had major Outcomes psychiatric illness such as Intervention The study did not demonstrate psychosis or schizophrenia. Duration: significant changes in Three months caregiver depression, hopelessness, or burden. Number of Patient Participants and Decedents: Screened - 45 Enrolled - 42 Completed - 36 Decedents - 5

Number of Caregiver Participants: Screened - 45 Enrolled - 42 Completed - 36

159

Evidence Table 2: Outcomes

McLean et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 82 al., 2013 Canada Depression - BDI-II Outcomes Study Arms - 2 Other QOL-Relevant Outcomes - Other QOL-Relevant Outcomes: Comparing a couples therapy intervention focused on Level of RDAS, BHS Marital functioning (0-69 scale, higher Randomization - patterns of interaction with is better): baseline, control, mean usual care for patients with Patients and 44.22 (SD 7.6); intervention, mean families/caregivers Caregiver Outcomes: metastatic cancer, the study 43.14 (SD 4.8); post-intervention, demonstrated a significant only Depression - BDI-II control, mean 43.44 (SD 10.3); improvement in the Study Duration - Other QOL-Relevant Outcomes - intervention, mean 56.27 (SD 4.6); intervention arm in marital Three months RDAS, Caregiver Burden Scale, three months, control, mean 44.17 (SD functioning. The study did not Risk of Bias - 5 BHS 10.2); intervention, mean 55.39 (SD demonstrate significant 6.3); post-intervention vs. baseline, changes in patient depression p<.001; three months vs. baseline, or hopelessness. p<.001.

Caregiver QOL-Relevant Caregiver Outcomes: Outcomes The study did not demonstrate Other QOL-Relevant Outcomes: significant changes in caregiver depression, Marital functioning (0-69 scale, higher hopelessness, or burden. is better): baseline, control, mean 42.94 (SD 7.2); intervention, mean 46.41 (SD 7.0); post-intervention, control, mean 42.39 (SD 6.9); intervention, mean 54.32 (SD 4.6); three months, control, mean 44.56 (SD 10.3); intervention, mean 54.72 (SD 5.8); post-intervention vs. baseline, p<.001; three months vs. baseline, p<.001.

160

Evidence Table 1: Populations and Interventions

Raaf et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 79 2013 EU Criteria: Cancer (mixed cancer) Fatigue in the last week (0-10 Control Arm - Outpatient clinics Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for scale, lower is better): control, case management Comparing a symptom mean 6.1 (SD 1.6); intervention, Intervention Arm - Outpatient clinics Level of inclusion if they had a solid mean 6.2 (SD 1.4); p>.05. assessment and management malignancy, received Prognosis: intervention with usual care for Randomization - Moderate or severe pain in the last treatment with palliative intent, Not described Multidisciplinarity and Clinical Palliative case patients with advanced Patients week: control, 28%; intervention, management had a fatigue score of at least 35%; p>.05. Moderate or severe Disciplines: cancer, the study Study Duration - 4 on an NRS of 0 to 10, had elements: demonstrated significant nausea in the last week: control, Control Arm - No multidisciplinary Three months an ECOG performance status improvements in the 13%; intervention, 27%; p>.05 team; physicians. Nurse-led case Risk of Bias - 1 of at most 2, had a life Moderate or severe vomiting in the management intervention group in symptom expectancy of at least four Intervention Arm - No burden, anxiety, fatigue, last week: control, 4%; months, and were able to intervention, 13%; p>.05. multidisciplinary team; physicians, fatigue interference, reduced write and speak Dutch. They Moderate or severe diarrhea in the nurses. Intervention Arm: activity due to fatigue, and were excluded if they had last week: control, 12%; Palliative case reduced motivation due to received any experimental intervention, 13%; p>.05. management, fatigue. The study did not drug in the last four weeks, Moderate or severe constipation in symptom assessment demonstrate significant their oncologist considered the last week: control, 5%; and management changes in HRQOL, their anxiety or depressive intervention, 16%; p>.05. depression, or physical or

symptoms to require Moderate or severe anorexia in the mental fatigue separately. psychiatric consultation, they last week: control, 28%; Palliative case had severe comorbidity intervention, 41%; p>.05. management causing fatigue, they lived in Moderate or severe dyspnea in the elements: nursing homes, or they were last week: control, 35%; Nurse-led case cognitively impaired. intervention, 24%; p>.05. management Moderate or severe cough in the last week: control, 14%; Number of Patient intervention, 13%; p>.05. Intervention Participants and Decedents: Moderate or severe dry mouth in Duration: Screened - Not described the last week: control, 24%; intervention, 23%; p>.05. Three months Enrolled - 152 Anxiety (0-21 scale, lower is Completed - 137 better): control, mean 6.4 (SD 3.6); Decedents - 26 intervention, mean 5.6 (SD 3.6); p>.05. Depression (0-21 scale, lower is better): control, mean 7.1 (SD 4.3); intervention, mean 6.0 (SD 4.1); p>.05.

Patient Functional Status: ECOG performance status (0-5 scale, lower is better): 1, control, 88%; intervention, 87%; 2, control, 12%; intervention, 13%; p>.05.

Caregiver Symptoms: None described

161

Evidence Table 2: Outcomes

Raaf et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 79 2013 EU Depression - HADS depression Outcomes Study Arms - 2 subscale Anxiety: Comparing a symptom Level of Anxiety - HADS anxiety subscale assessment and management Anxiety (0-21 scale, lower is better): intervention with usual care for Randomization - Other Symptoms - Study-specific three months vs. baseline, intervention Patients patients with advanced symptom burden composite scale vs. control, mean -1.23; group by time cancer, the study Study Duration - adapted from EORTC QLQ-C30- p<.001. demonstrated significant Three months style questions; study-specific improvements in the fatigue in the last week NRS; MFI intervention group in symptom Risk of Bias - 1 Other Symptoms: General Fatigue, reduced activity, burden, anxiety, fatigue, reduced motivation, physical fatigue, Symptom burden (0-100 scale, lower is fatigue interference, reduced and mental fatigue subscales; BFI-I better): three months vs. baseline, activity due to fatigue, and HRQOL - Study-specific QOL scale intervention vs. control, mean -5.25; reduced motivation due to derived from EORTC QLQ-C30 group by time p=.002. fatigue. The study did not demonstrate significant Fatigue in the last week (0-10 scale, lower is better): three months vs. changes in HRQOL, Caregiver Outcomes: baseline, intervention vs. control, mean depression, or physical or None described -1.03; group by time p<.001. mental fatigue separately. General fatigue (4-20 scale, lower is better): three months vs. baseline, intervention vs. control, mean -1.11; group by time p=.01. Reduced activity due to fatigue (4-20 scale, lower is better): three months vs. baseline, intervention vs. control, mean -1.41; group by time p=.005. Reduced motivation due to fatigue (4- 20 scale, lower is better): three months vs. baseline, intervention vs. control, mean -1.12; group by time p=.03. Fatigue interference (0-10 scale, lower is better): three months vs. baseline, intervention vs. control, mean -1.15; group by time p<.001.

Caregiver Outcomes: N/A

162

Evidence Table 1: Populations and Interventions

Abernethy et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 50 Usual care, palliative care team al., 2013 Australia Criteria: Cancer (mixed cancer) Usual pain in past 24 hours Control Arm - Outpatient clinics Outcomes

Study Arms - 2x2x2 Patients were eligible for (0-10 scale, lower is better): Comparing (1) palliative care Case Conferencing Intervention - Palliative care team elements: factorial design inclusion if they were newly all participants, mean 1.8 (SD Outpatient clinics, home plus case conferencing, (2) referred to the palliative care Prognosis: 2.1). Outpatient consultation palliative care plus patient and Level of Patient and Caregiver Pain Randomization - service, had any pain in the Not described Worst pain in past week (0-10 caregiver education about preceding three months, were Management Education Case Conferencing pain management, and (3) Providers scale, lower is better): all Intervention: expected to live at least 48 Intervention - Outpatient clinics, palliative care plus GP Study Duration - participants, mean 4.2 (SD home Palliative care team, palliative hours after referral, resided in 3.5). case management, case education about pain Not provided the service’s geographic area, GP Pain Management Education conferencing management, with palliative Risk of Bias - 1 and had demonstrated mental Intervention - Outpatient clinics, care alone for newly referred competence as determined by Patient Functional Status: home Palliative care team elements: patients, the study an MMSE score of at least 24 Outpatient consultation demonstrated a significant AKPS score (0-100 scale, or a GP-identified caregiver or improvement in the case higher is better): all legal health care proxy who Multidisciplinarity and Clinical Palliative case management conferencing intervention in participants, mean 61.0 (SD could provide informed Disciplines: elements: performance status and a 13.8). Nurse-led case management consent. Control Arm - Palliative care team; significant improvement in the

physicians, nurses, social workers. patient and caregiver pain Caregiver Symptoms: Patient and Caregiver Pain management education Number of Patient Case Conferencing Intervention - Management Education Intervention: intervention in total symptom Participants and Decedents: None described Palliative care team; physicians, burden. The study did not nurses, social workers, chaplains, Palliative care team, patient and Screened - 1948 caregiver education about end- demonstrate significant physical therapist, clinical of-life pain management changes in HRQOL or usual Enrolled - 461 pharmacist, representative from or worst pain. Completed - 147 another medical or home care Palliative care team elements:

agency relevant to the patient’s Outpatient consultation Decedents - Not described care, home equipment service. Health Care Use Patient and Caregiver Pain GP Pain Management Education Intervention: The study demonstrated a Management Education significant reduction in the Palliative care team, GP Intervention - Palliative care team; education about end-of-life pain case conferencing intervention physicians, nurses, social workers. management in hospital admissions.

GP Pain Management Education Intervention - Palliative care team; Palliative care team elements: physicians, nurses, social workers. Outpatient consultation

Intervention Duration: Case Conferencing Intervention - One-time intervention Patient and Caregiver Pain Management Education Intervention - Four weeks GP Pain Management Education Intervention - Four weeks

163

Evidence Table 2: Outcomes

Abernethy et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Case Conferencing Intervention Patient QOL-Relevant 50 al., 2013 Australia Pain - BPI daily usual pain and worst hospital Health Care Use: Outcomes Study Arms - 2x2x2 pain intensity subscales Patient and Caregiver Pain Hospital admissions per patient Comparing (1) palliative care factorial design Other Symptoms - Daily total Management Education Intervention over course of study: control, plus case conferencing, (2) Level of symptom burden (MSAS) mean 1.70 (SD 0.13); palliative care plus patient and Other Symptoms: caregiver education about Randomization - Functional Status - AKPS intervention, mean 1.26 (SD Providers Total symptom burden (0-152 scale, 0.10); p=.007. pain management, and (3) HRQOL - McGill QOL Questionnaire lower is better): at study completion, palliative care plus GP Study Duration - education about pain palliative care alone, mean 19.2 (SD Not provided 0.9); palliative care plus patient and management, with palliative Risk of Bias - 1 Caregiver Outcomes: caregiver education about pain care alone for newly referred None described management, mean 16.2 (SD 1.1); patients, the study p=.033. demonstrated a significant improvement in the case

conferencing intervention in Case Conferencing Intervention performance status and a Functional Status: significant improvement in the patient and caregiver pain Performance status (0-100 scale, management education higher is better): at study completion, intervention in total symptom palliative care alone, mean 51.7 (SD burden. The study did not 2.3); palliative care plus case demonstrate significant conferencing, mean 57.3 (SD 1.5); changes in HRQOL or usual p=.037. or worst pain.

Caregiver Outcomes: Health Care Use N/A The study demonstrated a significant reduction in the case conferencing intervention in hospital admissions.

164

Evidence Table 1: Populations and Interventions

Wallen et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 51 2012 US Criteria: Cancer (mixed cancer) Pain intensity (0-20 scale, Control Arm - Hospital Usual care Outcomes Study Arms - 2 Patients were eligible for lower is better): control, mean Intervention Arm - Hospital Comparing inpatient palliative inclusion if they had advanced 2.95 (SD 4.18); intervention, care with usual care for post- Level of Prognosis: Intervention Arm: Randomization - malignancies and were mean 2.91 (SD 4.34); p>.05. operative patients with Patients undergoing surgical Not described Pain unpleasantness (0-20 Multidisciplinarity and Clinical Palliative care team advanced cancer, the study procedures in NCI Surgery did not demonstrate any scale, lower is better): control, Disciplines: Study Duration - 12 Branch clinical trials. significant changes in pain mean 3.28 (SD 3.89); Control Arm - Post-operative team; months Palliative care team intensity, pain intervention, mean 3.31 (SD physicians, social workers, Risk of Bias - 3 elements: unpleasantness, depression, Number of Patient 4.34); p>.05. chaplains, nutritionists, recreational Inpatient consultation or overall symptom distress. Participants and Decedents: Symptom distress (distress therapists, occupational therapists, clinical psychiatrists. Screened - Not described associated with nausea, appetite change, insomnia, Intervention Arm - Palliative care Intervention Enrolled - 152 pain, fatigue, bowel changes, team; physicians, advanced Duration: Completed - 50 concentration, appearance, practice nurses, social workers, worry, breathing, and cough; Not provided Decedents - Not described chaplains, nutritionists, recreational 13-65 scale, lower is better): therapists, counselors, control, mean 22.19 (SD acupuncturists, acupressurists, 6.62); intervention, mean massage therapists, Reiki, 22.42 (SD 6.20); p>.05. rehabilitation medicine.

Patient Functional Status: None described

Caregiver Symptoms: None described

165

Evidence Table 2: Outcomes

Wallen et al., Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 51 2012 US Pain - GPS pain intensity and Outcomes Study Arms - 2 unpleasantness subscales Comparing inpatient palliative Level of Depression - CES-D care with usual care for post- operative patients with Randomization - Other Symptoms - SDS Patients advanced cancer, the study did not demonstrate any Study Duration - 12 significant changes in pain Caregiver Outcomes: months intensity, pain Risk of Bias - 3 None described unpleasantness, depression, or overall symptom distress.

166

Evidence Table 1: Populations and Interventions

Pirl et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 59 2012 US Criteria: Cancer (NSCLC) Major depression syndrome Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for (determined by PHQ-9 score): Intervention Arm - Outpatient clinics Comparing early palliative inclusion if they had confirmed all participants, 14%. care with usual care for Level of Prognosis: Intervention Arm: Randomization - metastatic NSCLC, an ECOG patients with newly diagnosed Patients performance status of 0-2, Not described Multidisciplinarity and Clinical Palliative care team NSCLC, the study and the ability to read and Patient Functional Status: demonstrated a significant Disciplines: Study Duration - 12 respond to questions in improvement in the ECOG performance status (0- Control Arm - No multidisciplinary weeks English. They were excluded if Palliative care team intervention group in 5 score, lower is better): all team; clinical disciplines not they were already receiving elements: depression. Risk of Bias - 4 participants, 0, 37.1%; 1, specified. palliative care. 53.6%; 2, 9.3%. Outpatient Intervention Arm - Palliative care consultation team; physicians, advanced Health Care Use

Number of Patient Caregiver Symptoms: practice nurses. The study did not demonstrate Participants and Decedents: None described Intervention a significant change in mental Screened - Not described Duration: health visits. Enrolled - 151 Not provided Completed - 104 Drugs Decedents - 27 The study did not demonstrate a significant change in antidepressant use.

167

Evidence Table 2: Outcomes

Pirl et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Mental health None Patient QOL-Relevant 59 2012 US Depression - PHQ-9 visits Outcomes Study Arms - 2 Depression: Drugs - Antidepressant use Comparing early palliative Level of care with usual care for Caregiver Outcomes: Depression (0-27 scale, lower is patients with newly diagnosed Randomization - better): 12 weeks vs. baseline, control, Patients None described NSCLC, the study mean +0.06 (SD 4.07); intervention, demonstrated a significant Study Duration - 12 mean -0.96 (SD 4.65); p<0.001. improvement in the weeks intervention group in depression. Risk of Bias - 4 Caregiver Outcomes:

N/A Health Care Use The study did not demonstrate a significant change in mental health visits.

Drugs The study did not demonstrate a significant change in antidepressant use.

168

Evidence Table 1: Populations and Interventions

Jones et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 92 2013 EU Criteria: Cancer (mixed cancer) None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient clinics Comparing a hospice-based Level of inclusion if they were at least residential care facilities rehabilitation intervention with 18 years old; had active, Prognosis: Patient Functional Status: Intervention Arm: usual care for cancer patients, Randomization - Patients progressive, recurrent Not described ECOG performance status (0- Complex the study demonstrated malignant breast or 5 scale, lower is better): 0, Multidisciplinarity and Clinical rehabilitation significant improvements in Study Duration - hematological disease; were control, 5%; intervention, 5%; Disciplines: intervention, including the intervention group in Three months at the end of treatment for first HRQOL, physical and daily 1, control, 45%; intervention, Control Arm - No multidisciplinary clinical assessment, or subsequent recurrence but living needs, psychological Risk of Bias - 2 29%; 2, control, 25%; team; physicians, nurses, social goal setting, referrals, not cured; required symptom needs, and patient care and intervention, 57%; 3, control, workers. weekly meetings, and management; had 25%; intervention, 9%; p>.05. discussion between support needs. The study did rehabilitation needs not Intervention Arm - Hospice care not demonstrate a significant patients and responsive to self- team; physicians, advanced clinicians to review change in psychological management; had Caregiver Symptoms: practice nurses, nurses, social progress distress. psychological, social, workers. None described financial, emotional, and spiritual needs not met by the Intervention Health Care Costs present care; and were able to Duration: The study demonstrated a reach the hospice day-care Three months significant increase in the unit on their own or through intervention group in both hospice-based transport. overall costs and QALYs.

Number of Patient Participants and Decedents: Screened - 81 Enrolled - 41 Completed - 36 Decedents - 2

169

Evidence Table 2: Outcomes

Jones et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 92 2013 EU Functional Status - Supportive Care Outcomes Study Arms - 2 Needs Survey Long Form physical Functional Status: Comparing a hospice-based Level of and daily living subscale rehabilitation intervention with Physical and daily living needs (0-100 usual care for cancer patients, Randomization - HRQOL - EQ-VAS, EQ-5D scale, lower is better): baseline, Patients the study demonstrated Other QOL-Relevant Outcomes - control, mean 44.5 (SD 26.5); significant improvements in Study Duration - Kessler Psychological Distress intervention, mean 50.3 (SD 24.2); the intervention group in Three months Scale, Supportive Care Needs three months, control, mean 47.1 (SD HRQOL, physical and daily Risk of Bias - 2 Survey Long Form psychological and 24.5); intervention, mean 37.0 (SD living needs, psychological patient care and support subscales 23.3); p=.022. needs, and patient care and support needs. The study did not demonstrate a significant Caregiver Outcomes: HRQOL: change in psychological None described HRQOL (0-100 scale, higher is better): distress. baseline, control, mean 58.8 (SD 17.1);

intervention, mean 61.5 (SD 18.0); three months, control, mean 54.2 (SD Health Care Costs 18.2); intervention, mean 67.0 (SD The study demonstrated a 13.4); p=.010. significant increase in the intervention group in both overall costs and QALYs. Other QOL-Relevant Outcomes: Psychological needs (0-100 scale, lower is better): baseline, control, mean 33.5 (SD 21.6); intervention, mean 35.8 (SD 24.9); three months, control, mean 38.8 (SD 21.4); intervention, mean 22.8 (SD 19.9); p=.006. Patient care and support needs (0-100 scale, lower is better): baseline, control, mean 20.9 (SD 9.3); intervention, mean 24.4 (SD 10.8); three months, control, mean 25.4 (SD 10.6); intervention, mean 20.8 (SD 10.8); p=.022.

Caregiver Outcomes: N/A

170

Evidence Table 1: Populations and Interventions

Dyar et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 130 2012 US Criteria: Cancer (mixed cancer) None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Not described Comparing an oncology inclusion if they were at least advanced registered nurse Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - 18 years old, had a diagnosis practitioner consultation Patients of metastatic cancer, had a Not described ECOG performance status (0- Multidisciplinarity and Clinical Consultation with intervention with usual care, treating oncologist who had a 5 scale, lower is better): 0, Disciplines: oncology advanced the study demonstrated a Study Duration - strong expectation of hospice significant improvement in the control, 7.1%; intervention, Control Arm - Oncology team; registered nurse One month referral within 12 months of intervention group in mental 8.3%; 1, control, 71.4%; physicians, disciplines of other practitioner enrollment, were willing to QOL and mixed effects on Risk of Bias - 1 intervention, 58.3%; 2, control, team members not specified. participate in two palliative 14.3%; intervention, 33.3%; 3, emotional well-being. The care consultations with an control, 7.1%; intervention, Intervention Arm - Oncology team Intervention study did not demonstrate advanced registered nurse 0.0%; p=.56. with advanced registered nurse Duration: significant changes in pain practitioner prior to hospice practitioners; physicians, advanced frequency or severity, fatigue One month enrollment and were willing to practice nurses, disciplines of other severity, functional status, undergo visits by an advanced Caregiver Symptoms: team members not specified. spiritual well-being, or overall registered nurse practitioner None described HRQOL. after enrollment, were able to complete questionnaires alone or with assistance, were willing and able to provide informed verbal consent, and had an identifiable primary caregiver. They were excluded if they were planning to enter hospice care with other hospice providers instead of the Community Hospice of Northeast Florida, which was a partner for the study.

Number of Patient Participants and Decedents: Screened - Not described Enrolled - 26 Completed - 18 Decedents - 4

171

Evidence Table 2: Outcomes

Dyar et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 130 2012 US Pain - LASA pain frequency and Outcomes Study Arms - 2 average pain subscales Other QOL-Relevant Outcomes: Comparing an oncology Level of Other Symptoms - LASA average advanced registered nurse Mental QOL (scale not described, practitioner consultation Randomization - fatigue subscale higher is better): one month vs. Patients intervention with usual care, Existential or Spiritual Well-Being - baseline, control, mean -10.0 (SD the study demonstrated a Study Duration - LASA spiritual QOL subscale 16.04); intervention, mean +19.0 (SD significant improvement in the One month Functional Status - FACT-G 32.47); difference, p=.022. intervention group in mental Risk of Bias - 1 functional well-being subscale Emotional well-being (0-24 scale, QOL and mixed effects on emotional well-being. The HRQOL - LASA overall QOL higher is better): one month vs. study did not demonstrate subscale baseline, control, mean -4.5 (SD 4.54); intervention, mean +1.2 (SD 2.94); significant changes in pain Other QOL-Relevant Outcomes - difference, p=.011. frequency or severity, fatigue LASA mental and emotional QOL severity, functional status,

subscales, FACT-G emotional well- spiritual well-being, or overall being subscale Caregiver Outcomes: HRQOL. N/A Caregiver Outcomes: None described

172

Evidence Table 1: Populations and Interventions

Au et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 108 2012 US Criteria: COPD Depression (scale not Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for described): control, 37.2%; Intervention Arm - Outpatient clinics Comparing an end-of-life inclusion if they had COPD as intervention, 37.8%; p=.91. communication intervention for Level of Prognosis: Intervention Arm: Randomization - defined by the GOLD criteria COPD patients using a Providers and identified a participating Not described Multidisciplinarity and Clinical Other care planning, patient-specific feedback form, clinician being primary Patient Functional Status: Disciplines: communication with usual care, the study Study Duration - responsible for their COPD demonstrated a significant None described Control Arm - No multidisciplinary intervention using a Two weeks care. They were excluded if improvement in the team; physicians, nonphysician patient-specific they had cognitive intervention group in quality of Risk of Bias - 4 clinicians. feedback form dysfunction, language Caregiver Symptoms: end-of-life communication and

barriers, or severe psychiatric None described Intervention Arm - No significant increases in disorders. multidisciplinary team; physicians, Intervention discussions with surrogates nonphysician clinicians. Duration: and discussions with clinicians One-time intervention about end-of-life treatment Number of Patient preferences. Participants and Decedents: Screened - 1173 Enrolled - 376 Completed - 306 Decedents - Not described

173

Evidence Table 2: Outcomes

Au et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 108 2012 US Communication or Care Planning - Outcomes Study Arms - 2 Quality of Communication about Communication or Care Planning: Comparing an end-of-life Level of End-of-Life Care Scale, study- communication intervention for specific discussions about treatment Communication about end-of-life care COPD patients using a Randomization - (0-100 scale, higher is better): control, Providers preferences with clinician item, patient-specific feedback form, study-specific discussions with baseline, mean 19.2 (95% CI 15.9- with usual care, the study Study Duration - surrogate item 22.4); two weeks, mean 25.5 (95% CI demonstrated a significant Two weeks 20.4-30.5); intervention, baseline, improvement in the Risk of Bias - 4 mean 23.3 (95% CI 19.9-26.8); two intervention group in quality of Caregiver Outcomes: weeks, mean 34.0 (95% CI 28.5-39.4); end-of-life communication and None described p=.03. significant increases in Discussions with clinicians about discussions with surrogates treatment preferences: control, and discussions with clinicians baseline, mean 17.4% (95% CI 10.5%- about end-of-life treatment 24.3%); two weeks, mean 30.8% (95% preferences. CI 5.3%-56.3%); intervention, baseline, mean 22.3% (95% CI 16.2%-28.4%); two weeks, mean 60.3% (95% CI 43.0%-77.6%); p<.001. Discussions with clinicians about treatment preferences at last clinic visit: two weeks, control, mean 15.9% (95% CI 7.5%-24.4%); intervention, mean 35.2% (95% CI 24.7%-45.7%); p<.001. Discussions with surrogate: control, baseline, mean 62.5% (95% CI 45.9%- 79.2%); two weeks, mean 75.2% (95% CI 58.5%-91.9%); intervention, baseline, mean 62.9% (95% CI 47.2%- 78.6%); two weeks, mean 86.2% (95% CI 74.3%-98.1%); p<.01.

Caregiver Outcomes: N/A

174

Evidence Table 1: Populations and Interventions

Capomolla et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 89 al., 2002 EU Criteria: CHF None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home, hospital Comparing a heart failure inclusion if they had a management program with Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of CHF and were usual care, the study Patients referred to a heart failure unit Not described NYHA class (I-IV scale, lower Multidisciplinarity and Clinical Palliative case demonstrated a significant and heart transplantation is better): I-II, control, 65.6%; Disciplines: management improvement in the Study Duration - program. intervention, 65.2%%; III-IV, intervention group in NYHA One year Control Arm - No multidisciplinary control, 34.4%; intervention, class. team; physicians. Palliative case Risk of Bias - 5 34.8%; p>.05. Number of Patient Intervention Arm - Heart failure management Participants and Decedents: management team; physicians, elements: Health Care Use Screened - Not described Caregiver Symptoms: nurses, physiotherapist, Nurse-led case The study demonstrated a psychologist, dietician, social Enrolled - 234 None described management significant decrease in hospital assistant. readmissions. Completed - 210

Decedents - 24 Intervention Duration: Drugs One year The study demonstrated significant increases in the intervention group dosages of long-acting and short-term ACE inhibitors, potassium- sparing diuretics, nitrates, and β-blockers. The study demonstrated significant decreases in dosages of loop diuretics and digitalis. The study demonstrated a significant increase in the rate of patients taking β-blockers but did not demonstrate significant changes in the rates of patients taking the other medications.

Health Care Costs The study demonstrated a significant decrease in the intervention group in hospital readmission costs and a significant increase in pharmacologic costs.

175

Evidence Table 2: Outcomes

Capomolla et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use Patient QOL-Relevant 89 al., 2002 EU Functional Status - NYHA class hospital Hospital readmission rate over one year: Outcomes control, 35%; intervention, 8%; p<.05. Study Arms - 2 Functional Status: Drugs - Dosages of, and rates of Comparing a heart failure patients taking, ACE inhibitors, Number of hospital readmissions over management program with Level of Caregiver Outcomes: NYHA class (I-IV scale, lower is one year: control (122 patients enrolled), Randomization - potassium-sparing diuretics, 78; intervention (112 patients enrolled), usual care, the study None described better): II vs. IIIV, one year vs. loop diuretics, nitrates, β- 14; p<.05. demonstrated a significant Patients baseline, control, 13% improved, 16% blockers, and digitalis improvement in the Study Duration - worsened, p>.05; intervention, 23% intervention group in NYHA Health Care Costs - Drugs One year improved, 11% worsened, p<.009. class. Pharmacologic costs, Long-acting ACE inhibitors (mg/day): Risk of Bias - 5 rehospitalization costs, overall control, baseline, mean 15 (SD 9); one Caregiver Outcomes: care management costs year, mean 12 (SD 10); intervention, baseline, mean 14 (SD 7); one year, Health Care Use N/A mean 20 (SD 8); p<.05. The study demonstrated a Short-term ACE inhibitors (mg/day): significant decrease in hospital control, baseline, mean 100 (SD 40); readmissions. one year, mean 103 (SD 39); intervention, baseline, mean 101 (SD 31); one year, mean 139 (SD 26); p<.05. Drugs Loop diuretics (mg/day): control, baseline, mean 74 (SD 50); one year, The study demonstrated mean 85 (SD 107); intervention, significant increases in the baseline, mean 70 (SD 57); one year, mean 61 (SD 44); p<.05. intervention group dosages of long-acting and short-term Potassium-sparing diuretics (mg/day): control, baseline, mean 40 (SD 31); one ACE inhibitors, potassium- year, mean 40 (SD 29); intervention, sparing diuretics, nitrates, and baseline, mean 35 (SD 33); one year, β-blockers. The study mean 78 (SD 37); p<.05. demonstrated significant Digitalis (mg/day): control, baseline, decreases in dosages of loop mean 0.140 (SD 0.05); one year, mean diuretics and digitalis. The 0.230 (SD 0.08); intervention, baseline, mean 0.130 (SD 0.07); one year, mean study demonstrated a 0.109 (SD 0.60); p<.05. significant increase in the rate Nitrates (mg/day): control, baseline, of patients taking β-blockers mean 68 (SD 40); one year, mean 40 but did not demonstrate (SD 29); intervention, baseline, mean 60 significant changes in the (SD 34); one year, mean 84 (SD 46); rates of patients taking the p<.05. other medications. β-blockers (mg/day): control, baseline, mean 13 (SD 12); one year, mean 14 (SD 29); intervention, baseline, mean 10 (SD 19); one year, mean 34 (SD 23); Health Care Costs p<.05. The study demonstrated a β-blockers (rate of patients taking): significant decrease in the control, baseline, 39%; one year, 40%; intervention group in hospital intervention, baseline, 40%; one year, 71%; p<.001. readmission costs and a significant increase in

pharmacologic costs. Health Care Costs Hospital readmission costs over one year: control, mean $1,332 (SD $1,992); intervention, mean $268 (SD $801); p<.001. Pharmacologic costs over one year: control, mean $490 (SD $164); intervention, mean $741 (SD $206); p<.001.

176

Evidence Table 1: Populations and Interventions

Doughty et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 75 al., 2002 New Zealand Criteria: CHF, prior myocardial None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for infarction, treated Intervention Arm - Outpatient clinics Comparing a palliative case inclusion if they were admitted hypertension, atrial management intervention for Level of Patient Functional Status: Intervention Arm: Randomization - to the general medical wards fibrillation, COPD stroke, patients with heart failure with Providers at Auckland Hospital with a diabetes NYHA class (I-IV scale, lower Multidisciplinarity and Clinical Palliative case usual care, the study primary diagnosis of heart is better): III, control, 24.0%; Disciplines: management, one- demonstrated a significant Study Duration - 12 failure. They were excluded if improvement in the intervention, 24.7%; IV, Control Arm - No multidisciplinary on-one and group months they had a surgically Prognosis: intervention group in physical control, 76%; intervention, team; physicians. heart failure Risk of Bias - 6 remediable cause of heart Not described 75.3%. education sessions functioning. failure, were considered for Intervention Arm - No

heart transplantation, were multidisciplinary team; physicians, unable to provide informed Caregiver Symptoms: nurses. Palliative ca se Health Care Use consent, had terminal cancer, None described management The study demonstrated or were participating in any elements: significant decreases in all- other clinical trial. Nurse-led case cause and heart failure management hospital readmissions after the first readmission. The study

Number of Patient did not demonstrate significant Participants and Decedents: Intervention changes in first all-cause or Screened - Not described Duration: heart failure readmissions or Enrolled - 197 12 months in hospital days. Completed - 153 Decedents - 43 Drugs The study did not demonstrate significant changes in the number of drug prescriptions or in rates of patients taking or dosages of furosemide, digoxin, or ACE inhibitors.

177

Evidence Table 2: Outcomes

Doughty et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 75 al., 2002 New Zealand Functional Status - MLHFQ physical hospital Number of all-cause hospital Outcomes Study Arms - 2 dimension Functional Status: Drugs - Number of drug readmissions after first Comparing a palliative case Level of prescriptions; rates of patients readmission over one year: management intervention for Physical functioning (0-40, lower is taking and dosages of control (97 patients enrolled), patients with heart failure with Randomization - Caregiver Outcomes: better): baseline, all participants, mean Providers furosemide, digoxin, and ACE 95; intervention (100 patients usual care, the study None described 25.6 (SD 12.4); 12 months vs. inhibitors enrolled), 56; p<.05. demonstrated a significant Study Duration - 12 baseline, control, mean -5.8; improvement in the months intervention, mean -11.1; p=.015. Number of heart failure hospital readmissions after first intervention group in physical Risk of Bias - 6 readmission over one year: functioning. Caregiver Outcomes: control (97 patients enrolled), 42; intervention (100 patients N/A Health Care Use enrolled), 15; p<.05. The study demonstrated significant decreases in all- cause and heart failure hospital readmissions after the first readmission. The study did not demonstrate significant changes in first all-cause or heart failure readmissions or in hospital days.

Drugs The study did not demonstrate significant changes in the number of drug prescriptions or in rates of patients taking or dosages of furosemide, digoxin, or ACE inhibitors.

178

Evidence Table 1: Populations and Interventions

Harrison et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 90 al., 2002 Canada Criteria: CHF None described Control Arm - Home Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home case management Comparing a transitional care Level of inclusion if they had a program designed to improve diagnosis of CHF and were Prognosis: Patient Functional Status: the transfer from hospital to Randomization - Palliative case Patients admitted to a large urban Not described NYHA class (I-IV scale, lower Multidisciplinarity and Clinical home for patients admitted to teaching hospital. They were is better): I, control, 2%; Disciplines: management the hospital with CHF, with Study Duration - 12 excluded if they came from or elements: usual care, the study intervention, 0%; II, control, Control Arm - No multidisciplinary weeks were being discharged to a demonstrated a significant 20%; intervention, 23%; III, team; physicians. Nurse-led case Risk of Bias - 5 long-term care facility; they control, 70%; intervention, management improvement in the lived outside the catchment Intervention Arm - No intervention group in HRQOL. 65%; IV, control, 8%; area for home care; they were intervention, 12%; p=.41. multidisciplinary team; physicians, The study demonstrated too ill, or deceased, shortly nurses. Intervention Arm: mixed effects on physical and MLHFQ physical dimension after admission; their first mental health and emotional (0-40 scale, lower is better): Palliative case language was not French or functioning. control, mean 25.45 (SD management, English; they were discharged 9.77); intervention, mean transitional care within 24 hours; or their program designed to 25.46 (SD 9.55); p=1.00. Health Care Use diagnosis was changed. improve the transfer SF-36 PCS (0-100 scale, from hospital to home The study demonstrated a higher is better): control, mean significant reduction in ED

Number of Patient 28.35 (SD 9.11); intervention, use. The study did not Participants and Decedents: 28.63 (SD 9.46); p=.85. Palliative case demonstrate a significant Screened - 483 management change in hospital admissions. elements: Enrolled - 192 Caregiver Symptoms: Nurse-led case Completed - 157 None described management Decedents - Not described

Intervention Duration: 12 weeks

179

Evidence Table 2: Outcomes

Harrison et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use Patient QOL-Relevant 90 al., 2002 Canada Functional Status - MLHFQ physical hospital, ED ED use over 12 weeks: control, Outcomes Study Arms - 2 domain, SF-36 PCS Functional Status: 46%; intervention, 29%; p=.03. Comparing a transitional care Level of HRQOL - MLHFQ program designed to improve Overall physical health (0-40 scale, the transfer from hospital to Randomization - Other QOL-Relevant Outcomes - lower is better): baseline, control, Patients home for patients admitted to MLHFQ emotional domain, SF-36 mean 25.45 (SD 9.77); intervention, the hospital with CHF, with Study Duration - 12 MCS mean 25.46 (SD 9.55); six weeks, usual care, the study weeks control, mean 21.46 (SD 11.15); demonstrated a significant intervention, mean 16.37 (SD 11.04); improvement in the Risk of Bias - 5 Patient Outcomes: 12 weeks, control, mean 23.21 (SD intervention group in HRQOL. None described 10.66); intervention, mean 14.89 (SD The study demonstrated 11.13); six weeks, p=.01; 12 weeks, mixed effects on physical and p<.001. mental health and emotional functioning. HRQOL: HRQOL (0-105 scale, lower is better): Health Care Use baseline, control, mean 44.56 (SD The study demonstrated a 19.50); intervention, mean 44.82 (SD significant reduction in ED 18.50); six weeks, control, mean 37.54 use. The study did not (SD 20.27); intervention, mean 27.25 demonstrate a significant (SD 19.07); 12 weeks, control, mean change in hospital admissions. 38.39 (SD 18.24); intervention, mean 25.76 (SD 19.44); six weeks, p=.002; 12 weeks, p<.001.

Other QOL-Relevant Outcomes: Emotional functioning (0-25 scale, lower is better): baseline, control, mean 6.98 (SD 5.89); intervention, mean 7.78 (SD 6.38); six weeks, control, mean 7.25 (SD 6.15); intervention, mean 4.52 (SD 5.17) six weeks, p=.006.

Caregiver Outcomes: N/A

180

Evidence Table 1: Populations and Interventions

Hoekstra et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant al., 2006131 EU Criteria: Fatigue: prevalence: control, 81.8%; Outcomes Cancer (mixed cancer) intervention, 84.1%; p>.05; severity (0-10 Control Arm - Home Usual care Study Arms - 2 Patients were eligible for scale, lower is better): control, mean 5.4 Intervention Arm - Home Comparing a symptom inclusion if they had small-cell (SD 2.4); intervention, mean 5.2 (SD 2.6); monitoring intervention, in Level of Prognosis: p>.05. Intervention Arm: Randomization - lung cancer with extensive which cancer patients filled out disease or stage IIIb or IV Pain: prevalence: control, 62.3%; weekly questionnaires, with Patients Identified by clinicians as intervention, 59.4%; p>.05; severity (0-10 Multidisciplinarity and Clinical Symptom monitoring NSCLC, advanced breast poor prognosis scale, lower is better): control, mean 4.6 Disciplines: questionnaires usual care, the study Study Duration - cancer, or colon, stomach, (SD 2.7); intervention, mean 4.4 (SD 2.3); demonstrated significant p>.05. Control Arm - No multidisciplinary Two months pancreatic, or esophagus improvements in the team; clinical disciplines not carcinoma with progressive Anorexia: prevalence: control, 46.8%; Intervention intervention group in the Risk of Bias - 3 intervention, 50.7%; p>.05; severity (0-10 specified. disease (metastases, local scale, lower is better): control, mean 4.9 Duration: constipation and vomiting. The recurrence, etc.); and were in (SD 2.5); intervention, mean 5.6 (SD 2.4); Intervention Arm - No Two months study did not demonstrate the palliative phase of their p>.05. multidisciplinary team; clinical significant changes in the disease (noncurable cancer Dyspnea: prevalence: control, 46.8%; disciplines not specified. fatigue, pain, anorexia, intervention, 47.8%; p>.05; severity (0-10 with life expectancy of 1-12 scale, lower is better): control, mean 3.9 dyspnea, coughing, months as judged by their (SD 2.1); intervention, mean 4.2 (SD 2.1); sleeplessness, nausea, or physician). They were p>.05. diarrhea. excluded if they were younger Coughing: prevalence: control, 37.7%; intervention, 37.7%; p>.05; severity (0-10 than 18, had poor knowledge scale, lower is better): control, mean 4.3 of Dutch, were mentally (SD 2.3); intervention, mean 3.4 (SD 2.2); incompetent to fill out the p>.05. questionnaires, or were too ill Insomnia: prevalence: control, 40.3%; intervention, 31.9%; p>.05; severity (0-10 to be interviewed (as judged scale, lower is better): control, mean 5.0 by the physician) without (SD 2.2); intervention, mean 4.3 (SD 2.1); having a closely related proxy. p>.05. Nausea: prevalence: control, 28.6%; intervention, 29.0%; p>.05; severity (0-10 scale, lower is better): control, mean 3.9 Number of Patient (SD 2.3); intervention, mean 3.8 (SD 2.5); Participants and Decedents: p>.05. Screened - 178 Constipation: prevalence: control, 29.9%; intervention, 24.6%; p>.05; severity (0-10 Enrolled - 159 scale, lower is better): control, mean 5.2 (SD 2.3); intervention, mean 4.1 (SD 2.5); Completed - 99 p>.05. Diarrhea: prevalence: control, 24.7%; Decedents - 28 intervention, 24.6%; p>.05; severity (0-10 scale, lower is better): control, mean 4.7 (SD 2.3); intervention, mean 4.5 (SD 3.0); p>.05. Vomiting: prevalence: control, 7.8%; intervention, 10.1%; p>.05; severity (0-10 scale, lower is better): control, mean 4.7 (SD 2.7); intervention, mean 4.7 (SD 2.7); p>.05.

Patient Functional Status: None described

Caregiver Symptoms: None described

181

Evidence Table 2: Outcomes

Hoekstra et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 131 al., 2006 EU Pain - NRS Outcomes Study Arms - 2 Dyspnea - NRS Other Symptoms: Comparing a symptom Level of monitoring intervention, in Other Symptoms - Fatigue, anorexia, Constipation prevalence: baseline, which cancer patients filled out Randomization - nausea, coughing, insomnia, control, 29.9%; intervention, 24.6%; Patients weekly questionnaires, with constipation, diarrhea, and vomiting two months, control, 38.6%; usual care, the study Study Duration - NRSs intervention, 14.3%; p<.05. demonstrated significant Two months Constipation severity (0-10 scale, improvements in the intervention group in the Risk of Bias - 3 Caregiver Outcomes: lower is better): baseline, control, mean 5.2 (SD 2.3), median 5; constipation and vomiting. The None described intervention, mean 4.1 (SD 2.5), study did not demonstrate median 3; two months, control, mean significant changes in the 3.2 (SD 2.2), median 3; intervention, fatigue, pain, anorexia, mean 3.7 (SD 2.0), median 4; test for dyspnea, coughing, medians, p<.05. sleeplessness, nausea, or diarrhea. Vomiting prevalence: baseline, control, 7.8%; intervention, 10.1%; two months, control, 22.8%; intervention, 4.8%; p<.05. Vomiting severity (0-10 scale, lower is better): baseline, control, mean 4.7 (SD 2.7), median 5; intervention, mean 4.7 (SD 2.7), median 5; two months, control, mean 2.9 (SD 2.5), median 2; intervention, mean 4.0 (SD 4.2), median 4; test for medians, p<.05.

Caregiver Outcomes: N/A

182

Evidence Table 1: Populations and Interventions

Jerant et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 23 2001 US Criteria: CHF, hypertension None described Control Arm - Not described Usual care Outcomes Study Arms - 3 Patients were eligible for Intervention Arm 1 - Home Comparing a home telecare inclusion if they were admitted intervention, a home Level of Prognosis: Patient Functional Status: Intervention Arm 2 - Home Intervention Arm 1: Randomization - to the UCD Hospital with a telephone intervention, and Patients primary diagnosis of CHF, Not described NYHA class (I-IV scale, lower Palliative case usual care, the study did not COPD exacerbation, dyspnea, demonstrate any significant is better): II, telecare, 69%; Multidisciplinarity and Clinical management, home Study Duration - or edema; were at least 40 differences in HRQOL, overall group telephone, 67%; usual Disciplines: telephone support 180 days years old; had an active physical health, or patient care, 58%; III, telecare, 23%; Risk of Bias - 7 telephone line in their home group telephone, 33%; usual Control Arm - No multidisciplinary satisfaction. team; physicians, nurses. Palliative case and had the vision and care, 42%; IV, telecare, 8%; hearing adequate to use a group telephone, 0%; usual Intervention Arm 1 - No management Health Care Use telephone or telecare care, 0%; p=.69. multidisciplinary team; nurses. elements: equipment (or had a caregiver SF-36 PCS (0-100 scale, Nurse-led case The study demonstrated a with adequate vision and Intervention Arm 2 - No higher is better): telecare, management significant decrease in the hearing); were English- multidisciplinary team; nurses. mean 30.5 (SD 11.4); group telecare and group telephone speaking; and had a primary telephone, mean 30.3 (SD groups in CHF-related ED care provider in the UCD Intervention Arm 2: 12.6); usual care, mean 31.1 visits. The study did not Health System. They were (SD 7.3); p=.98. Palliative case demonstrate significant excluded if they had a CCI of management, home changes in non-CHF-related at least 6, a GDS score of at telecare support ED visits, all-cause ED visits, least 7, an MMSE score of 20 Caregiver Symptoms: CHF-related hospital or less, or a score more than admissions, non-CHF-related None described two standard deviations below Palliative case hospital admissions, or all- the age- and education- management cause hospital admissions. adjusted mean SDMT; the elements: latter three conditions excluded patients who would Nurse-led case Health Care Costs management generally not be able to The study demonstrated a adequately participate in a significant decrease in the home telecare encounter. Intervention telecare and group telephone Duration: groups in CHF-related ED costs. The study did not Number of Patient 60 days demonstrate significant Participants and Decedents: changes in non-CHF-related Screened - 740 ED costs, all-cause ED costs, Enrolled - 37 CHF-related hospital admission costs, non-CHF- Completed - 36 related hospital admission Decedents - 2 costs, all-cause hospital admission costs, or total care costs.

183

Evidence Table 2: Outcomes

Jerant et al., Study Location - Patient Outcomes: None Health Care Use - Inpatient Health Care Use Patient QOL-Relevant 23 2001 US Experience or Satisfaction - CSQ hospital, ED CHF-related ED visits over 180 Outcomes Study Arms - 3 Functional Status - SF-36 PCS Health Care Costs - Total care days: telecare, mean 0.1 (SD Comparing a home telecare costs, ED costs, hospitalization 0.3); group telephone, mean 0.2 intervention, a home Level of HRQOL - MLHFQ Randomization - costs (SD 0.4); usual care, mean 0.7 telephone intervention, and Patients SD (0.9); p=.034. usual care, the study did not demonstrate any significant Caregiver Outcomes: Study Duration - differences in HRQOL, overall 180 days None described Health Care Costs physical health, or patient Risk of Bias - 7 CHF-related ED costs over 180 satisfaction. days: telecare, mean $399 (SD $1,438); group telephone, mean $1,036 (SD $2,387); usual care, Health Care Use mean 2,882 SD ($4,166); The study demonstrated a p=.049. significant decrease in the telecare and group telephone groups in CHF-related ED visits. The study did not demonstrate significant changes in non-CHF-related ED visits, all-cause ED visits, CHF-related hospital admissions, non-CHF-related hospital admissions, or all- cause hospital admissions.

Health Care Costs The study demonstrated a significant decrease in the telecare and group telephone groups in CHF-related ED costs. The study did not demonstrate significant changes in non-CHF-related ED costs, all-cause ED costs, CHF-related hospital admission costs, non-CHF- related hospital admission costs, all-cause hospital admission costs, or total care costs.

184

Evidence Table 1: Populations and Interventions

Kasper et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 91 2002 US Criteria: CHF, hypertension, None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for diabetes Intervention Arm - Outpatient Comparing a multidisciplinary Level of inclusion if they were admitted clinics, home outpatient management to the hospital with a primary Patient Functional Status: Intervention Arm: program with usual care for Randomization - Prognosis: Patients diagnosis of NYHA class III or NYHA class (I-IV scale, lower Palliative case patients with CHF, the study IV CHF and judged to be at Not described is better): II, control, 33.7%; Multidisciplinarity and Clinical management demonstrated significant Study Duration - high risk of CHF readmission. improvements in the intervention, 37.3%; III, Disciplines: Six months They were excluded if they intervention group in HRQOL, control, 61.2%; intervention, Control Arm - No multidisciplinary had valvular heart disease Palliative case physical functioning, and Risk of Bias - 9 55.9%. team; physicians. requiring surgical correction, management patient-assessed MLHFQ physical domain (0-40 active substance abuse, Intervention Arm - Team to prevent elements: symptomatology. scale, lower is better): control, peripartum cardiomyopathy, hospital readmission; physicians, median 35 (IQR 27-38); Nurse-led case hypertrophic cardiomyopathy nurses. intervention, median 34 (IQR management with left ventricular outflow Health Care Use 26-38). tract obstruction, restrictive The study demonstrated cardiomyopathy, constrictive Duke Activity Status score (0- Intervention significant decreases in all- pericarditis, psychiatric 12 scale, higher is better): Duration: cause and CHF-related disease or dementia likely to control, median 4.5 (IQR 3-7); Six months hospital admissions. The limit compliance, concurrent intervention, median 5 (IQR 3- study did not demonstrate noncardiac illness likely to 7). significant changes in inpatient cause repeat hospital or outpatient visits or inpatient admissions, heart length of stay. transplantation likely to occur Caregiver Symptoms:

within six months, uncorrected None described thyroid disease, serum Drugs creatinine at least 3.0 mg/dL, The study demonstrated a long-term intravenous significant increase in the inotropic therapy at home, intervention group in the cardiac surgery or myocardial percent of patients with infarction during the index systolic dysfunction prescribed admission, active participation target doses of vasodilators. in another research trial, The study did not demonstrate unwillingness to provide significant changes in the informed consent, or percent of all patients taking residence in a nursing home ACE inhibitors, ARBs, β- or rehabilitation facility outside blockers, digoxin, diuretics, the catchment area. hydralazine, or long-acting nitrates. Number of Patient Participants and Decedents: Health Care Costs Screened - 1452 The study did not demonstrate Enrolled - 200 significant changes in total inpatient costs or outpatient Completed - Not described pharmacy costs. Decedents - 20

185

Evidence Table 2: Outcomes

Kasper et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use Patient QOL-Relevant 91 2002 US Other Symptoms - Patient-assessed hospital, outpatient visits Number of all-cause hospital Outcomes Study Arms - 2 symptoms Other Symptoms: Drugs - Percent of patients with admissions over six months: Comparing a multidisciplinary systolic dysfunction prescribed control (98 patients enrolled), outpatient management Level of Functional Status - NYHA class, Patient-assessed symptoms: six Randomization - MLHFQ physical dimension, Duke target doses of vasodilators; 96; intervention (102 patients program with usual care for months, stable, control, 35%; percent of all patients taking enrolled), 77; p=.04. patients with CHF, the study Patients Activity Status intervention, 44%; improved, control, ACE inhibitors, ARBs, β- Number of CHF-related hospital demonstrated significant Study Duration - HRQOL - MLHFQ 31%; intervention, 43%; worse, control, blockers, digoxin, diuretics, improvements in the Six months 35%; intervention, 12%; missing, admissions over six months: hydralazine, and long-acting control (98 patients enrolled), intervention group in HRQOL, control, 1%; intervention, 2%; p=.003. nitrates physical functioning, and Risk of Bias - 9 Caregiver Outcomes: 59; intervention (102 patients Health Care Costs - Total enrolled), 43; p=.03. patient-assessed None described symptomatology. Functional Status: inpatient costs, outpatient pharmacy costs NYHA class (I-IV scale, lower is Drugs better): six months, I, control, 11%; Health Care Use Percent of patients with systolic intervention, 8%; II, control, 40%; The study demonstrated intervention, 67%; III, control, 41%; dysfunction prescribed target doses of vasodilators: six significant decreases in all- intervention, 20%; IV, control, 8%; cause and CHF-related intervention, 5%; missing, control, 0%; months, control, 60.6%; intervention, 92.5%; p<.001. hospital admissions. The intervention, 2%; p=.03. study did not demonstrate Physical functioning (0-40 scale, lower significant changes in inpatient is better): baseline, control, median 35 or outpatient visits or inpatient (IQR 27-38); intervention, median 34 length of stay. (IQR 26-38); six months, control, median 25 (IQR 11-32); intervention, 14 (IQR 5-21); p<.001. Drugs Physical functioning (0-12 scale, higher The study demonstrated a is better): six months, control, median significant increase in the 6 (IQR 4-8); intervention, median 6 intervention group in the (IQR 5-9); p=.05. percent of patients with systolic dysfunction prescribed target doses of vasodilators. HRQOL: The study did not demonstrate significant changes in the HRQOL (0-105 scale, lower is better): percent of all patients taking baseline, control, median 65.5 (IQR ACE inhibitors, ARBs, β- 52-78); intervention, median 68.5 (IQR blockers, digoxin, diuretics, 51-79); six months, control, median 51 hydralazine, or long-acting (IQR 22-64); intervention, 33 (IQR 14- nitrates. 52); p=.001.

Health Care Costs Caregiver Outcomes: The study did not demonstrate N/A significant changes in total inpatient costs or outpatient pharmacy costs.

186

Evidence Table 1: Populations and Interventions

Laramee et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 54 al., 2003 US Criteria: CHF, prior myocardial None described Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for infarction, myocardial Intervention Arm - Home, hospital Comparing a nursing case inclusion if they had CHF as infarction at current management intervention for Level of Patient Functional Status: Intervention Arm: Randomization - defined by either moderate-to- hospital admission, patients hospitalized with CHF Patients severe left ventricular COPD, diabetes, NYHA class (I-IV scale, lower Multidisciplinarity and Clinical Palliative case and then discharged, with dysfunction or radiographic peripheral vascular is better): I, control, 26%; Disciplines: management usual care, the study Study Duration - 12 evidence of pulmonary disease demonstrated significant intervention, 7%; II, control, Control Arm - No multidisciplinary weeks congestion and symptomatic improvements in the 36%; intervention, 55%; III, team; physicians, nurses, social improvement following Palliative case intervention group in overall Risk of Bias - 3 control, 35%; intervention, workers, physical therapist, diuresis; and were at risk for Prognosis: management satisfaction with care and 36%; IV, control, 3%; occupational therapist, dietary early readmission as defined elements: satisfaction with hospital care, Not described intervention, 2%; mean, services. by one or more of history of control, 2.2 (SD 0.8); Nurse-led case hospital discharge, care CHF, documented knowledge intervention, 2.5 (SD 0.6); Intervention Arm - No management instructions, and recovering at deficits of treatment plan or p<.001. multidisciplinary team; physicians, home. disease process, potential or nurses, social workers, physical

ongoing lack of adherence to therapist, occupational therapist, Intervention treatment plan, previous CHF Caregiver Symptoms: dietary services. Duration: Health Care Use hospital admission, living None described 12 weeks The study did not demonstrate alone, and four or more significant changes in rates of hospitalizations in the past five all-cause or CHF hospital years. They were excluded if readmissions. they were discharged to a

long-term care facility, had planned cardiac surgery, had Drugs cognitive impairment, had an The study did not demonstrate anticipated survival of less significant changes in rates of than three months, or were ACE inhibitor, ARB, or β- undergoing long-term blocker usage, or rates of hemodialysis. adherence to target doses of those medications. Number of Patient Participants and Decedents: Health Care Costs Screened - 589 The study did not demonstrate Enrolled - 287 significant changes in initial hospital admission costs, total Completed - 234 readmission costs, total Decedents - 28 inpatient costs, total outpatient costs, or total overall costs.

187

Evidence Table 2: Outcomes

Laramee et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient None Patient QOL-Relevant 54 al., 2003 US Experience or Satisfaction - Study- hospital Outcomes Study Arms - 2 specific patient satisfaction scale Experience or Satisfaction: Drugs - Rates of usage of, and Comparing a nursing case rates of adherence to target management intervention for Level of Satisfaction with hospital care (1-5 Randomization - doses of, ACE inhibitors, ARBs, patients hospitalized with CHF Caregiver Outcomes: scale, higher is better): four weeks, and β-blockers and then discharged, with Patients control, mean 4.0; intervention, mean None described Health Care Costs - Total overall usual care, the study Study Duration - 12 4.2; p=.003. demonstrated significant weeks costs, initial hospital admission Satisfaction with hospital discharge (1- costs, total readmission costs, improvements in the Risk of Bias - 3 5 scale, higher is better): four weeks, total inpatient costs, total intervention group in overall control, mean 4.0; intervention, mean outpatient costs satisfaction with care and 4.3; p<.001. satisfaction with hospital care, hospital discharge, care Satisfaction with care instructions (1-5 instructions, and recovering at scale, higher is better): four weeks, home. control, mean 3.4; intervention, mean 4.0; p<.001. Satisfaction with recovering at home Health Care Use (1-5 scale, higher is better): four The study did not demonstrate weeks, control, mean 3.9; intervention, significant changes in rates of mean 4.4; p<.001. all-cause or CHF hospital Overall satisfaction (1-5 scale, higher readmissions. is better): four weeks, control, mean 3.8; intervention, mean 4.2; p<.001. Drugs

The study did not demonstrate Caregiver Outcomes: significant changes in rates of N/A ACE inhibitor, ARB, or β- blocker usage, or rates of adherence to target doses of those medications.

Health Care Costs The study did not demonstrate significant changes in initial hospital admission costs, total readmission costs, total inpatient costs, total outpatient costs, or total overall costs.

188

Evidence Table 1: Populations and Interventions

McLachlan Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 98 et al., 2001 Australia Criteria: Cancer (mixed cancer) Depression (BDI, 0-39 scale, Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for higher is worse): none or Intervention Arm - Outpatient clinics Comparing an intervention in inclusion if they had a minimal (0-4), control, 67%; which patients used a touch Level of Prognosis: Intervention Arm: Randomization - diagnosis of lung, head and intervention, 61%; mild (5-7), screen to complete a needs Patients neck, genitourinary, skin, or Not described control, 21%; intervention, Multidisciplinarity and Clinical Palliative case assessment and a care other cancers managed in the 24%; moderate (8-15), control, Disciplines: management coordinator nurse made Study Duration - Peter MacCallum Cancer 11%; intervention, 12%; referrals to have needs met, Control Arm - No multidisciplinary Six months Institute; were not attending severe (16-39), control, 1%; with usual care, the study team; clinical disciplines not Risk of Bias - 2 for their first consultation; had intervention, 3%. Palliative case demonstrated a significant specified. management adequate English proficiency; improvement in the had ECOG performance Intervention Arm - Health care team elements: intervention group in Patient Functional Status: status of at most 2; were at that coordinated targeted Nurse-led case depression at six months for least 18 years old; had ECOG performance status (0- psychosocial interventions; management patients who were moderately adequate follow-up scheduled 5 scale, lower is better): 0, physicians, nurses. or severely depressed at

at the institute; gave written control, 36%; intervention, baseline. The study did not informed consent; and 41%; 1, control, 53%; Intervention demonstrate significant completed at least 90% of intervention, 48%; 2, control, Duration: changes in depression for all questionnaire items on patients, psychologic needs, 11%; intervention, 11%. Six months prestudy assessment. They spiritual or religious needs,

were excluded if they were physical functioning, emotional receiving treatment for a major Caregiver Symptoms: functioning, HRQOL, or psychiatric or cognitive None described patient satisfaction. disorder or had breast cancer (there were competing QOL studies for these patients).

Number of Patient Participants and Decedents: Screened - 3520 Enrolled - 450 Completed - 318 Decedents - Not described

189

Evidence Table 2: Outcomes

McLachlan Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 98 et al., 2001 Australia Depression - BDI Outcomes Study Arms - 2 Other Symptoms - CNQ psychologic Depression: Comparing an intervention in Level of needs domain which patients used a touch Depression (0-39 scale, lower is screen to complete a needs Randomization - Existential or Spiritual Well-Being - better), subset of patients who had a Patients assessment and a care Study-specific spirituality/religious score of 8 or above at baseline: six coordinator nurse made Study Duration - needs scale months vs. baseline, intervention mean referrals to have needs met, Six months Experience or Satisfaction - Study- change vs. control mean change, -5.1 with usual care, the study Risk of Bias - 2 specific patient satisfaction scale (95% CI 2.3-7.9), p=.001. demonstrated a significant improvement in the Functional Status - CNQ physical intervention group in and daily living needs domain, Caregiver Outcomes: depression at six months for EORTC QLQ-C30 physical N/A patients who were moderately functioning domain or severely depressed at HRQOL - EORTC QLQ-C30 global baseline. The study did not health status/QOL domain demonstrate significant Other QOL-Relevant Outcomes - changes in depression for all EORTC QLQ-C30 emotional patients, psychologic needs, functioning domain spiritual or religious needs, physical functioning, emotional functioning, HRQOL, or Caregiver Outcomes: patient satisfaction. None described

190

Evidence Table 1: Populations and Interventions

Riegel et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 24 2002 US Criteria: CHF, hypertension, None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for coronary artery disease, Intervention Arm - Outpatient Comparing a standardized Level of inclusion if they had a COPD, stroke, diabetes, clinics, home telephonic case management confirmed clinical diagnosis of renal disease without Patient Functional Status: Intervention Arm: intervention for patients with Randomization - Multidisciplinarity and Clinical Providers heart failure as the primary dialysis, peripheral NYHA class (I-IV scale, lower Palliative case heart failure, with usual care, reason for their hospital visit vascular disease, thyroid is better): II, control, 3.6%; Disciplines: management the study demonstrated a Study Duration - and spoke either English or disease intervention, 2.3%; III, control, Control Arm - No multidisciplinary significant increase in the Six months Spanish. They were excluded 38.4%; intervention, 35.9%; team; physicians. intervention group in patient if they had cognitive Palliative case satisfaction. Risk of Bias - 8 IV, control, 58.0%; Intervention Arm 1 - No impairment or psychiatric Prognosis: management intervention, 61.7%. multidisciplinary team; nurses. illness, severe renal failure elements: Not described SAS class (I-IV scale, lower is requiring dialysis, or terminal Health Care Use better): I, control, 8.0%; Nurse-led case disease; were discharged to a intervention, 11.5%; II, control, management The study demonstrated long-term care facility; or were 17.0%; intervention, 19.2%; significant decreases in the previously enrolled in a heart III, control, 59.8%; intervention group in heart failure disease management Intervention intervention, 53.8%; IV, failure hospital readmissions program. Duration: control, 15.2%; intervention, and heart failure hospital days. 15.4%. Six months The study did not demonstrate significant changes in all- Number of Patient cause hospital admissions, all- Participants and Decedents: Caregiver Symptoms: cause or heart failure hospital Screened - 1145 readmissions, all-cause None described Enrolled - 358 hospital days, time to hospital readmission, multiple Completed - 358 readmissions, ED visits, or Decedents - 48 physician office visits.

Health Care Costs The study did not demonstrate a significant change in inpatient heart failure costs.

191

Evidence Table 2: Outcomes

Riegel et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 24 2002 US Experience or Satisfaction - Study- hospital, ED, outpatient visits Heart failure hospital Outcomes Study Arms - 2 specific patient satisfaction measure Experience or Satisfaction: Health Care Costs - Inpatient readmission rate: three months, Comparing a standardized Level of heart failure costs control, mean 0.31 (SD 0.64); telephonic case management Patient satisfaction (scale not intervention, mean 0.17 (SD intervention for patients with Randomization - Caregiver Outcomes: described, higher is better): six Providers 0.43); p=.03; six months, heart failure, with usual care, None described months, control, mean 21.66 (SD control, mean 0.41 (SD 0.77); the study demonstrated a Study Duration - 3.44); intervention, mean 22.88 (SD intervention, mean 0.21 (SD significant increase in the Six months 2.85); p=.01. 0.50); p=.02. intervention group in patient Risk of Bias - 8 Heart failure hospital days: six satisfaction. Caregiver Outcomes: months, control, mean 2.1 (SD 4.6); intervention, mean 1.1 (SD N/A Health Care Use 3.1); p=.05. The study demonstrated significant decreases in the intervention group in heart failure hospital readmissions and heart failure hospital days. The study did not demonstrate significant changes in all- cause hospital admissions, all- cause or heart failure hospital readmissions, all-cause hospital days, time to hospital readmission, multiple readmissions, ED visits, or physician office visits.

Health Care Costs The study did not demonstrate a significant change in inpatient heart failure costs.

192

Evidence Table 1: Populations and Interventions

Tierney et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 85 al., 2001 US Criteria: Cancer (mixed cancer), None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for CHF, ischemic heart Intervention Arm - Outpatient clinics Comparing an EHR-based inclusion if they had been disease, chronic lung reminder to primary care Level of Patient Functional Status: Intervention Arm: Randomization - active for at least a year in a disease, cerebrovascular physicians to discuss advance Providers hospital-based academic disease, chronic renal None described Multidisciplinarity and Clinical Advance directive directives with scheduled primary care general internal insufficiency, chronic liver patients, with usual care, the Disciplines: completion Study Duration - medicine practice that serves disease study demonstrated unclear Control Arm - No multidisciplinary One year mainly inner-city indigent Caregiver Symptoms: changes in patient satisfaction team; physicians. Risk of Bias - 3 patients and were 75 or older None described Advance directive with primary care physician or between 50 and 74 with Prognosis: Intervention Arm - No completion elements: and patient satisfaction with ischemic heart disease, Not described multidisciplinary team; physicians. Computer-generated outpatient visits. chronic heart failure, COPD, reminders for primary cerebrovascular disease, care physicians to cancer, chronic renal discuss advance insufficiency, or chronic liver directives with failure. They were excluded if scheduled patients they had previously completed an advance directive with their current primary care physician Intervention or did not pass the Pfeiffer Duration: SPMSQ using standard One year scoring criteria.

Number of Patient Participants and Decedents: Screened - 1199 Enrolled - 1042 Completed - 686 Decedents - Not described

193

Evidence Table 2: Outcomes

Tierney et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 85 al., 2001 US Experience or Satisfaction - ABIM Outcomes Study Arms - 2 patient satisfaction instrument, MOS- Comparing an EHR-based Level of VSQ reminder to primary care Randomization - physicians to discuss advance Providers directives with scheduled Caregiver Outcomes: patients, with usual care, the Study Duration - None described study demonstrated unclear One year changes in patient satisfaction Risk of Bias - 3 with primary care physician and patient satisfaction with outpatient visits.

194

Evidence Table 1: Populations and Interventions

Velikova et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 86 al., 2004 EU Criteria: Cancer (mixed cancer) Number of symptoms (scale Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 3 Patients were eligible for not described, lower is better): Intervention Arm 1 - Outpatient Comparing completion of Level of inclusion if they were control, mean 2.9 (SD 1.80); clinics HRQOL questionnaires by attending the Leeds Cancer Prognosis: questionnaires only, mean 3.2 Intervention Arm 1: cancer patients before clinic Randomization - Intervention Arm 2 - Outpatient Patients Centre Medical Oncology Not described (SD 1.97); questionnaires and Completion of visits with results provided to Clinic at St. James’s Hospital, oncologist feedback, mean 2.8 clinics HRQOL oncologists, with completion of Study Duration - were diagnosed with cancer (SD 1.84). questionnaires before questionnaires only, and usual Six months and were commencing care, the study demonstrated Multidisciplinarity and Clinical clinic visits treatment, were expected to a significant improvement in Risk of Bias - 3 Disciplines: attend the clinic at least three Patient Functional Status: HRQOL in the questionnaires times, were fluent in English, ECOG performance status (0- Control Arm - No multidisciplinary Intervention Arm 2: and feedback group as were not taking part in 5 scale, lower is better): team; physicians. Completion of compared with the HRQOL studies, and did not control, 0, 19%; 1, 46%; 2, Intervention Arm 1 - No HRQOL questionnaires only and usual exhibit overt psychopathology. 26%; 3, 9%; questionnaires multidisciplinary team; physicians. questionnaires before care groups. The study did not only, 0, 9%; 1, 36%; 2, 43%; demonstrate a significant Intervention Arm 2 - No clinic visits, with 3, 13%; questionnaires and results provided to change in HRQOL in the Number of Patient multidisciplinary team; physicians. oncologist feedback, 0, 14%; oncologists questionnaires only group as Participants and Decedents: 1, 44%; 2, 35%; 3, 7%. compared with the usual care group. Screened - 439 Intervention Enrolled - 286 Caregiver Symptoms: Duration: Completed - 164 None described Six months Decedents - 95

195

Evidence Table 2: Outcomes

Velikova et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 86 al., 2004 EU HRQOL - FACT-G Outcomes Study Arms - 3 HRQOL: Comparing completion of Level of HRQOL questionnaires by Caregiver Outcomes: QOL (change in FACT-G scores): six cancer patients before clinic Randomization - months vs. baseline, control, Patients None described visits with results provided to improvement, 24%; no change, 45%; oncologists, with completion of Study Duration - deterioration, 31%; questionnaires questionnaires only, and usual Six months only, improvement, 32%; no change, care, the study demonstrated Risk of Bias - 3 61%; deterioration, 7%; questionnaires a significant improvement in and oncologist feedback, improvement, HRQOL in the questionnaires 40%; no change, 46%; deterioration, and feedback group as 14%; questionnaires and oncologist compared with the feedback vs. questionnaires only and questionnaires only and usual control, p=.001; questionnaires and care groups. The study did not oncologist feedback and demonstrate a significant questionnaires only vs. control, p=.003. change in HRQOL in the questionnaires only group as compared with the usual care Caregiver Outcomes: group. N/A

196

Evidence Table 1: Populations and Interventions

Allen et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 88 2008 US Criteria: Cancer (type not Depression (0-60 scale, lower is Control Arm - Home Supportive telephone Outcomes Study Arms - 2 Patients were eligible for described), heart disease, better): control, mean 13.92 (SD calls Comparing a legacy project 9.92); intervention, mean 18.80 Intervention Arm - Home Level of inclusion if they were at least ESRD, dementia, stroke, intervention with supportive 60 years old, were living in the general debility and (SD 12.06). telephone calls for community- Randomization - Intervention Arm: Patients and community, had a life-limiting multiple chronic illnesses Pain (0-10 scale, lower is Multidisciplinarity and Clinical dwelling patients, the study better): control, mean 1.93 (SD families/caregivers illness or combination of Disciplines: Legacy project demonstrated significant chronic illnesses, and were 1.86); intervention, mean 2.80 improvements in the legacy only Control Arm - No multidisciplinary intervention receiving assistance in ADLs Prognosis: (SD 2.32). project group in religious team; clinical disciplines not Study Duration - 10 or IADLs from a family meaning and dyspnea. The Not described Fatigue (0-10 scale, lower is specified. weeks caregiver. better): control, mean 1.79 (SD Intervention study did not demonstrate Risk of Bias - 1 1.85); intervention, mean 3.87 Intervention Arm - No Duration: significant changes in Caregivers were eligible for (SD 2.98). multidisciplinary team; clinical 10 weeks subjective well-being, daily inclusion if they had weekly Nausea (0-10 scale, lower is disciplines not specified. spiritual experiences, overall face-to-face contact with the better): control, mean 1.36 (SD symptom burden, depression, patient to provide assistance 1.34); intervention, mean 1.80 pain, fatigue, nausea, with ADL or IADL needs, were (SD 2.20). drowsiness, or anorexia. cognitively intact as measured Drowsiness (0-10 scale, lower is by a score of 24 or greater on better): control, mean 1.50 (SD Caregiver QOL-Relevant the MMSE, and were not 1.29); intervention, mean 2.79 currently participating in (SD 2.88). Outcomes another caregiver intervention Anorexia (0-10 scale, lower is The study demonstrated a study. better): control, mean 1.86 (SD significant improvement in the 1.88); intervention, mean 1.47 legacy project group in (SD 1.00). caregiver stress. The study did Number of Patient not demonstrate significant Participants and Decedents: Dyspnea (0-10 scale, lower is better): control, mean 3.00 (SD changes in caregiver Screened - 134 2.6); intervention, mean 3.00 subjective well-being, religious Enrolled - 42 (SD 2.19). meaning, daily spiritual experiences, or depression. Completed - 31 Symptom burden (comprised of pain, fatigue, nausea, Decedents - Not described drowsiness, anorexia, and dyspnea) (0-60 scale, lower is better): control, mean 11.43 (SD Number of Caregiver 5.88); intervention, mean 15.82 Participants: (SD 8.46). Screened - 134 Enrolled - 42 Patient Functional Status: Completed - 31 Katz ADL scale (0-6 scale, lower is better): control, mean 1.79 (SD 0.43); intervention, mean 1.59 (SD 0.80).

Caregiver Symptoms: Depression (0-60 scale, lower is better): control, mean 8.08 (SD 7.78); intervention, mean 13.45 (SD 8.38).

197

Evidence Table 2: Outcomes

Allen et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 88 2008 US Pain - ESAS item Outcomes Study Arms - 2 Dyspnea - ESAS item Dyspnea: Comparing a legacy project Level of intervention with supportive Depression - CES-D Dyspnea (0-10 scale, lower is better): telephone calls for community- Randomization - baseline, control, mean 3.00 (SD 2.60); Patients and Other Symptoms - Symptom burden dwelling patients, the study ESAS scale; fatigue, nausea, intervention, mean 3.00 (SD 2.19); 10 demonstrated significant families/caregivers weeks, control, mean 3.50 (SD 3.06); only drowsiness, and anorexia ESAS improvements in the legacy items intervention, mean 1.43 (SD 0.85); project group in religious Study Duration - 10 group by time p=.026. meaning and dyspnea. The Existential or Spiritual Well-Being - weeks study did not demonstrate BMMRS religious meaning and daily significant changes in Risk of Bias - 1 spiritual experiences subscales Existential or Spiritual Well-Being: subjective well-being, daily Other QOL-Relevant Outcomes - Religious meaning (2-8 scale, higher is spiritual experiences, overall Study-specific subjective well-being better): baseline, control, mean 6.93 symptom burden, depression, scale (SD 0.92); intervention, mean 6.64 (SD pain, fatigue, nausea, 1.58); 10 weeks, control, mean 6.43 drowsiness, or anorexia. (SD 1.28); intervention, mean 7.21 (SD

Caregiver Outcomes: 0.80); group by time p=.032. Caregiver QOL-Relevant Depression - CES-D Outcomes Existential or Spiritual Well-Being - Caregiver Outcomes: BMMRS religious meaning subscale The study demonstrated a significant improvement in the Other QOL-Relevant Outcomes - Other QOL-Relevant Outcomes: legacy project group in Caregiver Stressors Scale-Revised, caregiver stress. The study did study-specific subjective well-being Caregiver stress (scale not described, not demonstrate significant scale lower is better): baseline, control, changes in caregiver mean 43.64 (SD 7.13); intervention, subjective well-being, religious mean 49.65 (SD 7.35); 10 weeks, meaning, daily spiritual control, mean 46.57 (SD 7.10); experiences, or depression. intervention, mean 48.94 (SD 6.82); group by time p=.034.

198

Evidence Table 1: Populations and Interventions

Belle et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 71 Criteria: 2006 US Dementia None described Control Arm - Home Dementia and Outcomes Patients were eligible for Study Arms - 2 Intervention Arm - Home caregiving education Comparing a multicomponent inclusion if they had diagnosed information caregiver support, skills Level of Alzheimer’s disease or related Prognosis: Patient Functional Status: Randomization - training, and problem solving disorders. They were excluded if intervention with dementia and Patients and they were bedbound with an Not described Katz ADL scale (0-6 scale, Multidisciplinarity and Clinical Intervention Arm: caregiving education families/caregivers MMSE score of 0. lower is better): Hispanic or Disciplines: information for caregivers of only Latino, control, mean 3.1 (SD Control Arm - No multidisciplinary Multicomponent 2.2); intervention, mean 3.6 caregiver support, patients with dementia, the Study Duration - Caregivers were eligible for team; clinical disciplines not inclusion if they were Hispanic (SD 2.0); white or Caucasian, skills training, and study demonstrated significant Six months specified. or Latino, white or Caucasian, or control, mean 3.3 (SD 2.0); problem solving improvements in the Risk of Bias - 3 black or African-American; were intervention, mean 3.1 (SD Intervention Arm - No intervention intervention group in caregiver at least 21 years old; were living multidisciplinary team; clinical clinical depression for all 2.0); black or African- with or sharing cooking facilities American, control, mean 3.3 disciplines not specified. races, caregiver burden for with a relative with diagnosed (SD 2.0); intervention, mean Intervention black or African-American Alzheimer’s disease or related 3.3 (SD 2.0). Duration: caregivers only, social support disorders; were providing care for white or Caucasian Six months for the care recipient for at least caregivers only, and care four hours per day for at least Caregiver Symptoms: recipient problem behaviors the past six months; reported for Hispanic or Latino distress associated with None described caregivers only. The study did caregiving; had a telephone; planned to remain the not demonstrate significant geographic area for at least six changes in caregiver self-care months; and had competency in for any race or in the either English or Spanish. They aforementioned measures for were excluded if they were caregivers of different races involved in another caregiver and ethnicities. intervention study, had participated in REACH I, or had an illness that would prevent 6 Health Care Use months of study participation. The study did not demonstrate a significant change in rates of Number of Patient institutionalization. Participants and Decedents: Screened - 995 Enrolled - 642 Completed - 518 Decedents - Not described

Number of Caregiver Participants: Screened - 995 Enrolled - 642 Completed - 518

199

Evidence Table 2: Outcomes

Belle et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - None Caregiver QOL-Relevant 71 2006 US None described N/A Institutionalization Outcomes Study Arms - 2 Comparing a multicomponent caregiver support, skills Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - training, and problem solving Patients and Depression - CES-D intervention with dementia and caregiving education families/caregivers Other QOL-Relevant Outcomes - Depression: information for caregivers of only Modified ZBI, study-specific Caregiver depression (0-30 scale, patients with dementia, the caregiver self-care, scale, study- Study Duration - lower is better): six months vs. study demonstrated significant specific social support scale, study- Six months baseline, Hispanic or Latino, improvements in the specific care recipient problem intervention vs. control, difference in intervention group in caregiver Risk of Bias - 3 behaviors scale net improvement +39.0% (95% CI clinical depression for all 15.5%-59.6%); p<.05. races, caregiver burden for Caregiver clinical depression (score of black or African-American at least 15, 0-30 scale, lower is better): caregivers only, social support six months, all races and ethnicities, for white or Caucasian control, 22.7%; intervention, 12.6%; caregivers only, and care p=.001. recipient problem behaviors for Hispanic or Latino caregivers only. The study did Other QOL-Relevant Outcomes: not demonstrate significant changes in caregiver self-care Caregiver burden (0-44 scale, lower is for any race or in the better): six months vs. baseline, black aforementioned measures for or African-American, intervention vs. caregivers of different races control, difference in net improvement and ethnicities. +23.1% (95% CI 0.6%-45.7%); p<.05.

Social support (0-30 scale, higher is better): six months vs. baseline, white Health Care Use or Caucasian, intervention vs. control, The study did not demonstrate difference in net improvement +24.1% a significant change in rates of (95% CI 3.0%-46.4%); p<.05. institutionalization. Care recipient problem behaviors (3-15 scale, lower is better): six months vs. baseline, Hispanic or Latino, intervention vs. control, difference in net improvement +36.3% (95% CI 13.2%-56.7%); p<.05.

200

Evidence Table 1: Populations and Interventions

Brown et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 132 2004 Australia Criteria: Cancer (mixed cancer) Depression (0-39 scale, lower Control Arm - Outpatient clinics Informational booklet Outcomes Study Arms - 2 Patients were eligible for is better): control, mean 16.4 Intervention Arm - Outpatient clinics on living with cancer Comparing an informational Level of inclusion if they were female, (SD 3.7); intervention, mean booklet and video on decision- were at least 16 years old, Prognosis: 15.8 (SD 3.0); p>.05. making with an informational Randomization - Intervention Arm: Patients had been recently diagnosed Not described State anxiety (20-80 scale, Multidisciplinarity and Clinical booklet on living with cancer with cancer (any site or stage) lower is better): control, mean Disciplines: Decision support, for female cancer patients, the Study Duration - and were seeing a medical study did not demonstrate any 48.6 (SD 12.2); intervention, Control Arm - No multidisciplinary informational booklet Six months oncologist for the first time to significant changes in mean 40.2 (SD 11.2); p<.01. team; clinical disciplines not on clinical decision- discuss treatment options, depression, state anxiety, Risk of Bias - 3 specified. making and video of were able to speak and read oncologist discussing decisional conflict, satisfaction English, and were not too ill to Patient Functional Status: Intervention Arm - No treatment options with decision-making, or complete the questionnaires. multidisciplinary team; clinical with an actor satisfaction with the None described disciplines not specified. consultation.

Number of Patient Caregiver Symptoms: Decision support Participants and Decedents: elements: None described Screened - 71 Video-based decision Enrolled - 65 aid Completed - Not described Decedents - Not described Intervention Duration: One-time intervention

201

Evidence Table 2: Outcomes

Brown et al., Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 132 2004 Australia Depression - BDI-SF Outcomes Study Arms - 2 Anxiety - STAI Comparing an informational Level of booklet and video on decision- Communication or Care Planning - making with an informational Randomization - O’Connor’s Decisional Conflict Scale Patients booklet on living with cancer Experience or Satisfaction - Patient for female cancer patients, the Study Duration - Satisfaction with Consultation Scale, study did not demonstrate any Six months SWD scale significant changes in depression, state anxiety, Risk of Bias - 3 decisional conflict, satisfaction Caregiver Outcomes: with decision-making, or None described satisfaction with the consultation.

202

Evidence Table 1: Populations and Interventions

Curtis et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 133 2011 US Criteria: Not described None described Control Arm - Hospital Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Hospital Comparing a clinician inclusion if they died in an ICU behavior intervention with Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - or within 30 hours of transfer usual care for ICU patients, Sites to another hospital location. Not described None described Multidisciplinarity and Clinical Clinician behavior the study did not demonstrate They were excluded if they any significant changes in Disciplines: intervention involving Study Duration - had ICU stays shorter than six caregiver-assessed quality of Control Arm - No multidisciplinary clinician education, Four years hours. Caregiver Symptoms: dying, caregiver satisfaction team; clinical disciplines not local champions, with ICU care, caregiver Risk of Bias - 3 None described specified. academic detailing, clinician feedback of satisfaction with decision- Number of Patient Intervention Arm - No quality data, and making, or caregiver total Participants and Decedents: multidisciplinary team; clinical system supports satisfaction with the ICU disciplines not specified. experience. Screened - 2318

Enrolled - 1924 Intervention Completed - 822 Duration: Health Care Use Decedents - 1924 20 months The study did not demonstrate a significant change in ICU

length of stay. Number of Caregiver

Participants: Life-Extending Devices or Screened - 2318 Procedures Enrolled - 1924 The study did not demonstrate Completed - 822 a significant change in the time from ICU admission to withdrawal of mechanical ventilation.

203

Evidence Table 2: Outcomes

Curtis et al., Study Location - Patient Outcomes: None Health Care Use - ICU None Caregiver QOL-Relevant 133 2011 US None described Life-Extending Devices or Outcomes Study Arms - 2 Procedures - Withdrawal of Comparing a clinician mechanical ventilation behavior intervention with Level of Caregiver Outcomes: Randomization - usual care for ICU patients, Sites Experience or Satisfaction - Total the study did not demonstrate satisfaction with ICU experience, any significant changes in Study Duration - satisfaction with ICU care, and caregiver-assessed quality of Four years satisfaction with decision-making FS- dying, caregiver satisfaction Risk of Bias - 3 ICU survey subscales with ICU care, caregiver satisfaction with decision- Other QOL-Relevant Outcomes - making, or caregiver total Caregiver-assessed quality of dying satisfaction with the ICU (QODD Questionnaire) experience.

Health Care Use The study did not demonstrate a significant change in ICU length of stay.

Life-Extending Devices or Procedures The study did not demonstrate a significant change in the time from ICU admission to withdrawal of mechanical ventilation.

204

Evidence Table 1: Populations and Interventions

Egan et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 21 2002 Australia Criteria: COPD None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home, hospital Comparing a nursing case inclusion if they were at least management intervention for Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - 18 years old; had a history of patients hospitalized with Patients chronic bronchitis (with Not described None described Multidisciplinarity and Clinical Palliative case COPD and then discharged, infection), emphysema, with usual care, the study Disciplines: management Study Duration - chronic airway obstruction, demonstrated significant Control Arm - No multidisciplinary Three months chronic asthma, or a Caregiver Symptoms: declines in the intervention team; clinical disciplines not combination of these; had Palliative case group in affectionate social Risk of Bias - 2 None described specified. FEV1 measured on admission management support between baseline and prior to initializing intravenous Intervention Arm - No elements: one month, but not between medications to determine the multidisciplinary team; nurses. Nurse-led case one month and three months; severity of disease; had management and in activities that cause or cognitive function at the are limited by breathlessness function at the time of study between one and three entry adequate to understand Intervention months, but not between and complete a questionnaire; Duration: baseline and one month. The were admitted to a respiratory study did not demonstrate Six weeks unit bed within 72 hours of significant changes in anxiety, admission to the hospital; and depression, respiratory gave informed consent in symptoms, subjective well- writing. being, tangible social support, positive social interaction, emotional/information support, Number of Patient or impacts on social Participants and Decedents: functioning and psychological Screened - 129 disturbances resulting from airway disease. Enrolled - 66 Completed - 53 Decedents - 5

205

Evidence Table 2: Outcomes

Egan et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 21 2002 Australia Dyspnea - SGRQ respiratory Outcomes Study Arms - 2 symptoms domain Experience or Satisfaction: Comparing a nursing case Level of Depression - HADS depression management intervention for Affectionate social support (scale not patients hospitalized with Randomization - subscale described, higher is better): one month Patients COPD and then discharged, Anxiety - HADS anxiety subscale vs. baseline, control, median 0; with usual care, the study Study Duration - Experience or Satisfaction - Social intervention, median -6.7; p=.034. demonstrated significant Three months Support Survey affectionate support declines in the intervention group in affectionate social Risk of Bias - 2 and tangible support subscales Other QOL-Relevant Outcomes: support between baseline and Other QOL-Relevant Outcomes - Activities that cause or are limited by one month, but not between Subjective Well-Being Scale, breathlessness (0-100 scale, lower is one month and three months; impacts on social functioning and better): three months vs. one month, and in activities that cause or psychological disturbances resulting control, median -6.4; intervention, are limited by breathlessness from airway disease (SGRQ impacts median 0; p=.01. between one and three domain), SGRQ activities domain, months, but not between Social Support Survey positive social baseline and one month. The interaction and emotional information Caregiver Outcomes: study did not demonstrate support subscales N/A significant changes in anxiety, depression, respiratory Caregiver Outcomes: symptoms, subjective well- being, tangible social support, None described positive social interaction, emotional/information support, or impacts on social functioning and psychological disturbances resulting from airway disease.

206

Evidence Table 1: Populations and Interventions

Fortinsky et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 134 al., 2009 US Criteria: Dementia None described Control Arm - Not described Usual care, Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home, educational materials Comparing an individualized Level of inclusion if they had a Alzheimer’s association chapter sent to caregivers but dementia education and care physician’s diagnosis of any Prognosis: Patient Functional Status: not discussed consultation intervention for Randomization - Sites type of irreversible dementia, Not described None described family caregivers with usual resided in a home setting care, the study did not Multidisciplinarity and Clinical Study Duration - 12 outside of a nursing home or Disciplines: Intervention Arm: demonstrate any significant months assisted living facility, Caregiver Symptoms: changes in caregiver Control Arm - No multidisciplinary Individualized ambulated without constant depression, physical signs and Risk of Bias - 4 None described team; clinical disciplines not dementia education human assistance, used the symptoms, burden, or self- specified. and care consultation referring physician as their efficacy for symptom

regular source of medical Intervention Arm - No management or accessing care, and had at least one multidisciplinary team; clinical Intervention community support services. disciplines not specified. Duration: identifiable family caregiver. 12 months Caregivers were eligible for Health Care Use inclusion if they were related The study did not demonstrate to eligible patients, had a significant change in nursing primary or shared home admissions. responsibility for patients’ health-related needs, and were known by the referring physician’s office staff.

Number of Caregiver Participants: Screened - 178 Enrolled - 84 Completed - 75

207

Evidence Table 2: Outcomes

Fortinsky et Study Location - Patient Outcomes: None Health Care Use - Nursing home None Caregiver QOL-Relevant 134 al., 2009 US None described Outcomes Study Arms - 2 Comparing an individualized dementia education and care Level of Caregiver Outcomes: Randomization - consultation intervention for Sites Depression - CES-D family caregivers with usual care, the study did not Other Symptoms - Hopkins Study Duration - 12 demonstrate any significant Symptoms Checklist months changes in caregiver Risk of Bias - 4 Other QOL-Relevant Outcomes - depression, physical signs and Revised Caregiver Burden Scale, symptoms, burden, or self- study-specific self-efficacy for efficacy for symptom symptom management measure, management or accessing study-specific self-efficacy for community support services. accessing community support

services measure Health Care Use The study did not demonstrate a significant change in nursing home admissions.

208

Evidence Table 1: Populations and Interventions

Gant et al., Study Location - Inclusion and Exclusion Criteria: Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 65 2007 US Patients were eligible for inclusion Dementia None described Control Arm - Home Dementia educational Outcomes Study Arms - 2 if they had no lifetime history of booklet and biweekly Comparing a video, workbook, schizophrenia, bipolar disorder, Intervention Arm - Home check-in telephone and telephone coaching Level of suicidality, Huntington’s disease, Prognosis: Randomization - Korsakoff’s disease, multiple calls intervention on behavior Patients and sclerosis, HIV, or alcohol abuse; Not described Patient Functional Status: Multidisciplinarity and Clinical strategies for male dementia had an established relationship caregivers, with a dementia families/caregivers Katz ADL scale (0-6 scale, Disciplines: with a clinic or primary care Intervention Arm: educational booklet and only lower is better): all Control Arm - No multidisciplinary physician whom they had visited 10-session video telephone check-in calls, the within the last year; and had a participants, mean 2.3 (SD team; clinical disciplines not Study Duration - series on behavioral study did not demonstrate Not provided diagnosis of dementia confirmed 2.0). specified. by a physician. strategies to manage significant changes in Risk of Bias - 2 Intervention Arm - No challenging caregiver caregiver affect, caregiver Caregivers were eligible for Caregiver Symptoms: multidisciplinary team; situations, stress, caregiver distress inclusion if they were male; were geropsychologists, masters-level accompanying concerning specific patient None described willing to be assigned by chance to gerontology student. workbook, and behaviors, or caregiver self- either group (video vs. instructional efficacy. booklet); were the primary weekly telephone caregiver for a family member coaching sessions diagnosed with a dementing illness; were living in same residence as care recipient; were Intervention between 30 and 85; had no plans Duration: to place the family member in a nursing home or hospice services 12 weeks within the next six months; had visual acuity sufficient to read a book; had the auditory ability to hear voice on the telephone and TV; had no current alcohol abuse or serious suicidal ideation; had no history of suicide attempt; had no current involvement in another caregiver intervention project; had access to telephone, TV, and VCR; had an established relationship with a clinic or primary care physician with whom caregiver had visited within the last year; had at least two care recipient memory or behavior problems that occurred in the past week; and had at least a moderate level of distress (upset or annoyance) following two or more care recipient behaviors.

Number of Caregiver Participants: Screened - 73 Enrolled - 32 Completed - 28

209

Evidence Table 2: Outcomes

Gant et al., Study Location - Patient Outcomes: None None N/A Caregiver QOL-Relevant 65 2007 US None described Outcomes Study Arms - 2 Comparing a video, workbook, and telephone coaching Level of Caregiver Outcomes: Randomization - intervention on behavior Patients and Other QOL-Relevant Outcomes - strategies for male dementia families/caregivers Caregiver stress (Target Complaints caregivers, with a dementia only Interview), distress concerning educational booklet and specific patient behaviors (RMBPC), telephone check-in calls, the Study Duration - Revised Scale for Caregiving Self- study did not demonstrate Not provided Efficacy, Positive and Negative significant changes in Risk of Bias - 2 Affect Scale caregiver affect, caregiver stress, caregiver distress concerning specific patient behaviors, or caregiver self- efficacy.

210

Evidence Table 1: Populations and Interventions

Given et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 22 2002 US Criteria: Cancer (mixed cancer) Symptom count in the last two Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for weeks (comprised of nausea, Intervention Arm - Not described Comparing a supportive care inclusion if they were at least vomiting, pain, fatigue, nursing intervention for cancer Level of Prognosis: Intervention Arm: Randomization - 40 years old; were within 56 insomnia, dyspnea, patients that utilized problem- Patients and days of initiating their first Not described coordination problems, fever, Multidisciplinarity and Clinical Palliative case solving approaches to families/caregivers cycle of chemotherapy cough, dry mouth, Disciplines: management symptom management and following a new diagnosis of constipation, anorexia, improving physical functioning only Control Arm - No multidisciplinary colon cancer, breast cancer, diarrhea, and mouth sores) (0- and emotional health, with team; clinical disciplines not Study Duration - 20 lung cancer, non-Hodgkin’s 14 scale, lower is better): Palliative case usual care, the study specified. weeks lymphoma, or other solid control, mean 6.8 (SD 1.5); management demonstrated significant Risk of Bias - 0 tumors; reported both pain intervention, mean 7.3 (SD Intervention Arm - No elements: improvements in the and fatigue at baseline 2.8); p=.25. multidisciplinary team; nurses. Nurse-led case intervention group in symptom interview; were cognitively management count and physical role intact; were able to read functioning. The study did not English to comprehend Patient Functional Status: demonstrate significant consent forms; were willing to SF-36 physical role Intervention changes in social role participate in the 10-session functioning subscale (0-100 Duration: functioning, pain, or fatigue. nursing intervention that scale, higher is better): 20 weeks covered 20 weeks; and had a control, mean 11.0 (SD 22.0); caregiver who was willing to intervention, mean 11.0 (SD participate in the study. They 21.0); p=.92. were excluded if they were not

expected to survive the 20- week intervention and the 12- Caregiver Symptoms: week follow-up data collection None described period.

Number of Patient Participants and Decedents: Screened - Not described Enrolled - 113 Completed - Not described Decedents - Not described

211

Evidence Table 2: Outcomes

Given et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 22 2002 US Pain - Pain in the last two weeks Outcomes Study Arms - 2 (Symptom Experience Scale) Other Symptoms: Comparing a supportive care Level of Other Symptoms - Symptom count nursing intervention for cancer Symptom count in the last two weeks patients that utilized problem- Randomization - and fatigue in the last two weeks (comprised of nausea, vomiting, pain, Patients and (Symptom Experience Scale) solving approaches to fatigue, insomnia, dyspnea, symptom management and families/caregivers Functional Status - SF-36 physical coordination problems, fever, cough, only improving physical functioning role functioning subscale dry mouth, constipation, anorexia, and emotional health, with Study Duration - 20 Other QOL-Relevant Outcomes - diarrhea, and mouth sores) (0-14 usual care, the study weeks SF-36 social role functioning scale, lower is better): baseline, demonstrated significant Risk of Bias - 0 subscale control, mean 6.8 (SD 2.1); improvements in the intervention, mean 7.3 (SD 2.8); 20 intervention group in symptom

weeks, control, mean 4.4 (SD 2.7); count and physical role Caregiver Outcomes: intervention, mean 3.3 (SD 2.6); functioning. The study did not p=.051. None described demonstrate significant changes in social role functioning, pain, or fatigue. Functional Status: Physical role functioning (0-100 scale, higher is better): baseline, control, mean 11 (SD 22); intervention, mean 11 (SD 22); 20 weeks, control, mean 31 (SD 36); intervention, mean 50 (SD 41); p=.053.

Caregiver Outcomes: N/A

212

Evidence Table 1: Populations and Interventions

Given et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 42 2006 US Criteria: Cancer (mixed cancer) Depressive symptoms (0-60 Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for scale, lower is better): control, Intervention Arm - Outpatient clinics Comparing a cognitive inclusion if they spoke mean 13.5 (SD 9.05); behavioral intervention aimed Level of Prognosis: Intervention Arm: Randomization - English, were within their first intervention, mean 12.1 (SD at lowering family members’ Patients and two cycles of chemotherapy Not described 9.05). Multidisciplinarity and Clinical Cognitive behavioral reactions and increasing their families/caregivers for treatment of a new tumor, Pain: all participants, 63.1%. Disciplines: intervention aimed at involvement in assisting and were cognitively intact. cancer patients undergoing only Control Arm - No multidisciplinary lowering family Fatigue: all participants, chemotherapy, with usual team; clinical disciplines not members’ reactions Study Duration - 20 Caregivers were eligible for 81.0%. care, the study demonstrated specified. and increasing their weeks inclusion if they saw the Nausea: all participants, involvement in a significant improvement in Risk of Bias - 1 patient at least daily, would 51.0%. Intervention Arm - No assisting cancer the intervention group in agree to assist them during multidisciplinary team; nurses. caregiver reactions to patient Vomiting: all participants, patients undergoing cancer care, and were symptoms. 19.0%. chemotherapy cognitively intact. Insomnia: all participants,

59.7%. Intervention Number of Patient Dyspnea: all participants, Duration: Participants and Decedents: 24.7%. 20 weeks Screened - 609 Diarrhea: all participants, Enrolled - 237 25.5%. Completed - 191 Coordination problem: all Decedents - Not described participants, 14.8%. Anorexia: all participants, 51.3%. Number of Caregiver Participants: Cough: all participants, 22.4%. Screened - 609 Dry mouth: all participants, 30.8%. Enrolled - 237 Constipation: all participants, Completed - 191 35.7%. Inability to concentrate: all participants, 26.2%.

Patient Functional Status: None described

Caregiver Symptoms: Depressive symptoms (0-60 scale, lower is better): control, mean 11.3 (SD 10.86); intervention, mean 10.6 (SD 9.15).

213

Evidence Table 2: Outcomes

Given et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 42 2006 US None described N/A Outcomes Study Arms - 2 Comparing a cognitive behavioral intervention aimed Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - at lowering family members’ Patients and Other QOL-Relevant Outcomes - reactions and increasing their involvement in assisting families/caregivers Study-specific caregiver reaction per Other QOL-Relevant Outcomes: only symptom measure cancer patients undergoing Caregiver reaction score per symptom chemotherapy, with usual Study Duration - 20 (scale not described, lower is better): care, the study demonstrated weeks 10 weeks, control, mean 2.16 (95% CI a significant improvement in Risk of Bias - 1 1.92-2.40); intervention, mean 1.78 the intervention group in (95% CI 1.53-2.03); p=.04. caregiver reactions to patient symptoms.

214

Evidence Table 1: Populations and Interventions

Graff et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 66 2007 EU Criteria: Dementia Depression (0-15 scale, lower Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for is better): control, mean 7.5 Intervention Arm - Home Comparing an occupational inclusion if they were at least (SD 3.0); intervention, mean therapy intervention for Level of Prognosis: Intervention Arm: Randomization - 65 years old, had been 6.9 (SD 3.0) patients with dementia and Patients and diagnosed with mild-to- Not described Depression (0-38 scale, lower Multidisciplinarity and Clinical Occupational therapy their informal caregivers with moderate dementia, were usual care, the study families/caregivers is better): control, mean 8.1 Disciplines: living in the community, and demonstrated significant only (SD 4.6); intervention, mean Control Arm - No multidisciplinary had a primary caregiver who Intervention improvements in the 8.3 (SD 6.2). team; clinical disciplines not Study Duration - cared for them at least once a Duration: intervention group in patient Three months specified. week. They were excluded if Five weeks HRQOL, depression, and Risk of Bias - 3 they had a score of 12 or Patient Functional Status: Intervention Arm - No health status. more on the GDS, had severe multidisciplinary team; occupational None described behavioral or psychological therapist. symptoms in dementia, had Caregiver QOL-Relevant Outcomes severe illnesses as judged by Caregiver Symptoms: a geriatrician, had The study demonstrated Depression (0-60 scale, lower occupational therapy goals significant improvements in is better): control, mean 11.4 which could not defined after the intervention group in (SD 7.2); intervention, mean comprehensive goal setting, caregiver HRQOL, 11.7 (SD 8.3). or were not on stable depression, health status, and treatment by an antidementia sense of control over life. drug.

Caregivers were excluded if they had severe illnesses.

Number of Patient Participants and Decedents: Screened - 275 Enrolled - 135 Completed - 105 Decedents - 1

Number of Caregiver Participants: Screened - 275 Enrolled - 135 Completed - 105

215

Evidence Table 2: Outcomes

Graff et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 66 2007 EU Depression - CSDD Outcomes Depression: Study Arms - 2 HRQOL - Dqol Comparing an occupational Depression (0-38 scale, lower is better): baseline, therapy intervention for Level of Other QOL-Relevant Outcomes - control, mean 8.1 (SD 4.6); intervention, mean 8.3 (SD 6.2); six weeks, control, mean 9.2 (SD 6.4); patients with dementia and Randomization - GHQ-12 Patients and intervention, mean 6.5 (SD 5.3); p<.001; three their informal caregivers with months, control, mean 9.2 (SD 6.1); intervention, usual care, the study families/caregivers mean 6.2 (SD 4.6); p<.001. demonstrated significant only Caregiver Outcomes: improvements in the Study Duration - Depression - CES-D HRQOL: intervention group in patient Three months QOL (1-5 scale, higher is better): baseline, control, HRQOL, depression, and HRQOL - Dqol mean 3.3 (SD 1.3); intervention, mean 3.4 (SD 0.7); Risk of Bias - 3 six weeks, control, mean 3.1 (SD 0.8); intervention, health status. Other QOL-Relevant Outcomes - mean 4.0 (SD 0.6); p<.001; three months, control, GHQ-12, Mastery Scale mean 3.1 (SD 0.9); intervention, mean 4.0 (SD 0.5); p<.001. Caregiver QOL-Relevant Outcomes Other QOL-Relevant Outcomes: The study demonstrated Health status (0-36 scale, lower is better): baseline, significant improvements in control, mean 12.3 (SD 5.7); intervention, mean 10.7 (SD 4.1); six weeks, control, mean 11.8 (SD 5.6); the intervention group in intervention, mean 7.8 (SD 4.2); p<.001; three caregiver HRQOL, months, control, mean 14.0 (SD 6.0); intervention, mean 9.1 (SD 4.2); p=.001. depression, health status, and sense of control over life.

Caregiver Outcomes:

Depression: Depression (0-60 scale, lower is better): baseline, control, mean 11.4 (SD 7.2); intervention, mean 11.7 (SD 8.3); six weeks, control, mean 12.6 (SD 8.5); intervention, mean 5.8 (SD 4.8); p<.001; three months, control, mean 13.1 (SD 9.1); intervention, mean 5.4 (SD 4.5); p<.001.

HRQOL: QOL (1-5 scale, higher is better): baseline, control, mean 3.4 (SD 0.8); intervention, mean 3.3 (SD 0.9); six weeks, control, mean 3.4 (SD 0.7); intervention, mean 4.0 (SD 0.6); p<.001; three months, control, mean 3.4 (SD 0.8); intervention, mean 4.1 (SD 0.6); p<.001.

Other QOL-Relevant Outcomes: Health status (0-36 scale, lower is better): baseline, control, mean 11.3 (SD 4.0); intervention, mean 12.0 (SD 4.9); six weeks, control, mean 11.0 (SD 3.9); intervention, mean 7.0 (SD 3.9); p<.001; three months, control, mean 12.1 (SD 5.0); intervention, mean 7.1 (SD 3.5); p<.001. Sense of control over life (5-25 scale, higher is better): baseline, control, mean 12.0 (SD 3.0); intervention, mean 12.6 (SD 3.6); six weeks, control, mean 12.6 (SD 3.0); intervention, mean 16.6 (SD 3.0); p<.001; three months, control, mean 12.3 (SD 2.8); intervention, mean 16.7 (SD 2.7); p<.001.

216

Evidence Table 1: Populations and Interventions

Graff et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 67 2006 EU Criteria: Dementia Depression (0-15 scale, lower Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for is better): control, mean 7.5 Intervention Arm - Home Comparing an occupational inclusion if they were at least (SD 3.0); intervention, mean therapy intervention for Level of Prognosis: Intervention Arm: Randomization - 65 years old, had been 6.9 (SD 3.0). patients with dementia and Patients and diagnosed with mild-to- Not described Depression (0-38 scale, lower Multidisciplinarity and Clinical Occupational therapy their informal caregivers, with moderate dementia, were usual care, the study families/caregivers is better): control, mean 8.1 Disciplines: living in the community, and demonstrated significant only (SD 4.6); intervention, mean Control Arm - No multidisciplinary had a primary caregiver who Intervention improvements in the 8.3 (SD 6.2). team; clinical disciplines not Study Duration - cared for them at least once a Duration: intervention group in Three months specified. week. They were excluded if Five weeks occupational performance Risk of Bias - 6 they had a score of 12 or Patient Functional Status: Intervention Arm - No process skills and the need for more on the GDS, had severe multidisciplinary team; occupational assistance in daily activities. IDDD performance scale (0-44 behavioral or psychological therapist. scale, lower is better): control, symptoms in dementia, had mean 24.5 (SD 8.7); severe illnesses as judged by Caregiver QOL-Relevant intervention, mean 23.5 (SD a geriatrician, had Outcomes 7.9). occupational therapy goals The study demonstrated a which could not defined after AMPS motor scale (-3 to 4 significant improvement in the comprehensive goal setting, scale, higher is better): intervention group in caregiver or were not on stable control, mean 1.1 (SD 1.0); sense of competence. treatment by an antidementia intervention, mean 1.0 (SD drug. 1.1). AMPS process scale (-3 to 4 Caregivers were excluded if scale, higher is better): they had severe illnesses. control, mean 0.3 (SD 0.8); intervention, mean 0.2 (SD 0.8). Number of Patient Participants and Decedents: Screened - 275 Caregiver Symptoms: Enrolled - 135 Depression (0-60 scale, lower is better): control, mean 11.4 Completed - 105 (SD 7.2); intervention, mean Decedents - 1 11.7 (SD 8.3).

Number of Caregiver Participants: Screened - 275 Enrolled - 135 Completed - 105

217

Evidence Table 2: Outcomes

Graff et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 67 2006 EU Functional Status - IDDD Outcomes Study Arms - 2 performance scale, AMPS process Functional Status: Comparing an occupational Level of scale therapy intervention for Need for assistance in daily activities patients with dementia and Randomization - (0-44 scale, lower is better): baseline, Patients and their informal caregivers, with Caregiver Outcomes: control, mean 24.5 (SD 8.7); usual care, the study families/caregivers intervention, mean 23.5 (SD 7.9); six only Other QOL-Relevant Outcomes - demonstrated significant SCQ weeks, control, mean 25.3 (SD 8.6); improvements in the Study Duration - intervention, mean 14.4 (SD 6.1); intervention group in Three months p<.001; three months, control, mean occupational performance Risk of Bias - 6 27.2 (SD 8.9); intervention, mean 13.6 process skills and the need for (SD 6.0); p<.001. assistance in daily activities. Occupational performance process skills (-3 to 4 scale, higher is better): baseline, control, mean 0.3 (SD 0.8); Caregiver QOL-Relevant intervention, mean 0.2 (SD 0.8); six Outcomes weeks, control, mean 0.2 (SD 0.8); The study demonstrated a intervention, mean 1.2 (SD 0.7); significant improvement in the p<.001; three months, control, mean intervention group in caregiver 0.0 (SD 0.7); intervention, mean 1.2 sense of competence. (SD 0.8); p<.001.

Caregiver Outcomes:

Other QOL-Relevant Outcomes: Sense of competence (27-135 scale, higher is better): baseline, control, mean 90.4 (SD 13.6); intervention, mean 89.7 (SD 14.9); six weeks, control, mean 88.4 (SD 13.7); intervention, mean 104.6 (SD 13.4); p<.001; three months, control, mean 89.4 (SD 14.4); intervention, mean 107.3 (SD 13.6); p<.001.

218

Evidence Table 1: Populations and Interventions

Grande et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 135 al., 2004 EU Criteria: Not described None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Caregivers were eligible for Intervention Arm - Home Comparing a hospice at home inclusion if they were intervention with usual care, Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - bereaved informal carers of the study did not demonstrate Patients terminally ill patients who took Referred to or receiving None described Multidisciplinarity and Clinical Hospice any significant changes in part in an RCT of the caregivers’ feelings, actions, hospice Disciplines: Study Duration - 15 Cambridge Hospice at Home or relationships both Control Arm - No multidisciplinary months program. Caregiver Symptoms: Intervention immediately following the team; clinical disciplines not Duration: patient’s death and six months Risk of Bias - 5 None described specified. Two weeks after death; or in caregivers’ Number of Patient Intervention Arm - Hospice care overall physical or mental Participants and Decedents: team; clinical disciplines not health at six months after Screened - 262 specified. death. Enrolled - 242 Completed - 229 Decedents - 229

Number of Caregiver Participants: Screened - 198 Enrolled - 115 Completed - 96

219

Evidence Table 2: Outcomes

Grande et Study Location - Patient Outcomes: None None N/A Caregiver QOL-Relevant 135 al., 2004 EU None described Outcomes Study Arms - 2 Comparing a hospice at home intervention with usual care, Level of Caregiver Outcomes: Randomization - the study did not demonstrate Patients Other QOL-Relevant Outcomes - any significant changes in Bereavement’s effect on feelings, caregivers’ feelings, actions, Study Duration - 15 actions, and relationships (TRIG); or relationships both months SF-36 PCS and MCS immediately following the Risk of Bias - 5 patient’s death and six months after death; or in caregivers’ overall physical or mental health at six months after death.

220

Evidence Table 1: Populations and Interventions

Haley et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 81 2008 US Criteria: Dementia None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Caregivers were eligible for Intervention Arm - Outpatient clinics Comparing a counseling and inclusion if they had support group intervention for Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - experienced the death of their caregivers of dementia Families or care recipient during Not described None described Multidisciplinarity and Clinical Counseling sessions, patients, with usual care, the involvement in the NYU study demonstrated a caregivers Disciplines: ad hoc telephone Caregiver Intervention study, significant increase in the Control Arm - No multidisciplinary counseling, and Study Duration - had provided post- Caregiver Symptoms: intervention group in the team; clinical disciplines not weekly support group Two years bereavement outcome data, resilient pattern of depressive Depression (0-15 scale, lower specified. meetings were primarily responsible for symptoms during Risk of Bias - 0 is better): control, mean 11.1 caregiving, were caring for (SD 7.2); intervention, mean Intervention Arm - No bereavement, and a significant their spouse at home at 9.4 (SD 5.7); p<.05. multidisciplinary team; counselors. Intervention decrease in chronically baseline, and had at least one Duration: depressed pattern of other family member or Two years depressive symptoms during relative living in the vicinity. bereavement.

Number of Caregiver Participants: Screened - Not described Enrolled - 406 Completed - 254

221

Evidence Table 2: Outcomes

Haley et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 81 2008 US None described N/A Outcomes Study Arms - 2 Comparing a counseling and support group intervention for Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - caregivers of dementia Families or Depression - GDS patients, with usual care, the study demonstrated a caregivers Depression: significant increase in the Study Duration - Resilient pattern of depressive intervention group in the Two years symptoms post-bereavement (scores resilient pattern of depressive Risk of Bias - 0 11 or less on the GDS [0-15 scale, symptoms during lower is better] at patient’s death and bereavement, and a significant one and two years post-death): control, decrease in chronically 42.9%; intervention, 60.0%; p=.023. depressed pattern of Chronically depressed pattern of depressive symptoms during depressive symptoms post- bereavement. bereavement (scores of greater than 11 on the GDS [0-15 scale, lower is better] at patient’s death and one and two years post-death): control, 17.9%; intervention, 7.8%; p=.044.

222

Evidence Table 1: Populations and Interventions

Kalauokalani Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 73 et al., 2007 US Criteria: Cancer (mixed cancer) Average pain (0-10 scale, Control Arm - Outpatient clinics Standardized Outcomes Study Arms - 2 Patients were eligible for lower is better): control, mean Intervention Arm - Outpatient clinics education session Comparing an education and inclusion if they spoke 5.2 (SD 1.6); intervention, emphasizing coaching session for cancer Level of Prognosis: Randomization - English, were between 18 and mean 5.3 (SD 1.8). fundamental patients designed to address Patients 75, had a diagnosis of cancer, Not described Multidisciplinarity and Clinical principles of cancer misconceptions about pain experienced an average pain Disciplines: pain control, following treatment and to encourage Study Duration - score of least 30 out of 100 on Patient Functional Status: the outline of a talking to the oncologist about Control Arm - No multidisciplinary Two weeks the VAS over the prior two pamphlet produced pain control, with a SF-12 PCS (0-100 scale, team; masters-level psychology weeks, and had no scheduled by the AHRQ standardized education Risk of Bias - 1 higher is better): control, mean student, fourth-year medical major surgical procedures session emphasizing pain 29.1 (SD 8.0); intervention, student. during the follow-up interval of mean 26.1 (SD 8.5). control, the study two weeks. They were Intervention Arm - No Intervention Arm: demonstrated a significant

excluded if they were enrolled multidisciplinary team; masters- Education and improvement in the in hospice or if they were Caregiver Symptoms: level psychology student, fourth- coaching session intervention group in minority under the care of the Pain None described year medical student. designed to address patients’ average pain as Management Service, which misconceptions about compared with the average would have made them less pain treatment and to pain of both white patients in likely to benefit from the encourage talking to the intervention group and intervention. the oncologist about minority patients in the control pain control group. Number of Patient Participants and Decedents: Intervention Screened - 1073 Duration: Enrolled - 78 One-time intervention Completed - 67 Decedents - 1

223

Evidence Table 2: Outcomes

Kalauokalani Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 73 et al., 2007 US Pain - VAS Outcomes Study Arms - 2 Pain: Comparing an education and Level of coaching session for cancer Caregiver Outcomes: Average pain (0-10 scale, lower is patients designed to address Randomization - better): baseline, white patients, mean Patients None described misconceptions about pain 5.038; minority patients, mean 6.000; treatment and to encourage Study Duration - p=.049; two weeks, control, white talking to the oncologist about Two weeks patients, mean 4.65; minority patients, pain control, with a Risk of Bias - 1 mean 6.42; p=.012; two weeks, standardized education intervention, white patients, mean 4.31; session emphasizing pain minority patients, mean 4.00; p=.71. control, the study demonstrated a significant improvement in the Caregiver Outcomes: intervention group in minority N/A patients’ average pain as compared with the average pain of both white patients in the intervention group and minority patients in the control group.

224

Evidence Table 1: Populations and Interventions

Keefe et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 27 2005 US Criteria: Cancer (mixed cancer) None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a partner-guided inclusion if they were at least pain management training Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - 18 years old, had an program with usual care for Patients and advanced cancer diagnosis Identified by clinicians as FACT-G physical well-being Multidisciplinarity and Clinical Partner-guided pain cancer patients, the study did families/caregivers (i.e., metastatic or widespread poor prognosis, referred subscale (0-4 scale, higher is Disciplines: management training not demonstrate any disease) with disease-related significant changes in pain or only to or receiving hospice better): control, mean 2.08 Control Arm - No multidisciplinary pain, a worst pain rating of at physical well-being. (SD 0.77); intervention, mean team; clinical disciplines not Study Duration - least three on the BPI, a life Intervention 2.02 (SD 0.77). specified. One month expectancy of less than six Duration: Caregiver QOL-Relevant Risk of Bias - 10 months (i.e., they met the Intervention Arm - No Two weeks Medicare hospice benefit Caregiver Symptoms: multidisciplinary team; nurses. Outcomes definition for hospice None described The study demonstrated eligibility), no change in significant improvements in disease treatment planned. the intervention group in caregivers’ reported self- efficacy for control of pain and Number of Patient other symptoms. The study Participants and Decedents: did not demonstrate significant Screened - 173 changes in caregiver strain or Enrolled - 78 mood. Completed - 56 Decedents - 12

Number of Caregiver Participants: Screened - 173 Enrolled - 78 Completed - 56

225

Evidence Table 2: Outcomes

Keefe et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 27 2005 US Pain - BPI usual pain and worst pain None Outcomes Study Arms - 2 over last week scales Comparing a partner-guided pain management training Level of Functional Status - FACT-G physical Caregiver Outcomes: Randomization - well-being subscale program with usual care for Patients and cancer patients, the study did not demonstrate any families/caregivers Other QOL-Relevant Outcomes: only Caregiver Outcomes: significant changes in pain or Self-efficacy for pain control in patient physical well-being. Study Duration - Other QOL-Relevant Outcomes - (10-100 scale, higher is better): one

One month Chronic Pain Self-Efficacy Scale, month, control, mean 43.68 (SD Caregiver Strain Index, mood Risk of Bias - 10 17.92); intervention, mean 57.90 (SD Caregiver QOL-Relevant (POMS-B) 17.97); p=.006. Outcomes Self-efficacy for control of other The study demonstrated symptoms in patient (10-100 scale, significant improvements in higher is better): one month, control, the intervention group in mean 54.03 (SD 18.20); intervention, caregivers’ reported self- mean 66.93 (SD 18.21); p=.012. efficacy for control of pain and other symptoms. The study did not demonstrate significant changes in caregiver strain or mood.

226

Evidence Table 1: Populations and Interventions

Kissane et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 56 al., 2006 Australia Criteria: Cancer (mixed cancer) None described Control Arm - Home Usual care, hospice Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home, hospital Comparing a family focused inclusion if they had a grief therapy intervention for Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of cancer, their families of terminal cancer Families or treating physician gave a Identified by clinicians as None described Multidisciplinarity and Clinical Hospice, family patients with usual care, the prognosis of six months, they study demonstrated significant caregivers poor prognosis Disciplines: focused grief therapy were between 35 and 70, they improvements in the Control Arm - No multidisciplinary Study Duration - 13 had adequate command of Caregiver Symptoms: intervention group in team; clinical disciplines not months after English, were geographically Intervention depression and general None described specified. patient death accessible, had a living Duration: symptom severity among the Risk of Bias - 4 partner, and had one or more Intervention Arm 1 - No 18 months most distressed families, but children at least 12 years old. multidisciplinary team; social not among all families. The Families were eligible for workers. study did not demonstrate inclusion if they were at risk of significant changes in poor psychosocial outcomes caregiver experience of as indicated by scores of four bereavement, caregiver social or less on cohesiveness or functioning, or family nine or less on the Family functioning. Relationships Index. Health Care Use Number of Patient The study did not demonstrate Participants and Decedents: significant changes in GP Screened - 483 visits or counseling use. Enrolled - 81 Completed - 81 Drugs Decedents - 67 The study did not demonstrate significant changes in the use of antidepressants, hypnotics, Number of Caregiver or tranquilizers. Participants: Screened - 483 Enrolled - 81 Completed - 67

227

Evidence Table 2: Outcomes

Kissane et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Outpatient None Caregiver QOL-Relevant 56 al., 2006 Australia None described N/A visits, counseling Outcomes Study Arms - 2 Drugs - Use of antidepressants, Comparing a family focused hypnotics, and tranquilizers grief therapy intervention for Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - families of terminal cancer Families or Depression - BDI-SF patients with usual care, the study demonstrated significant caregivers Other Symptoms - Brief Symptom Depression: improvements in the Inventory Study Duration - 13 Depression among most distressed intervention group in months after Other QOL-Relevant Outcomes - families (scale not described): six depression and general patient death Social functioning (Social Adjustment months vs. baseline, control, mean symptom severity among the Risk of Bias - 4 Scale), family functioning (Family +0.43; intervention, +5.92; p<.01; 13 most distressed families, but Environment Scale and Family months vs. baseline, control, mean not among all families. The Assessment Device), Bereavement +3.86; intervention, mean +5.13; study did not demonstrate Phenomenology Questionnaire p<.01. significant changes in caregiver experience of bereavement, caregiver social Other Symptoms: functioning, or family General symptom severity among most functioning. distressed families (scale not described): six months vs. baseline, control, mean +0.16; intervention, Health Care Use mean +0.83; p<.01; 13 months vs. The study did not demonstrate baseline control, mean +0.30; significant changes in GP intervention, mean +0.81; p<.01. visits or counseling use.

Drugs The study did not demonstrate significant changes in the use of antidepressants, hypnotics, or tranquilizers.

228

Evidence Table 1: Populations and Interventions

Lautrette et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 110 al., 2007 EU Criteria: Cancer (type not None described Control Arm - Hospital Usual care Outcomes Study Arms - 2 Patients were eligible for described), CHF, COPD, Intervention Arm - Hospital Comparing a family inclusion if the physician in ESLD conference and bereavement Level of Patient Functional Status: Intervention Arm: Randomization - charge believed they would brochure intervention with die within a few days. They Poor performance status Multidisciplinarity and Clinical End-of-life family usual care for family members Families or Prognosis: caregivers were excluded if they were (scale not described): control, Disciplines: conference and of patients dying in the ICU, younger than 18. the study demonstrated Identified by clinicians as 44%; intervention, 43%; p=.61 Control Arm - No multidisciplinary brochure on Study Duration - 90 significant improvements in poor prognosis team; clinical disciplines not bereavement days the intervention group in specified. Risk of Bias - 8 Family members were Caregiver Symptoms: anxiety, depression, and excluded if they had None described Intervention Arm - No Intervention PTSD symptoms. insufficient knowledge of multidisciplinary team; clinical Duration: French for a telephone disciplines not specified. One-time intervention interview. Health Care Use The study did not demonstrate Number of Patient significant changes in hospital Participants and Decedents: length of stay or ICU length of stay. Screened - 132

Enrolled - 126 Drugs Completed - 125 The study demonstrated Decedents - 125 significant increases in the intervention group in withdrawal of vasopressors Number of Caregiver and in rates of new Participants: psychotropic drug Screened - 132 prescriptions for family Enrolled - 126 members after death of the patient. Completed - 108

Life-Extending Devices or Procedures The study demonstrated a significant increase in the intervention group in withdrawal of mechanical ventilation. The study did not demonstrate a significant change in withdrawal of dialysis.

229

Evidence Table 2: Outcomes

Lautrette et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Drugs Caregiver QOL-Relevant 110 al., 2007 EU None described N/A hospital, ICU Vasopressors withdrawn: Outcomes Study Arms - 2 Drugs - Withdrawal of control, 30%; intervention, 51%; Comparing a family vasopressors, rate of new p=.01. conference and bereavement Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - psychotropic drug prescriptions Family members received newly brochure intervention with Families or Depression - HADS depression for family members after death prescribed psychotropic drugs usual care for family members of the patient of patients dying in the ICU, caregivers subscale Depression: after death of patient: 90 days, Life-Extending Devices or control, 23%; intervention, 11%; the study demonstrated Study Duration - 90 Anxiety - HADS anxiety subscale Symptoms of depression (0-21 scale, Procedures - Withdrawal of p=.05. significant improvements in days Other Symptoms - IES lower is better): score greater than 8, mechanical ventilation, the intervention group in 90 days, control, 56%; intervention, Risk of Bias - 8 withdrawal of dialysis anxiety, depression, and 29%; p=.003. PTSD symptoms.

Anxiety: Health Care Use Symptoms of anxiety (0-21 scale, The study did not demonstrate lower is better): score greater than 8, significant changes in hospital 90 days, control, 67%; intervention, length of stay or ICU length of 45%; p=.02. stay.

Other Symptoms: Drugs PTSD symptoms (0-75 scale, lower is The study demonstrated better): 90 days, control, median 39 significant increases in the (IQR 25-48); intervention, median 27 intervention group in (IQR 18-42); p=.02. withdrawal of vasopressors and in rates of new psychotropic drug prescriptions for family members after death of the patient.

230

Evidence Table 1: Populations and Interventions

Miaskowski Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 29 et al., 2007 US Criteria: Cancer (mixed cancer) None described Control Arm - Home Agency for Health Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Care Policy and Comparing a pain education inclusion if they were oncology Research’s Cancer and management program for Level of Prognosis: Patient Functional Status: Randomization - outpatients at least 18 years Pain Guideline given cancer patients, with giving Patients old, were able to read and Not described KPS score (0-100 scale, Multidisciplinarity and Clinical to patients patients a clinical pain write English, had KPS scores higher is better): control, mean Disciplines: guideline, the study Study Duration - of at least 50, had an average demonstrated a significant 71.9 (SD 12.2); intervention, Control Arm - No multidisciplinary Six weeks pain intensity score of at least Intervention Arm: improvement in the mean 68.7 (SD 11.4); p>.05. team; nurses. 2.5 on the BPI, and had intervention group in bodily Risk of Bias - 10 PRO-SELF Pain radiographic evidence of bone Intervention Arm - No Control Program, a pain. The study did not metastasis. Caregiver Symptoms: multidisciplinary team; nurses. pain education and demonstrate significant None described management changes in pain interference, program depression, or overall physical Number of Patient or mental health. Participants and Decedents: Screened - Not described Intervention Duration: Enrolled - 212 Six weeks Completed - 174 Decedents - 10

231

Evidence Table 2: Outcomes

Miaskowski Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 29 et al., 2007 US Pain - BPI pain interference items, Outcomes Study Arms - 2 SF-36 bodily pain subscale Pain: Comparing a pain education Level of Depression - POMS and management program for Bodily pain (0-100 scale, higher is cancer patients, with giving Randomization - Functional Status - SF-36 PCS and better): six weeks, control, mean 39.6 Patients patients a clinical pain physical functioning subscale (SD 19.8); intervention, mean 46.8 (SD guideline, the study Study Duration - Other QOL-Relevant Outcomes - 22.1); p=.005. demonstrated a significant Six weeks SF-36 MCS and mental health improvement in the intervention group in bodily Risk of Bias - 10 subscale Caregiver Outcomes: pain. The study did not

N/A demonstrate significant Caregiver Outcomes: changes in pain interference, None described depression, or overall physical or mental health.

232

Evidence Table 1: Populations and Interventions

Miaskowski Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 30 et al., 2004 US Criteria: Cancer (mixed cancer) Average pain (0-10 scale, Control Arm - Home Agency for Health Outcomes Study Arms - 2 Patients were eligible for lower is better): control, mean Intervention Arm - Home Care Policy and Comparing a pain education inclusion if they were oncology 4.2 (SD 1.8); intervention, Research’s Cancer and management program for Level of Prognosis: Randomization - outpatients at least 18 years mean 4.1 (SD 1.7); p>.05. Pain Guideline given cancer patients with giving Patients old, were able to read and Not described Least pain (0-10 scale, lower Multidisciplinarity and Clinical to patients patients a clinical pain write English, had KPS scores is better): control, mean 1.8 Disciplines: guideline, the study Study Duration - of at least 50, had an average demonstrated significant (SD 1.7); intervention, mean Control Arm - No multidisciplinary Six weeks pain intensity score of at least Intervention Arm: improvements in the 1.9 (SD 1.6); p>.05. team; nurses. Risk of Bias - 10 2.5 on the BPI, and had PRO-SELF Pain intervention group in average, radiographic evidence of bone Worst pain (0-10 scale, lower Intervention Arm - No Control Program, a least, and worst pain. is better): control, mean 6.7 multidisciplinary team; nurses. metastasis. They were pain education and excluded if they had an (SD 2.4); intervention, mean management identified family caregiver who 6.7 (SD 2.1); p>.05. program Drugs could not or would not The study demonstrated a participate in the intervention. significant increase in the Patient Functional Status: Intervention intervention group in both KPS score (0-100 scale, Duration: around the clock and as Number of Patient higher is better): control, mean Six weeks needed opioid prescriptions. Participants and Decedents: 71.5 (SD 12.3); intervention, The study did not demonstrate Screened - Not described mean 69.1 (SD 11.4); p>.05. significant changes in only Enrolled - 212 around the clock opioid prescriptions, only as needed Caregiver Symptoms: Completed - 174 opioid prescriptions, no opioid Decedents - 10 None described prescriptions, total amount of opioid prescriptions prescribed and taken, amount of around the clock opioid prescriptions prescribed and taken, or prescriptions for NSAIDs, corticosteroids, pamidronate, or strontium.

233

Evidence Table 2: Outcomes

Miaskowski Study Location - Patient Outcomes: Patient Outcomes: Drugs - Around the clock opioid Drugs Patient QOL-Relevant 30 et al., 2004 US Pain - VAS prescriptions, as needed opioid Both around the clock and as Outcomes Study Arms - 2 prescriptions, or both; amount of needed opioid prescriptions: Comparing a pain education Pain: total, around the clock, and as Level of baseline, control, 28.8%; and management program for Caregiver Outcomes: Average pain (0-10 scale, lower is needed opioid prescriptions intervention, 28.3%; six weeks, cancer patients with giving Randomization - better): six weeks vs. baseline, control, prescribed and taken; Patients None described control, 32.5%; intervention, patients a clinical pain mean +1.9%; intervention, mean - prescriptions for NSAIDs, 37.0%; group by time p=.008. guideline, the study Study Duration - 32.5%; group by time p<.001. corticosteroids, pamidronate, or demonstrated significant strontium Six weeks Least pain (0-10 scale, lower is better): improvements in the Risk of Bias - 10 six weeks vs. baseline, control, mean intervention group in average, +14.6%; intervention, mean -28.4%; least, and worst pain. group by time p<.001. Worst pain (0-10 scale, lower is better): Drugs six weeks vs. baseline, control, mean - 1.2%; intervention, mean -27.0%; The study demonstrated a group by time p<.001. significant increase in the intervention group in both around the clock and as Caregiver Outcomes: needed opioid prescriptions. The study did not demonstrate N/A significant changes in only around the clock opioid prescriptions, only as needed opioid prescriptions, no opioid prescriptions, total amount of opioid prescriptions prescribed and taken, amount of around the clock opioid prescriptions prescribed and taken, or prescriptions for NSAIDs, corticosteroids, pamidronate, or strontium.

234

Evidence Table 1: Populations and Interventions

Mittelman et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 112 al., 2008 Australia Criteria: Dementia None described Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a counseling inclusion if they met NINCDS- intervention for caregivers of Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - ADRDA and DSM IV criteria Alzheimer’s patients with Patients and for probable Alzheimer’s Not described None described Multidisciplinarity and Clinical In-person and ad hoc usual care, the study Disease; had a Global demonstrated a significant families/caregivers Disciplines: telephone counseling Deterioration Scale score of 4 improvement in the only Control Arm - No multidisciplinary to 5, indicating mild to Caregiver Symptoms: intervention group in caregiver Study Duration - team; clinical disciplines not moderate dementia at Depression (0-63 scale, lower Intervention depression. Two years specified. Duration: enrollment; had no is better): minimal depression Risk of Bias - 4 contraindication to donepezil; (0-13), control, 81.8%; Intervention Arm - No Two years Drugs were stable with other intervention, 81.0%; mild multidisciplinary team; counselors. medications; were in good depression (14-19), control, The study did not demonstrate general physical health; were 12.2%; intervention, 14.3%; a significant change in able to give informed consent, moderate depression (20-28), caregiver antidepressant use. or if not able, not object to control, 5.1%; intervention, participating; and were 3.9%; severe depression (29- residing in the community with 63), control, 1.4%; their spouse. intervention, 0.0%; p=.49.

Caregivers were eligible for inclusion if they were the patients’ spouses, were self- defined as the primary caregiver, and gave informed consent. They were excluded if they had previously received formal caregiver counseling.

Number of Patient Participants and Decedents: Screened - 169 Enrolled - 158 Completed - 82 Decedents - 20

Number of Caregiver Participants: Screened - 169 Enrolled - 158 Completed - 82

235

Evidence Table 2: Outcomes

Mittelman et Study Location - Patient Outcomes: Patient Outcomes: Drugs - Caregiver None Caregiver QOL-Relevant 112 al., 2008 Australia None described N/A antidepressant use Outcomes Study Arms - 2 Comparing a counseling intervention for caregivers of Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - Alzheimer’s patients with Patients and Depression - BDI usual care, the study demonstrated a significant families/caregivers Depression: only improvement in the Depression (0-63 scale, lower is intervention group in caregiver Study Duration - better): two years vs. baseline, depression. Two years intervention vs. control, linear

Risk of Bias - 4 regression coefficient -0.38 (SE 0.17); group by time p=.031. Drugs The study did not demonstrate a significant change in caregiver antidepressant use.

236

Evidence Table 1: Populations and Interventions

Mittelman et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 113 al., 2004 US Criteria: Dementia None described Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a counseling inclusion if they had a intervention for caregivers of Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of Alzheimer’s Alzheimer’s patients with Patients and Disease. Not described None described Multidisciplinarity and Clinical In-person and ad hoc usual care, the study demonstrated a significant families/caregivers Disciplines: telephone counseling Caregivers were eligible for improvement in the only Control Arm - No multidisciplinary inclusion if they were living Caregiver Symptoms: intervention group in caregiver Study Duration - team; clinical disciplines not with the patient at baseline Depression (0-30 scale, lower Intervention depression. Five years specified. and they or the patient had at is better): all participants, Duration: Risk of Bias - 2 least one relative living in the mean 9.8 (SD 6.6). Intervention Arm - No Five years New York City metropolitan multidisciplinary team; counselors. area.

Number of Patient Participants and Decedents: Screened - Not described Enrolled - 406 Completed - 223 Decedents - Not described

Number of Caregiver Participants: Screened - Not described Enrolled - 406 Completed - 223

237

Evidence Table 2: Outcomes

Mittelman et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 113 al., 2004 US None described N/A Outcomes Study Arms - 2 Comparing a counseling intervention for caregivers of Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - Alzheimer’s patients with Patients and Depression - GDS usual care, the study demonstrated a significant families/caregivers Depression: only improvement in the Depression (0-30 scale, lower is intervention group in caregiver Study Duration - better): one year vs. baseline, linear depression. Five years regression intercept 9.70 (SE 0.45); Risk of Bias - 2 group by time interaction -1.38 (SE 0.44); p=.002; intervention vs. control, lower depression through 3.1 years; p<.05.

238

Evidence Table 1: Populations and Interventions

Nobili et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 68 2004 EU Criteria: Dementia None described Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a psychologist and inclusion if they had a clinical occupational therapist Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of dementia; had a intervention for caregivers of Patients and score 2 or more on at least Not described Katz ADL scale (0-6 scale, Multidisciplinarity and Clinical Home visit by patients with dementia, with families/caregivers one problem behavior, as lower is better): control, mean Disciplines: psychologist and usual care, the study did not assessed by the SBI-C; lived demonstrate a significant only 2.1 (SD 1.8); intervention, Control Arm - No multidisciplinary occupational therapist within 10 miles of Milan; and change in patient functional mean 2.6 (SD 2.0); p=.60. team; clinical disciplines not to develop a care Study Duration - 12 had a principal caregiver. plan status. months specified.

Risk of Bias - 3 Caregiver Symptoms: Intervention Arm - No Number of Patient Caregiver QOL-Relevant None described multidisciplinary team; psychologist, Intervention Participants and Decedents: occupational therapist. Duration: Outcomes Screened - 1595 One-time intervention The study did not demonstrate Enrolled - 69 a significant change in caregiver stress. Completed - 39

Decedents - 8 Drugs

The study did not demonstrate Number of Caregiver a significant change in the use Participants: of psychotropic drugs Screened - 103 (neuroleptics, benzodiazepines, or Enrolled - 69 antidepressants). Completed - 39

239

Evidence Table 2: Outcomes

Nobili et al., Study Location - Patient Outcomes: None Drugs - Use of psychotropic None Patient QOL-Relevant 68 2004 EU Functional Status - Katz ADL scale drugs (neuroleptics, Outcomes benzodiazepines, or Study Arms - 2 Comparing a psychologist and antidepressants) occupational therapist Level of Caregiver Outcomes: Randomization - intervention for caregivers of Patients and Other QOL-Relevant Outcomes - patients with dementia, with families/caregivers RSS usual care, the study did not only demonstrate a significant change in patient functional Study Duration - 12 status. months

Risk of Bias - 3 Caregiver QOL-Relevant Outcomes The study did not demonstrate a significant change in caregiver stress.

Drugs The study did not demonstrate a significant change in the use of psychotropic drugs (neuroleptics, benzodiazepines, or antidepressants).

240

Evidence Table 1: Populations and Interventions

Northouse et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 31 al., 2005 US Criteria: Cancer (breast) None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a family-based inclusion if they had a education and support Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - confirmed diagnosis of intervention for patients with Patients and recurrent and/or progressive Not described KPS score (0-100 scale, Multidisciplinarity and Clinical FOCUS program, a recurrent breast cancer and families/caregivers breast cancer within the higher is better): intervention Disciplines: family-based their caregivers, with usual previous month, were at least care, the study demonstrated only vs. control, no difference; Control Arm - No multidisciplinary education and 21 years old, were mentally a significant improvement in p=.49. team; clinical disciplines not support intervention Study Duration - and physically able to the intervention group in specified. Six months participate, were able to read hopelessness at three months Risk of Bias - 2 and understand English, had a Caregiver Symptoms: Intervention Arm - No Intervention but not six months. The study life expectancy of at least six None described multidisciplinary team; nurses. Duration: did not demonstrate significant months determined by a Four months changes in symptom distress, physician or protocol nurse, physical QOL, mental QOL, or and were able to identify a coping effectiveness. family caregiver to participate in the study. Caregiver QOL-Relevant Caregivers were eligible for Outcomes inclusion if they were at least The study did not demonstrate 21 years old, were mentally significant changes in and physically able to caregiver hopelessness, participate, were able to speak physical QOL, mental QOL, or and understand English, and coping effectiveness. were identified by the patient as the primary caregiver.

Number of Patient Participants and Decedents: Screened - 250 Enrolled - 200 Completed - 134 Decedents - 38

Number of Caregiver Participants: Screened - 250 Enrolled - 200 Completed - 134

241

Evidence Table 2: Outcomes

Northouse et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 31 al., 2005 US Other Symptoms - Symptom distress Outcomes Study Arms - 2 (OSQ) Other QOL-Relevant Outcomes: Comparing a family-based Level of Functional Status - Physical QOL education and support Hopelessness (0-20 scale, lower is intervention for patients with Randomization - (study-specific scale comprised of better): baseline, control, mean 2.98 Patients and physical SF-36 and FACT-G recurrent breast cancer and (SD 4.0); intervention, mean 4.52 (SD their caregivers, with usual families/caregivers subscales) 4.8); three months, control, mean 3.96 only care, the study demonstrated HRQOL - FACT-B (SD 4.1); intervention, mean 3.56 (SD a significant improvement in Study Duration - Other QOL-Relevant Outcomes - 4.3); group by time p=.002. the intervention group in Six months BHS, coping effectiveness (Brief hopelessness at three months but not six months. The study Risk of Bias - 2 COPE Inventory), mental QOL Caregiver Outcomes: (study-specific scale comprised of did not demonstrate significant mental and emotional SF-36 and None changes in symptom distress, FACT-G subscales) physical QOL, mental QOL, or coping effectiveness.

Caregiver Outcomes: Caregiver QOL-Relevant HRQOL - FACT-G Outcomes Other QOL-Relevant Outcomes - The study did not demonstrate BHS, coping effectiveness (Brief significant changes in COPE Inventory), physical QOL caregiver hopelessness, (study-specific scale comprised of physical QOL, mental QOL, or physical SF-36 and FACT-G coping effectiveness. subscales), mental QOL (study- specific scale comprised of mental and emotional SF-36 and FACT-G subscales)

242

Evidence Table 1: Populations and Interventions

Northouse et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 32 al., 2007 US Criteria: Cancer (prostate) None described Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a family-based inclusion if they had newly education and support Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosed, biochemically intervention for patients with Patients and recurrent, or advanced Not described None described Multidisciplinarity and Clinical FOCUS program, a prostate cancer and their prostate cancer; were at least caregivers, with usual care, families/caregivers Disciplines: family-based 30 years old; had a life the study demonstrated a only Control Arm - No multidisciplinary education and expectancy of at least 12 Caregiver Symptoms: significant improvement in the team; clinical disciplines not support intervention Study Duration - 12 months; had a spouse or live- intervention group in None described specified. months in partner; and lived within 75 communication about the Risk of Bias - 3 miles of participating cancer Intervention Arm - No Intervention illness with a spouse at four centers. They were excluded if multidisciplinary team; nurses. Duration: months but not at eight or 12 they had second primary Four months months. The study did not cancers. demonstrate significant changes in symptom distress, Caregivers were eligible for prostate-specific symptoms, inclusion if they were at least overall physical or mental 21 years old, were identified health, hopelessness, by patients as their primary HRQOL, or coping self- caregiver. They were efficacy. excluded if they were spouses and had been diagnosed with cancer within the prior year or Caregiver QOL-Relevant were receiving cancer Outcomes treatment. The study demonstrated significant improvements in spousal communication about Number of Patient the illness with the patient; Participants and Decedents: caregiver symptom distress, Screened - 429 HRQOL, overall mental health, Enrolled - 263 and hopelessness at four months but not at eight or Completed - 218 twelve months; and caregiver Decedents - 15 coping self-efficacy as four and 12 months but not at eight

months. Number of Caregiver Participants: Screened - 429 Enrolled - 263 Completed - 218

243

Evidence Table 2: Outcomes

Northouse et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 32 al., 2007 US Other Symptoms - Symptom distress Outcomes Study Arms - 2 (Symptom Scale of the OSQ), Communication or Care Planning: Comparing a family-based Level of prostate-specific symptoms (EPIC) Communication about illness with spouse (0- education and support Randomization - Communication or Care Planning - 5 scale, higher is better): four months, intervention for patients with Patients and MISS control, mean 3.69 (SD 0.52); intervention, prostate cancer and their mean 3.80 (SD 0.46); p=.03. caregivers, with usual care, families/caregivers Functional Status - SF-12 PCS only the study demonstrated a HRQOL - FACT-G Caregiver Outcomes: significant improvement in the Study Duration - 12 intervention group in Other QOL-Relevant Outcomes - months communication about the BHS, coping self-efficacy (CASE), Other Symptoms: illness with a spouse at four Risk of Bias - 3 SF-12 MCS Symptom distress (scale not described, months but not at eight or 12 lower is better): four months, control, mean months. The study did not Caregiver Outcomes: 6.28 (SD 3.6); intervention, mean 5.10 (SD demonstrate significant 3.4); p=.006. changes in symptom distress, Other Symptoms - Symptom distress prostate-specific symptoms, (Symptom Scale of the OSQ) Communication or Care Planning: overall physical or mental Communication or Care Planning - health, hopelessness, Communication about illness with spouse (0- MISS 5 scale, higher is better): four months, HRQOL, or coping self- efficacy. HRQOL - FACT-G control, mean 3.57 (SD 0.56); intervention, mean 3.74 (SD 0.53); p=.002; eight months, Other QOL-Relevant Outcomes - control, mean 3.52 (SD 0.56); intervention, BHS, coping self-efficacy (CASE), mean 3.66 (SD 0.54); p=.01; 12 months, Caregiver QOL-Relevant SF-12 MCS control, mean 3.50 (SD 0.56); intervention, Outcomes mean 3.66 (SD 0.54); p=.009. The study demonstrated significant improvements in HRQOL: spousal communication about QOL (scale not described, higher is better): the illness with the patient; four months, control, mean 83.5 (SD 11.4); caregiver symptom distress, intervention, mean 86.5 (SD 11.3); p=.004. HRQOL, overall mental health, and hopelessness at four months but not at eight or Other QOL-Relevant Outcomes: twelve months; and caregiver Hopelessness (0-20 scale, lower is better): coping self-efficacy as four four months, control, mean 3.07 (SD 2.4); and 12 months but not at eight intervention, mean 2.47 (SD 2.1); p=.03. months. Coping self-efficacy (scale not described, higher is better): four months, control, mean 138.8 (SD 22.3); intervention, mean 144.1 (SD 17.8); p=.02; 12 months, control, mean 137.8 (SD 22.6); intervention, mean 143.8 (SD 17.9); p=.02. Overall mental health (0-100 scale, higher is better): four months, control, mean 49.0 (SD 7.5); intervention, mean 50.9 (SD 7.5); p=.03.

244

Evidence Table 1: Populations and Interventions

Oliver et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 74 2001 US Criteria: Cancer (type not Average pain (0-100 scale, Control Arm - Outpatient clinics Usual care, education Outcomes Study Arms - 2 Patients were eligible for described) lower is better): control, mean Intervention Arm - Outpatient clinics session on cancer Comparing a Level of inclusion if they had a 51.8 (SD 15.5); intervention, pain limited to multidimensional, diagnosis of cancer, were mean 53.2 (SD 18.0). discussion about an individualized education and Randomization - Prognosis: Patients between 18 and 75 years old, Impairment due to pain (0-100 Multidisciplinarity and Clinical existing AHRQ coaching session on cancer spoke English, experienced at Not described scale, lower is better): control, Disciplines: guideline pain that addressed Study Duration - least a moderate level of pain knowledge, attitudes, mean 62.2 (SD 18.6); Control Arm - No multidisciplinary Two weeks as indicated by a baseline misconceptions, and problem- intervention, mean 64.3 (SD team; masters-level psychology Verbal Analog Scale average Intervention Arm: solving, with an education Risk of Bias - 7 17.6). student, fourth-year medical pain score of 30 out of 100 or session limited to discussion student. Individualized greater over the previous two Pain frequency (0-100 scale, education and about an existing AHRQ weeks, and had no scheduled lower is better): control, mean Intervention Arm - No coaching session on guideline, the study major surgical procedures 91.7 (SD 19.4); intervention, multidisciplinary team; masters- cancer pain using a demonstrated a significant during the follow-up interval. mean 93.4 (SD 16.6). level psychology student, fourth- variety of improvement in the They were excluded if they year medical student. interpersonal intervention group in average were enrolled in hospice or pain. The study did not Patient Functional Status: techniques, including under the care of the Pain problem solving, demonstrate significant Management Service. SF-12 PCS (0-100 scale, knowledge and changes in impairment due to higher is better): control, mean attitudinal pain, pain frequency, or pain- 29.1 (SD 8.0); intervention, clarification, and related knowledge. Number of Patient mean 26.1 (SD 8.5). facilitation of clinician Participants and Decedents: interactions Screened - 1073 Caregiver Symptoms: Enrolled - 78 None described Intervention Completed - 67 Duration: Decedents - 1 One-time intervention

245

Evidence Table 2: Outcomes

Oliver et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 74 2001 US Pain - VAS, impairment due to pain Outcomes Study Arms - 2 (MOS-PAQ pain effects subscale), Pain: Comparing a Level of pain frequency (MOS-PAQ single multidimensional, item), pain-related knowledge (scale Average pain over the past two weeks individualized education and Randomization - (0-10 scale, lower is better): two weeks Patients adapted from the APS Guidelines for coaching session on cancer the Treatment of Pain Patient vs. baseline, control, mean -0.1; pain that addressed Study Duration - Outcome questionnaire) intervention, mean -0.65; p=.014. knowledge, attitudes, Two weeks Average pain over the past two weeks misconceptions, and problem- Risk of Bias - 7 (0-100 scale, lower is better): two solving, with an education Caregiver Outcomes: weeks vs. baseline, intervention vs. session limited to discussion None described control, mean -8.96 (95% CI -16.0 to - about an existing AHRQ 1.89); p<.05. guideline, the study demonstrated a significant

improvement in the Caregiver Outcomes: intervention group in average N/A pain. The study did not demonstrate significant changes in impairment due to pain, pain frequency, or pain- related knowledge.

246

Evidence Table 1: Populations and Interventions

Porter et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 69 2011 US Criteria: Cancer (lung) None described Control Arm - Home Education and Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home support Comparing an education and Level of inclusion if they had a support intervention with a diagnosis of early-stage lung Prognosis: Patient Functional Status: caregiver-assisted coping Randomization - Intervention Arm: Patients and cancer (NSCLC stages I-III or Not described None described Multidisciplinarity and Clinical skills training intervention for limited-stage small-cell lung patients with lung cancer, the families/caregivers Disciplines: Caregiver-assisted cancer), no other cancers in study demonstrated significant only Control Arm - No multidisciplinary coping skills training the past five years, the ability Caregiver Symptoms: improvements in the education Study Duration - team; physicians, nurses. to read and speak English, None described and support group in Four months and a caregiver who also was Intervention Arm 1 - No Intervention depression, anxiety, functional Risk of Bias - 9 willing to participate. multidisciplinary team; physicians, Duration: well-being, and self-efficacy nurses. Eight months for managing pain, symptoms, and function. The study did Number of Patient not demonstrate significant Participants and Decedents: changes in usual pain, worst Screened - 1098 pain, or HRQOL. Stage I cancer patients benefited Enrolled - 233 more from the education Completed - 140 intervention, while Stage IIII Decedents - 33 cancer patients benefited more from the coping skills training intervention. Number of Caregiver Participants: Caregiver QOL-Relevant Screened - 1098 Outcomes Enrolled - 233 The study demonstrated Completed - 140 significant improvements in caregiver strain and self- efficacy for helping patients manage symptoms. The study did not demonstrate significant changes in caregiver depression or anxiety.

247

Evidence Table 2: Outcomes

Porter et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 69 2011 US Pain - BPI usual pain and worst pain Outcomes Study Arms - 2 scales Depression: Comparing an education and Level of Depression - BDI Depressive symptoms (0-63 scale, lower is support intervention with a better): eight months vs. baseline, education caregiver-assisted coping Randomization - Anxiety - STAI Trait Version Patients and and support vs. coping skills training, linear skills training intervention for Functional Status - FACT-L regression coefficient -2.38 (SE 0.86); group patients with lung cancer, the families/caregivers by time by cancer stage (I vs. IIII) p=.006. only functional well-being subscale study demonstrated significant improvements in the education Study Duration - HRQOL - FACT-L Anxiety: and support group in Four months Other QOL-Relevant Outcomes - depression, anxiety, functional Study-specific self-efficacy for Trait anxiety (20-80 scale, lower is better): Risk of Bias - 9 eight months vs. baseline, education and well-being, and self-efficacy managing pain, symptoms, and support vs. coping skills training, linear for managing pain, symptoms, function scale regression coefficient -8.28 (SE 2.85); group and function. The study did by time by cancer stage (I vs. IIII) p=.005. not demonstrate significant changes in usual pain, worst Caregiver Outcomes: pain, or HRQOL. Stage I Functional Status: Depression - POMS-B depression cancer patients benefited subscale Functional well-being (0-28 scale, higher is more from the education better): eight months vs. baseline, education intervention, while Stage IIII Anxiety - POMS-B tension/anxiety and support vs. coping skills training, linear cancer patients benefited subscale regression coefficient 2.22 (SE 0.80); group by time by cancer stage (I vs. IIII) p=.006. more from the coping skills Other QOL-Relevant Outcomes - training intervention. Strain (CSI), study-specific self-

efficacy for helping patient manage Other QOL-Relevant Outcomes: symptoms scale Self-efficacy for managing pain, symptoms, Caregiver QOL-Relevant and function (10-100 scale, higher is better): Outcomes eight months vs. baseline, education and The study demonstrated support vs. coping skills training, linear regression coefficient 6.26 (SE 3.20); group significant improvements in by time by cancer stage (I vs. IIII) p=.05. caregiver strain and self- efficacy for helping patients manage symptoms. The study Caregiver Outcomes: did not demonstrate significant changes in caregiver Other QOL-Relevant Outcomes: depression or anxiety. Caregiver strain (0-13 scale, lower is better): eight months vs. baseline, education and support vs. coping skills training, linear regression coefficient -0.99 (SE 0.50); group by time by cancer stage (I vs. IIII) p=.05. Self-efficacy for helping patient manage symptoms (10-100 scale, higher is better): eight months vs. baseline, education and support vs. coping skills training, linear regression coefficient 6.48 (SE 3.03); group by time by cancer stage (I vs. IIII) p=.03.

248

Evidence Table 1: Populations and Interventions

Rea et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 102 2004 New Zealand Criteria: CHF, ischemic heart Dyspnea (scale not described, Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for disease, COPD higher is better): control, mean Intervention Arm - Outpatient Comparing a COPD disease Level of inclusion if they had a 16.3; intervention, mean 17.3. clinics, home, hospital management program with diagnosis of moderate to Intervention Arm: usual care, the study Randomization - Prognosis: Fatigue (scale not described, Sites severe COPD. They were higher is better): control, mean Disease management demonstrated a significant excluded if they had chronic Not described 15.3; intervention, mean 15.3. Multidisciplinarity and Clinical program, including improvement in the Study Duration - 12 asthma, bronchiectasis, or Disciplines: patient-specific care intervention group in fatigue. months comorbidity more significant The study did not demonstrate Control Arm - No multidisciplinary plan than COPD; were unable to significant changes in bodily Risk of Bias - 3 Patient Functional Status: team; physicians. give informed consent; were pain, dyspnea, physical None described no longer enrolled with Intervention Arm - Primary care Intervention functioning, mental health, or participating GP practice or team plus respiratory specialist Duration: emotional functioning. physician and respiratory nurse moved out of area; had a Caregiver Symptoms: 12 months prognosis of less than 12 specialist; physicians, nurses. months; were on long-term None described Health Care Use oxygen therapy or were too The study demonstrated a unwell; or were deceased. significant decrease in hospital days for respiratory admissions. The study did not GP practices were excluded if demonstrate significant they were unable to contact changes in all-cause hospital the patient, or had insufficient admissions, respiratory practice nurse resource. hospital admissions, hospital days for all admissions, or ED Number of Patient visits that did not lead to Participants and Decedents: hospital admissions. Screened - 700 Enrolled - 135 Drugs Completed - 117 The study did not demonstrate significant changes in Decedents - 6 prescriptions for antibiotics, oral steroids, inhaled steroids, anticholinergics, β2-agonists, or anticholinergics and β2- agonists combined.

249

Evidence Table 2: Outcomes

Rea et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use Patient QOL-Relevant 102 2004 New Zealand Pain - SF-36 bodily pain subscale hospital, ED Hospital days for respiratory Outcomes Study Arms - 2 Dyspnea - CRQ dyspnea subscale Other Symptoms: Drugs - Prescriptions for admissions: 12 months pre- Comparing a COPD disease Level of antibiotics, oral steroids, inhaled intervention to baseline, control, management program with Other Symptoms - CRQ fatigue Fatigue (scale not described, higher is steroids, anticholinergics, β2- mean 3.5; intervention, mean usual care, the study Randomization - subscale better): baseline, control, mean 15.3; Sites agonists, and anticholinergics 2.8; baseline to 12 months, demonstrated a significant Functional Status - SF-36 physical intervention, mean 15.3; 12 months, and β2-agonists combined control, mean 4.0; intervention, improvement in the Study Duration - 12 functioning subscale control, mean 15.7; intervention, mean mean 1.1; p=.03. intervention group in fatigue. months 17.7; p=.010. The study did not demonstrate Other QOL-Relevant Outcomes - significant changes in bodily Risk of Bias - 3 SF-36 mental health subscale, CRQ pain, dyspnea, physical emotional function subscale Caregiver Outcomes: functioning, mental health, or N/A emotional functioning.

Caregiver Outcomes: Health Care Use None described The study demonstrated a significant decrease in hospital days for respiratory admissions. The study did not demonstrate significant changes in all-cause hospital admissions, respiratory hospital admissions, hospital days for all admissions, or ED visits that did not lead to hospital admissions.

Drugs The study did not demonstrate significant changes in prescriptions for antibiotics, oral steroids, inhaled steroids, anticholinergics, β2-agonists, or anticholinergics and β2- agonists combined.

250

Evidence Table 1: Populations and Interventions

Syrjala et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 87 2008 US Criteria: Cancer (mixed cancer) Usual pain (0-10 scale, lower Control Arm - Outpatient clinics, Nutrition education Outcomes Study Arms - 2 Patients were eligible for is better): nutrition control, home video and written Comparing a pain education inclusion if they were at least mean 4.14 (SD 1.97); pain materials and management intervention Level of Prognosis: Intervention Arm - Outpatient Randomization - 18 years old, had a cancer training, mean 4.07 (SD 1.62); clinics, home with a nutrition education diagnosis with disease-related Not described p>.05. intervention for cancer Patients Intervention Arm: persistent pain; had a life Worst pain (0-10 scale, lower patients, the study Study Duration - expectancy of at least six Multidisciplinarity and Clinical Pain education and demonstrated a significant Six months is better): nutrition control, months, had ambulatory mean 5.97 (SD 2.63); pain Disciplines: management video improvement in the pain functional status, were education and management Risk of Bias - 0 training, mean 5.74 (SD 2.30); Control Arm - No multidisciplinary and written materials receiving cancer treatment group in usual pain. The study p>.05. team; physicians, nurses. that was expected to be stable did not demonstrate significant over the next six months, and Intervention Arm - No Intervention changes in worst pain or had English reading and Patient Functional Status: multidisciplinary team; physicians, Duration: overall symptom severity. writing proficiency adequate to nurses. ECOG performance status (0- One-time intervention participate in the intervention 5 scale, lower is better): and assessments. They were Drugs nutrition control, mean 1.03 excluded if they had active (SD 0.79); pain training, mean The study demonstrated a alcohol or other substance 1.30 (SD 0.77); p>.05. significant increase in the pain abuse or a major psychiatric education and management diagnosis for which treatment group in opioid use. was being received. Caregiver Symptoms: None described Number of Patient Participants and Decedents: Screened - 226 Enrolled - 93 Completed - 78 Decedents - 10

251

Evidence Table 2: Outcomes

Syrjala et al., Study Location - Patient Outcomes: Patient Outcomes: Drugs - Opioid use Drugs Patient QOL-Relevant 87 2008 US Pain - BPI usual pain and worst pain Total daily morphine equivalent Outcomes Study Arms - 2 scales Pain: dose (log 10 dose ratio): pain Comparing a pain education Level of Other Symptoms - Symptom severity training vs. nutrition control, one and management intervention Usual pain (0-10 scale, lower is better): month vs. baseline, 0.06; two with a nutrition education Randomization - (MSAS) six months vs. baseline, pain training Patients months vs. baseline, 0.30; three intervention for cancer vs. nutrition control, mean -0.81 (SE months vs. baseline, 0.49; patients, the study Study Duration - Caregiver Outcomes: 0.36); p=.03. p<.001. demonstrated a significant Six months improvement in the pain None described education and management Risk of Bias - 0 Caregiver Outcomes: group in usual pain. The study N/A did not demonstrate significant changes in worst pain or overall symptom severity.

Drugs The study demonstrated a significant increase in the pain education and management group in opioid use.

252

Evidence Table 1: Populations and Interventions

Vallieres et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 76 al., 2006 Canada Criteria: Cancer (mixed cancer) Average pain over past 48 Control Arm - Outpatient Clinics Usual care Outcomes Study Arms - 2 Patients were eligible for hours (0-10 scale, lower is Intervention Arm - Outpatient Comparing a pain Level of inclusion if they had cancer, better): control, mean 4.5 (SD Clinics management intervention with spoke French, were at least Prognosis: 2.2); intervention, mean 5.2 Intervention Arm: usual care for cancer patients Randomization - Patients 18 years old, were Not described (SD 2.2); p>.05. Pain education, diary, undergoing radiotherapy, the experiencing any pain or Multidisciplinarity and Clinical and access to doctor study demonstrated significant Study Duration - Worst pain over past 48 hours taking analgesics, were (0-10 scale, lower is better): Disciplines: to adjust medication improvements in the Three weeks treated with radiotherapy for intervention group in average control, mean 6.9 (SD 2.7); Control Arm - No multidisciplinary Risk of Bias - 0 more than one week (greater intervention, mean 7.3 (SD pain and worst pain. team; physicians. Intervention than five fractions) on an 2.4); p>.05. outpatient basis. Intervention Arm - No Duration: Weakest pain over past 48 multidisciplinary team; physicians. Three weeks hours (0-10 scale, lower is Number of Patient better): control, mean 1.7 (SD Participants and Decedents: 1.4); intervention, mean 2.4 Screened - 72 (SD 2.3); p>.05. Enrolled - 64 Completed - Not described Patient Functional Status: Decedents - Not described Not described

Caregiver Symptoms: None described

253

Evidence Table 2: Outcomes

Vallieres et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 76 al., 2006 Canada Pain - APS pain scale Outcomes Study Arms - 2 Pain: Comparing a pain Level of management intervention with Caregiver Outcomes: Average pain over past 48 hours (0-10 usual care for cancer patients Randomization - scale, lower is better): baseline, Patients None described undergoing radiotherapy, the control, mean 4.5 (SD 2.2); study demonstrated significant Study Duration - intervention, mean 5.2 (SD 2.2); three improvements in the Three weeks weeks, control, mean 4.4 (SD 2.6); intervention group in average Risk of Bias - 0 intervention, 2.9 (SD 2.4); group by pain and worst pain. time p<.001. Worst pain over past 48 hours (0-10 scale, lower is better): baseline, control, mean 6.9 (SD 2.7); intervention, mean 7.3 (SD 2.4); three weeks, control, mean 5.5 (SD 2.7); intervention, 4.2 (SD 2.9); group by time p=.03.

Caregiver Outcomes: N/A

254

Evidence Table 1: Populations and Interventions

Ward et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 33 2008 US Criteria: Cancer (mixed cancer) Pain severity (0-10 scale, Control Arm - Outpatient clinics, Standard pain Outcomes Study Arms - 2 Patients were eligible for lower is better): control, mean home educational Comparing the RIDcancerPain inclusion if they were at least 3.05 (SD 2.11); intervention, information pain educational intervention Level of Prognosis: Intervention Arm - Outpatient Randomization - 21 years old, had a diagnosis mean 2.63 (SD 1.99). clinics, home with standard pain educational of metastatic cancer, and had Not described information for cancer Patients Pain severity (0-3 scale, lower Intervention Arm: a rating of one or greater on a is better): control, mean 1.52 patients, the study Study Duration - 0 to 10 scale for worst cancer Multidisciplinarity and Clinical RIDcancerPain pain demonstrated mixed effects in Two months (SD 0.80); intervention, mean related pain experienced 1.43 (SD 0.78). Disciplines: education the intervention group on pain during the past two weeks. severity. The study did not Risk of Bias - 1 Control Arm - No multidisciplinary intervention They were excluded if they demonstrate a significant team; nurses. were unable to read English, Patient Functional Status: change in overall well-being. were incarcerated, or had an Intervention Arm - No Intervention ECOG performance status (0- ECOG performance status multidisciplinary team; nurses. Duration: 5 scale, lower is better): all score of 3 or 4. participants, 0 to 2. One-time intervention

Number of Patient Caregiver Symptoms: Participants and Decedents: None described Screened - 403 Enrolled - 222 Completed - 150 Decedents - 22

255

Evidence Table 2: Outcomes

Ward et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 33 2008 US Pain - TPQM adapted pain severity Outcomes Study Arms - 2 item; mean of BPI worst pain, least Pain: Comparing the RIDcancerPain Level of pain, and pain now scales pain educational intervention Pain severity (0-3 scale, lower is with standard pain educational Randomization - HRQOL - QLI-CV better): baseline, control, mean 1.52 Patients information for cancer (SD 0.80); intervention, mean 1.43 (SD patients, the study Study Duration - Caregiver Outcomes: 0.78); two months, control, mean 1.47 demonstrated mixed effects in Two months (SD 0.77); intervention, mean 1.39 (SD the intervention group on pain None described Risk of Bias - 1 0.75); p=.009. severity. The study did not demonstrate a significant change in overall well-being. Caregiver Outcomes: N/A

256

Evidence Table 1: Populations and Interventions

Ward et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 77 2009 US Criteria: Cancer (mixed cancer) Pain severity (0-100 scale, Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 3 Patients were eligible for lower is better): all Intervention Arm 1 - Outpatient Comparing a pain Level of inclusion if they were at least participants, mean 49.67 (SD clinics management and planning 18 years old, were diagnosed Prognosis: 9.96). Intervention Arm 1: intervention in dyads or Randomization - Intervention Arm 2 - Outpatient Patients and with cancer, had moderate to Not described RIDcancerPAIN+, a patients only with usual care severe pain in the past two clinics pain management for cancer patients, the study families/caregivers Patient Functional Status: only weeks, had an ECOG score of education and demonstrated significant 0-2, and had a caregiver who improvements in pain relief ECOG performance status (0- Multidisciplinarity and Clinical planning intervention Study Duration - could participate in the study. from pain management 5 scale, lower is better): all Disciplines: delivered in dyads Nine weeks They were excluded if they activities in the dyad group at participants, 0-2. could not read and write Control Arm - No multidisciplinary five weeks and negative Risk of Bias - 2 English. team; clinical disciplines not Intervention Arm 2: attitudes about reporting pain Caregiver Symptoms: specified. RIDcancerPAIN+, a and using analgesics in both Caregivers were eligible for None described Intervention Arm 1 - No pain management the dyad and patient only inclusion if they were at least multidisciplinary team; nurses, education and groups at nine weeks. The 18 years old and were psychologists. planning intervention study did not demonstrate identified by the patient as a significant changes in pain Intervention Arm 2 - No delivered to patients friend, spouse, or other family only severity, pain interference, member who had the most multidisciplinary team; nurses, global HRQOL, or negative influence on the patient’s psychologists. mood. illness and treatment. Intervention Duration: Caregiver QOL-Relevant Number of Patient Four weeks Outcomes Participants and Decedents: The study did not demonstrate Screened - 543 a significant change in Enrolled - 161 negative attitudes about reporting pain and using Completed - 109 analgesics. Decedents - 25

Number of Caregiver Participants: Screened - 543 Enrolled - 161 Completed - 109

257

Evidence Table 2: Outcomes

Ward et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 77 2009 US Pain - Pain relief single item, pain Outcomes Study Arms - 3 severity composite score, pain Pain: Comparing a pain Level of interference (BPI-SF) management and planning Pain relief from pain management intervention in dyads or Randomization - HRQOL - Global QOL composite activities in the past week (0-4 scale, Patients and score patients only with usual care lower is better): five weeks vs. for cancer patients, the study families/caregivers Other QOL-Relevant Outcomes - baseline, control, mean -0.05 (SD only demonstrated significant Negative attitudes about reporting 0.90); RIDcancerPAIN+ dyad, mean - improvements in pain relief Study Duration - pain and using analgesics (BQ-II), 0.22 (SD 0.94); RIDcancerPAIN+ from pain management Nine weeks EORTC QLQ-C30 mood subscale patient only, mean +0.24 (SD 0.84); activities in the dyad group at Risk of Bias - 2 p=.008. five weeks and negative attitudes about reporting pain Caregiver Outcomes: and using analgesics in both Other QOL-Relevant Outcomes: Other QOL-Relevant Outcomes - the dyad and patient only Negative attitudes about reporting Negative attitudes about reporting pain groups at nine weeks. The pain and using analgesics (BQ-II) and using analgesics (scale not study did not demonstrate described, higher is better): nine weeks significant changes in pain vs. baseline, control, mean +0.08 (SD severity, pain interference, 0.75); RIDcancerPAIN+ dyad, mean global HRQOL, or negative +0.72 (SD 0.88); RIDcancerPAIN+ mood. patient only, mean +0.43 (SD 0.64);

p<.05. Caregiver QOL-Relevant

Outcomes Caregiver Outcomes: The study did not demonstrate None a significant change in negative attitudes about reporting pain and using analgesics.

258

Evidence Table 1: Populations and Interventions

Ward et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 120 2009 US Criteria: Cancer (mixed cancer) Pain severity (0-10 scale, Control Arm - Home Usual care Outcomes Study Arms - 3 Patients were eligible for lower is better): assessment Intervention Arm 1 - Home Comparing an educational inclusion if they were CIS only, mean 3.63 (SD 2.20); intervention tailored to cancer Level of Prognosis: Intervention Arm 2 - Home Intervention Arm 1: Randomization - callers, could participate in TBI, mean 3.65 (SD 2.30). patients’ attitudinal barriers to Patients English, were at least 18 Not described Pain duration (1-5 scale, lower Baseline assessment reporting pain and using years old, were diagnosed analgesics with baseline is better): control, mean 3.00 Multidisciplinarity and Clinical of pain and attitudinal Study Duration - 28 with cancer, and had barriers to reporting assessment only and usual days (SD 1.13); assessment only, Disciplines: moderate to severe pain mean 3.06 (SD 1.14); TBI, pain and using care, the study demonstrated Control Arm - No multidisciplinary Risk of Bias - 3 related to cancer in the past mean 3.03 (SD 1.14). analgesics a significant improvement in week. They were excluded if team; clinical disciplines not the tailored intervention group

they had received pain Pain interference (0-10 scale, specified. in attitudinal barriers to lower is better): assessment information as part of usual Intervention Arm 1 - No Intervention Arm 2: reporting pain and using only, mean 5.03 (SD 2.85); CIS service or had previously multidisciplinary team; clinical TBI, an educational analgesics. The study did not TBI, mean 4.87 (SD 2.88). called during the study accrual disciplines not specified. intervention tailored demonstrate significant period. to participants’ changes in pain severity, Intervention Arm 2 - No duration, or interference. Patient Functional Status: multidisciplinary team; clinical attitudinal barriers to reporting pain and Number of Patient disciplines not specified. Not described using analgesics Participants and Decedents: Health Care Costs

Screened - 4318 The study demonstrated Caregiver Symptoms: Intervention Enrolled - 1256 similar costs of delivering the None described Duration: interventions in the tailored Completed - 970 One-time intervention intervention and baseline Decedents - 5 assessment only groups, but significance was not evaluated.

259

Evidence Table 2: Outcomes

Ward et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Costs - Intervention None Patient QOL-Relevant 120 2009 US Pain - Pain severity (three intensity costs Outcomes Study Arms - 3 items from the BPI), pain duration Other QOL-Relevant Outcomes: Comparing an educational single item, pain interference (seven intervention tailored to cancer Level of Attitudinal barriers to reporting pain Randomization - interference items from the BPI) patients’ attitudinal barriers to and using analgesics (0-8 scale, lower reporting pain and using Patients Other QOL-Relevant Outcomes - is better): control, 28 days, mean 4.01 BQ-II analgesics with baseline Study Duration - 28 (SD 2.01); assessment only, baseline, assessment only and usual days mean 4.28 (SD 1.86); 28 days, mean care, the study demonstrated Risk of Bias - 3 Caregiver Outcomes: 3.90 (SD 2.00); TBI, baseline, mean a significant improvement in 4.13 (SD 1.86); 28 days, mean 2.98 the tailored intervention group None described (SD 1.95); TBI vs. control, p<.001; TBI in attitudinal barriers to vs. assessment only, p<.001. reporting pain and using analgesics. The study did not demonstrate significant Caregiver Outcomes: changes in pain severity, N/A duration, or interference.

Health Care Costs The study demonstrated similar costs of delivering the interventions in the tailored intervention and baseline assessment only groups, but significance was not evaluated.

260

Evidence Table 1: Populations and Interventions

Wells et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 34 2003 US Criteria: Cancer (type not Average pain (0-10 scale, Control Arm - Not described Usual care, pain Outcomes Study Arms - 3 Patients were eligible for described) lower is better): all Intervention Arm 1 - Home education video Comparing a pain education Level of inclusion if they were participants, mean 4.18 (SD video plus routine provider- experiencing cancer pain that 2.10). Intervention Arm 2 - Home initiated telephone follow-up Randomization - Prognosis: Intervention Arm 1: Patients and was managed with an Least pain (0-10 scale, lower calls, access to a pain hot line, analgesic (opioid or non- or nothing further, the study families/caregivers Not described is better): all participants, Multidisciplinarity and Clinical Pain education video opioid), had onset or demonstrated a significant only mean 2.66 (SD 2.23). Disciplines: and access to a pain escalation of pain within the hot line change in all groups in Study Duration - past three months, had Worst pain (0-10 scale, lower Control Arm - No multidisciplinary attitudinal barriers to

Six months access to a telephone at is better): all participants, team; clinical disciplines not communication with providers Risk of Bias - 0 home, had a primary caregiver mean 6.35 (SD 2.50). specified. Intervention Arm 2: about pain directly following who was willing to participate Pain interference (0-10 scale, Intervention Arm 1 - No Pain education video the video. The study did not in the study, and were at least lower is better): all participants multidisciplinary team; nurses. and routine provider- demonstrate significant 18 years old. They were mean 4.75 (SD 2.64). initiated telephone changes in attitudinal barriers excluded if they had Intervention Arm 2 - No to analgesic use or in pain follow-up calls documented cognitive multidisciplinary team; nurses. intensity, interference, or impairment or a life Patient Functional Status: relief. expectancy of less than six KPS score (0-100 scale, Intervention months. higher is better): control, mean Duration: Caregiver QOL-Relevant 94.32; hot line, mean 92.50; One month Outcomes weekly call, mean 95.63; Number of Patient p>.05. The study demonstrated a Participants and Decedents: significant improvement in all

Screened - 84 groups in caregiver beliefs Caregiver Symptoms: Enrolled - 64 about using analgesics directly None described following the video. Completed - Not described

Decedents - 8 Drugs

The study did not demonstrate Number of Caregiver a significant change in Participants: analgesic use. Screened - Not described Enrolled - 64 Completed - Not described

261

Evidence Table 2: Outcomes

Wells et al., Study Location - Patient Outcomes: Patient Outcomes: Drugs - Analgesic use None Patient QOL-Relevant 34 2003 US Pain - BPI-SF pain intensity, Outcomes Study Arms - 3 interference, and relief scales Other QOL-Relevant Outcomes: Comparing a pain education Level of Other QOL-Relevant Outcomes - video plus routine provider- Attitudinal barriers to communication initiated telephone follow-up Randomization - BQ-r communication subscale, BQ-r with providers about pain (0-5 scale, Patients and analgesic subscale calls, access to a pain hot line, lower is better): all participants, or nothing further, the study families/caregivers baseline, mean 1.79 (SD 0.94); post- only demonstrated a significant Caregiver Outcomes: education video, mean 1.23 (SD 0.97); change in all groups in Study Duration - p<.001. attitudinal barriers to Other QOL-Relevant Outcomes - Six months communication with providers FPQ knowledge/attitude subscale about pain directly following Risk of Bias - 0 Caregiver Outcomes: the video. The study did not demonstrate significant Other QOL-Relevant Outcomes: changes in attitudinal barriers to analgesic use or in pain Beliefs about using analgesics (0-10 intensity, interference, or scale, higher is better): all participants, relief. baseline, mean 5.19 (SD 2.26); post- education video, mean 6.21 (SD 1.87); p<.001. Caregiver QOL-Relevant Outcomes The study demonstrated a significant improvement in all groups in caregiver beliefs about using analgesics directly following the video.

Drugs The study did not demonstrate a significant change in analgesic use.

262

Evidence Table 1: Populations and Interventions

Anderson et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 35 al., 2004 US Criteria: Cancer (mixed cancer) Severe pain (BPI score of 7 or Control Arm - Outpatient, clinics, Nutrition education Outcomes Study Arms - 2 Patients were eligible for higher, 0-10 scale, lower is home Comparing a pain education inclusion if they were an better): control, 72%; intervention with nutrition Level of Prognosis: Intervention Arm - Outpatient, Intervention Arm: Randomization - outpatient at the participating intervention, 68%. clinics, home education for minority cancer Patients institutions were a member of Not described Pain management patients, the study did not a black/African American or education demonstrate significant Study Duration - 10 Hispanic minority group, had a Patient Functional Status: changes in pain intensity, pain Multidisciplinarity and Clinical weeks socioeconomic disadvantage ECOG performance status (0- Disciplines: interference, perceived control as confirmed by a billing code Intervention over pain, adequacy of Risk of Bias - 4 5 scale, lower is better): 0 or Control Arm - No multidisciplinary indicating low-income status, Duration: analgesic prescriptions, 1, nutrition education, 72%; team; nurses. had a pathologic diagnosis of pain education, 56%; p=.12. Three days overall physical or mental cancer, had chronic cancer- Intervention Arm - No health, or HRQOL. SF-12 PCS (0-100 scale, related pain, had a pain worst multidisciplinary team; nurses. higher is better): nutrition score on the BPI of at least 4, education, mean 31.26 (SD had no major surgery within 8.53); pain education, mean the past 30 days, had an 27.93 (SD 8.34); p>.05. ECOG performance status of 0-2, and were at least 18 years old. Caregiver Symptoms: None described Number of Patient Participants and Decedents: Screened - 1000 Enrolled - 101 Completed - 67 Decedents - 31

263

Evidence Table 2: Outcomes

Anderson et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 35 al., 2004 US Pain - BPI pain interference and pain Outcomes Study Arms - 2 intensity scales, perceived control Comparing a pain education Level of over pain (Pain Control Scale of the intervention with nutrition Randomization - Survey of Pain Attitudes), adequacy education for minority cancer Patients of analgesic prescriptions (PMI) patients, the study did not Functional Status - SF-12 PCS demonstrate significant Study Duration - 10 changes in pain intensity, pain weeks HRQOL - SF-12 interference, perceived control Risk of Bias - 4 Other QOL-Relevant Outcomes - over pain, adequacy of SF-12 MCS analgesic prescriptions, overall physical or mental

health, or HRQOL. Caregiver Outcomes: None described

264

Evidence Table 1: Populations and Interventions

Beauchamp Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant et al., US Criteria: Dementia None described Control Arm - Not described Usual care Outcomes 2005111 Study Arms - 2 Caregivers were eligible for Intervention Arm - Home Comparing a web-based inclusion if they were multimedia intervention that Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - employed at least part-time modeled positive caregiving Families or and had at least four contacts Not described Not described Multidisciplinarity and Clinical Web-based strategies with usual care for a month caring for a family caregivers of dementia caregivers Disciplines: multimedia member with substantial patients, the study Control Arm - No multidisciplinary intervention that Study Duration - 30 memory problems. Caregiver Symptoms: demonstrated significant team; clinical disciplines not modeled positive days improvements in the Depressive symptoms (0-27 specified. caregiving strategies Risk of Bias - 0 scale, lower is better): control, intervention group in caregiver Number of Caregiver mean 15.4 (SD 10.9); Intervention Arm - No depressive symptoms, state Participants: intervention, mean 16.8 (SD multidisciplinary team; clinical Intervention anxiety, stress, strain, gain, Screened - 329 11.3). disciplines not specified. Duration: and self-efficacy. Enrolled - 307 State anxiety (0-30 scale, 30 days Completed - 299 lower is better): control, mean 14.7 (SD 6.4); intervention, mean 15.6 (SD 5.8).

265

Evidence Table 2: Outcomes

Beauchamp Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant et al., US None described N/A Outcomes 2005111 Study Arms - 2 Comparing a web-based multimedia intervention that Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - modeled positive caregiving Families or Depression - Study-specific scale strategies with usual care for caregivers of dementia caregivers comprised of eight CES-D items and Depression: one item assessing suicidal ideation patients, the study Study Duration - 30 based on CES-D scale format Depressive symptoms (0-27 scale, demonstrated significant days lower is better): control, baseline, improvements in the Anxiety - STAI Risk of Bias - 0 mean 15.4 (SD 10.9); 30 days, mean intervention group in caregiver Other QOL-Relevant Outcomes - 16.4 (SD 11.2); intervention, baseline, depressive symptoms, state Study-specific caregiver stress mean 16.8 (SD 11.3); 30 days, mean anxiety, stress, strain, gain, measure, Caregiver Strain 15.4 (SD 9.7); group by time p=.009. and self-efficacy. Instrument, Positive Aspects of Caregiving survey, study-specific caregiver self-efficacy measure Anxiety: State anxiety (0-30 scale, lower is better): control, baseline, mean 14.7 (SD 6.4); 30 days, mean 15.6 (SD 6.4); intervention, baseline, mean 15.6 (SD 5.8); 30 days, mean 14.7 (SD 5.6); group by time p=.030.

Other QOL-Relevant Outcomes: Caregiver stress (scale not described, lower is better): control, baseline, mean 12.3 (SD 7.0); 30 days, mean 11.6 (SD 6.4); intervention, baseline, mean 14.2 (SD 7.7); 30 days, mean 10.8 (SD 6.2); group by time p<.001. Caregiver strain (0-70 scale, lower is better): control, baseline, mean 39.8 (SD 9.7); 30 days, mean 39.9 (SD 9.7); intervention, baseline, mean 41.2 (SD 10.3); 30 days, mean 39.1 (SD 9.5); group by time p=.028. Caregiver gain (9-45 scale, higher is better): control, baseline, mean 30.8 (SD 6.7); 30 days, mean 30.9 (SD 6.7); intervention, baseline, mean 30.9 (SD 7.0); 30 days, mean 32.6 (SD 7.1); group by time p=.021. Caregiver self-efficacy (0-42 scale, higher is better): control, baseline, mean 22.7 (SD 7.9); 30 days, mean 23.5 (SD 8.1); intervention, baseline, mean 23.2 (SD 7.9); 30 days, mean 25.7 (SD 7.4); group by time p=.016.

266

Evidence Table 1: Populations and Interventions

Boesen et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 78 al., 2005 EU Criteria: Cancer (cutaneous Depression (scale not Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for malignant melanoma) described, lower is better): Intervention Arm - Outpatient clinics Comparing a health Level of inclusion if they were 18 to 70 control, mean 5.9 (SD 7.8); management and years old, had T1-4, N1a-2a, intervention, mean 5.2 (SD Intervention Arm: psychological support Randomization - Prognosis: Patients M0 cutaneous malignant 6.9); p=.43. Multidisciplinarity and Clinical Health education, intervention with usual care for melanoma diagnosed and Not described Fatigue (scale not described, Disciplines: enhancement of patients with cutaneous Study Duration - 12 treated in eastern Denmark, malignant melanoma, the lower is better): control, mean Control Arm - No multidisciplinary problem-solving months and had undergone surgery study demonstrated significant 5.2 (SD 4.8); intervention, team; clinical disciplines not skills, stress within 12 weeks of study improvements in the Risk of Bias - 2 mean 5.1 (SD 3.9); p=.92. specified. management, and participation. psychological support intervention group in total Confusion (scale not Intervention Arm - intervention mood disturbance, fatigue, described, lower is better): Psychoeducational intervention vigor, behavioral coping, and

Number of Patient control, mean 1.9 (SD 4.1); team; physician, nurse, and cognitive coping at six months Participants and Decedents: intervention, mean 1.8 (SD psychologist Intervention but not at 12 months. The Screened - 420 4.2); p=.84. Duration: study did not demonstrate significant changes in Enrolled - 262 Anxiety (scale not described, Six weeks lower is better): control, mean depression, anxiety, Completed - 241 4.9 (SD 6.6); intervention, confusion, or avoidance Decedents - 1 mean 4.2 (SD 5.8); p=.42. coping.

Patient Functional Status: Not described

Caregiver Symptoms: None described

267

Evidence Table 2: Outcomes

Boesen et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 78 al., 2005 EU Depression - POMS depression Outcomes Study Arms - 2 subscale Other Symptoms: Comparing a health Level of Anxiety - POMS anxiety subscale management and Total mood disturbance (scale not psychological support Randomization - Other Symptoms - POMS total mood described, lower is better): six months Patients intervention with usual care for disturbance, fatigue, and confusion vs. baseline, control, mean -2.64 (SD patients with cutaneous Study Duration - 12 subscales 22.1); intervention, mean -8.43 (SD malignant melanoma, the months Other QOL-Relevant Outcomes - 24.3); p=.04. study demonstrated significant Risk of Bias - 2 POMS vigor subscale; DWI-R Fatigue (scale not described, lower is improvements in the behavioral coping, cognitive coping, better): six months vs. baseline, intervention group in total and avoidance coping subscales control, mean +0.46 (SD 4.6); mood disturbance, fatigue, vigor, behavioral coping, and intervention, mean -0.80 (SD 4.3); p=.04. cognitive coping at six months Caregiver Outcomes: but not at 12 months. The

None described study did not demonstrate Other QOL-Relevant Outcomes: significant changes in Vigor (scale not described, higher is depression, anxiety, better): six months vs. baseline, confusion, or avoidance control, mean -0.22 (SD 4.5); coping. intervention, mean +2.36 (SD 6.6); p=.003. Behavioral coping (scale not described, higher is better): six months vs. baseline, control, mean -1.33 (SD 6.4); intervention, mean +1.81 (SD 6.5); p<.001. Cognitive coping (scale not described, higher is better): six months vs. baseline, control, mean -3.40 (SD 8.6); intervention, mean -0.01 (SD 6.9); p<.001.

Caregiver Outcomes: N/A

268

Evidence Table 1: Populations and Interventions

Bourgeois et Study Location - Inclusion and Exclusion Criteria: Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 63 al., 2002 US Patients were eligible for inclusion if Dementia None described Control Arm - Home Visitation control Outcomes they were at least 45 years old; met Study Arms - 3 the ADRDA-NINCDS criteria for Intervention Arm 1 - Home group: workshop that Comparing a workshop focused Level of probable Alzheimer’s disease; scored emulated a typical on changing Alzheimer’s 20 or less on the MMSE within four Prognosis: Patient Functional Status: Intervention Arm 2 - Home Randomization - support group disease patient habits (patient- weeks of enrollment in the study; meeting with change), a workshop focused Patients and obtained an average score (from Not described OARS physical ADL questions supporting materials, on changing caregiver habits families/caregivers primary and secondary caregiver (0-14 scale, higher is better): Multidisciplinarity and Clinical ratings) of at least 1.0 (mild behavioral weekly home visits (self-change), and a workshop only visitation control, mean 9.60 Disciplines: disturbance) on the Global Rating (SD 3.54); self-change, mean that emulated a typical support Item of the BEHAVE-AD; Study Duration - 9.90 (SD 3.09); patient- Control Arm - No multidisciplinary group meeting (visitation Nine months demonstrated a stable medical Intervention Arm 1: condition as determined by medical change, mean 9.59 (SD 3.09). team; clinical disciplines not control), all of which included Risk of Bias - 3 chart review of clinical history, specified. Self-change group: supporting materials and physical examination, and laboratory Intervention Arm 1 - No workshop focused on follow-up weekly visits, the evidence; and resided with a spousal Caregiver Symptoms: study demonstrated caregiver. They were excluded if they multidisciplinary team; clinical changing caregiver had a documented history of None described disciplines not specified. habits with supporting inconsistent effects by alcoholism, schizophrenia, materials, weekly outcome, group, and time. The Intervention Arm 2 - No Parkinson’s disease, head trauma home visits study demonstrated a with cognitive sequelae, or MRI or CT multidisciplinary team; clinical significant improvement in evidence of focal stroke. disciplines not specified. caregiver mood in the self- Caregivers were eligible for inclusion Intervention Arm 2: change group at three, six, and if they were spouses, did not Patient-change nine months, and in the patient- demonstrate any major debilitating change group at three and nine health problems, agreed to random group: workshop assignment to the study conditions, focused on changing months but not at six months. spent at least eight daytime hours in patient behaviors with The study demonstrated a the home with the patient, and supporting materials, significant improvement in identified a secondary caregiver caregiver strain in the self- (either an adult relative, neighbor, or weekly home visits friend) who spent at least four hours change group at six and nine per week of direct with the patient and months but not at three caregiver and who agreed to Intervention months, and in the patient- complete a subsample of the change group at six months but assessment battery at each of the Duration: not at three or nine months. four assessment periods. Three months The study demonstrated

significant improvements in Number of Patient Participants and caregiver depression in both Decedents: patient- and self-change Screened - 93 groups, and in caregiver Enrolled - 76 perceived stress in the patient- Completed - 63 change group, at three months Decedents - Not described but not at six or nine months. The study did not demonstrate significant changes in caregiver Number of Caregiver Participants: anxiety, anger, or perceived Screened - 93 health. Enrolled - 76 Completed - 63

269

Evidence Table 2: Outcomes

Significant First Clinical Significant Clinical Economic Author, Quality of Life Measures and Economic Measures and Year Study Elements Outcome(s) and Measures Result(s) Outcome(s) Result(s) Overall Study Result

Bourgeois et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 63 al., 2002 US None described N/A Outcomes Study Arms - 3 Comparing a workshop focused on changing Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - Alzheimer’s disease patient Patients and Depression - CES-D habits (patient-change), a workshop focused on changing families/caregivers Anxiety - STAI Depression: only caregiver habits (self-change), Other QOL-Relevant Outcomes - Depression (0-60 scale, lower is better): baseline, and a workshop that emulated Study Duration - Caregiver Strain Scale, Mood Rating visitation control, mean 12.4 (SD 7.6); self-change, a typical support group Nine months Scale, Perceived Stress Scale, mean 12.5 (SD 8.6); patient-change, mean 10.0 (SD meeting (visitation control), all Risk of Bias - 3 Spielberger Anger Expression Scale, 5.5); three months, visitation control, mean 13.5 (SD of which included supporting perceived health (Caregiver Health 6.5); self-change, mean 10.1 (SD 5.7); patient- materials and follow-up weekly Index) change, mean 9.4 (SD 5.9); three months, self- visits, the study demonstrated change vs. visitation control, p<.05; patient-change inconsistent effects by vs. visitation control, p<.05. outcome, group, and time. The study demonstrated a significant improvement in Other QOL-Relevant Outcomes: caregiver mood in the self- Caregiver strain (0-164 scale, lower is better): change group at three, six, and baseline, visitation control, mean 71.9 (SD 17.2); nine months, and in the self-change, mean 68.6 (SD 16.3); patient-change, patient-change group at three mean 62.9 (SD 19.0); six months, visitation control, and nine months but not at six mean 77.8 (SD 19.1); self-change, mean 66.1 (SD months. The study 15.7); patient-change, mean 65.2 (SD 19.3); nine demonstrated a significant months, visitation control, mean 78.0 (SD 19.8); self- improvement in caregiver change, mean 63.2 (SD 19.8); six months, self- strain in the self-change group change vs. visitation control, p<.05; patient-change at six and nine months but not vs. visitation control, p<.01; nine months, self- at three months, and in the change vs. visitation control, p<.01. patient-change group at six months but not at three or nine Mood (1-9 scale, higher is better): baseline, months. The study visitation control, mean 5.6 (SD 1.1); self-change, demonstrated significant mean 5.2 (SD 2.4); patient-change, mean 6.1 (SD improvements in caregiver 1.3); three months, visitation control, mean 5.5 (SD depression in both patient- and 1.5); self-change, mean 6.8 (SD 1.3); patient- self-change groups, and in change, mean 6.5 (SD 1.4); six months, visitation caregiver perceived stress in control, mean 4.9 (SD 1.9); self-change, mean 6.2 the patient-change group, at (SD 1.4); nine months, visitation control, mean 4.5 three months but not at six or (SD 1.9); self-change, mean 6.3 (SD 1.4); patient- nine months. The study did not change, mean 6.7 (SD 1.5); three months, self- demonstrate significant change vs. visitation control, p<.05. changes in caregiver anxiety, Perceived stress: baseline, visitation control, mean anger, or perceived health. 6.1 (SD 2.9); self-change, mean 6.7 (SD 3.8); patient-change, mean 5.2 (SD 2.4); three months, visitation control, mean 5.8 (SD 2.3); patient- change, mean 4.0 (SD 2.3); three months, patient- change vs. visitation control, p<.01.

270

Evidence Table 1: Populations and Interventions

Cornbleet et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 136 al., 2002 EU Criteria: Cancer (mixed cancer) None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Not described Comparing a patient-held inclusion if they had a record intervention with usual Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of cancer, were at care in cancer patients, the Patients least 18 years old, were Identified by clinicians as Ability to do light activities Multidisciplinarity and Clinical Patient-held record study did not demonstrate any attending an oncology significant changes in poor prognosis, referred (scale not described): able to Disciplines: Study Duration - outpatient clinic or receiving satisfaction with to or receiving hospice do light activities, control, Control Arm - No multidisciplinary Six months hospice home care or day Intervention communication, including 75%; intervention, 76%; team; clinical disciplines not care, had a prognosis of at Duration: communication with primary Risk of Bias - 4 unable to do light activities, specified. least 6 months, and were able control, 25%; intervention, Six months care staff, communication with to read English. They were 24%; p=1.00. Intervention Arm - No the main outpatient doctor, excluded if they had a multidisciplinary team; clinical views of communication

psychological condition or disciplines not specified. between hospital staff, views sensory impairment that would Caregiver Symptoms: of communication between the make an interview unwise or None described GP and hospital doctors, or impossible. satisfaction with communication between all hospice and home care staff Number of Patient involved in the patient’s care. Participants and Decedents: Screened - Not described Enrolled - 231 Completed - 177 Decedents - 38

271

Evidence Table 2: Outcomes

Cornbleet et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 136 al., 2002 EU Communication or Care Planning - Outcomes Study Arms - 2 Undescribed communication with Comparing a patient-held Level of primary care staff measure, record intervention with usual Randomization - undescribed communication with care in cancer patients, the Patients main outpatient doctor measure, study did not demonstrate any undescribed views of communication significant changes in Study Duration - between hospital staff measure, satisfaction with Six months undescribed views of communication communication, including Risk of Bias - 4 between GP and hospital doctors communication with primary measure, undescribed satisfaction care staff, communication with with communication between all the main outpatient doctor, hospice and home care staff views of communication involved in the patient’s care between hospital staff, views measure of communication between the GP and hospital doctors, or satisfaction with Caregiver Outcomes: communication between all None described hospice and home care staff involved in the patient’s care.

272

Evidence Table 1: Populations and Interventions

Dalton et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 43 2004 US Criteria: Cancer (mixed cancer) Pain interference (0-10 scale, Control Setting of Care - Not Usual care Outcomes Study Arms - 3 Patients were eligible for lower is better): control, mean Specified Comparing a profile-tailored inclusion if they were at least 5.0; standard CBT, mean 4.9; CBT intervention for cancer Level of Prognosis: Intervention Arm 1 - Outpatient Intervention Arm 1: Randomization - 18 years old; were able to profile-tailored CBT, mean clinics pain with standard CBT and Patients speak and read English; had Not described 5.2. Standard CBT usual care, the study an elevated score on at least Intervention Arm 2 - Outpatient demonstrated significant

Study Duration - one BPP scale; were clinics improvements in both profile- Six months experiencing cancer-related Patient Functional Status: Intervention Arm 2: tailored and standard CBT chronic pain (longer than six groups in interference of pain Risk of Bias - 4 SF-12 PCS (0-100 scale, Multidisciplinarity and Clinical Profile-tailored CBT weeks) associated with higher is better): control, mean with mood and sleep, Disciplines: disease progression, adjuvant 31.9; standard CBT, mean composite pain interference, treatment, or surgical 29.5; profile-tailored CBT, Control Arm - No multidisciplinary Intervention and physical functioning. The exploration not controlled with mean 28.7. team; clinical disciplines not Duration: profile-tailored CBT group current treatment; were specified. showed improvement Five weeks receiving treatment for cancer Intervention Arm 1 - No immediately and at one month pain; and were able to Caregiver Symptoms: multidisciplinary team; nurses. post-intervention, which participate in a pain evaluation None described diminished at six months; by and treatment program. They Intervention Arm 2 - No contrast, the standard CBT were excluded if they had multidisciplinary team; nurses. group showed little change chart-documented immediately and at one month psychopathology or confusion post-intervention, but then or had ever been enrolled in a showed improvement at six multi-week CBT program with months. The study did not a health professional. demonstrate significant changes in average, worst, or least pain; fatigue; anorexia; Number of Patient vigor; confusion; performance Participants and Decedents: status; or mental health. Screened - Not described Enrolled - 131 Completed - 28 Decedents - 26

273

Evidence Table 2: Outcomes

Dalton et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 43 2004 US Pain - BPI interference of pain with Pain: Outcomes Study Arms - 3 mood, interference of pain with Interference of pain with mood (0-10 Comparing a profile-tailored Level of sleep, and composite interference scale, lower is better): control, baseline, CBT intervention for cancer Randomization - subscales; BPI average pain, worst mean 4.8; immediate follow-up, mean 4.9; pain with standard CBT and Patients pain, and least pain scales one month, mean 4.0; six months, mean usual care, the study 4.2; standard CBT, baseline, mean 4.9; Other Symptoms - SDS fatigue and demonstrated significant Study Duration - immediate follow-up, mean 5.3; one improvements in both profile- Six months anorexia subscales, POMS month, mean 4.6; six months, mean 2.5; confusion subscale tailored and standard CBT Risk of Bias - 4 profile-tailored CBT, baseline, mean 5.0; groups in interference of pain Functional Status - SF-12 PCS, KPS immediate follow-up, mean 2.7; one with mood and sleep, month, mean 3.6; six months, mean 3.9; composite pain interference, Other QOL-Relevant Outcomes - group by time p<.05. SF-12 MCS, POMS vigor subscale and overall physical health. Interference of pain with sleep (0-10 The profile-tailored CBT group

scale, lower is better): control, baseline, showed improvement Caregiver Outcomes: mean 5.5; immediate follow-up, mean 4.4; immediately and at one month one month, mean 4.2; six months, mean None described post-intervention, which 4.1; standard CBT, baseline, mean 4.0; diminished at six months; by immediate follow-up, mean 4.3; one contrast, the standard CBT month, mean 4.4; six months, mean 2.9; profile-tailored CBT, baseline, mean 5.8; group showed little change immediate follow-up, mean 1.9; one immediately and at one month month, mean 3.0; six months, mean 3.0; post-intervention, but then group by time p<.05. showed improvement at six months. The study did not Composite pain interference (0-10 scale, demonstrate significant lower is better): control, baseline, mean 5.0; immediate follow-up, mean 4.9; one changes in average, worst, or month, mean 4.8; six months, mean 4.3; least pain; fatigue; anorexia; standard CBT, baseline, mean 4.9; vigor; confusion; performance immediate follow-up, mean 4.8; one status; or overall mental month, mean 5.7; six months, mean 3.3; health. profile-tailored CBT, baseline, mean 5.2; immediate follow-up, mean 2.7; one month, mean 3.7; six months, mean 3.8; group by time p<.05.

Functional Status: Overall physical health (0-100 scale, higher is better): control, baseline, mean 31.9; immediate follow-up, mean 29.6; one month, mean 28.2; six months, mean 34.1; standard CBT, baseline, mean 29.5; immediate follow-up, mean 27.9; one month, mean 28.9; six months, mean 31.4; profile-tailored CBT, baseline, mean 28.7; immediate follow-up, mean 31.0; one month, mean 31.1; six months, mean 30.0; group by time p<.05.

Caregiver Outcomes: N/A

274

Evidence Table 1: Populations and Interventions

Done et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 72 2001 EU Criteria: Dementia None described Control Arm - Home, residential Informational booklet Outcomes Study Arms - 2 Patients were eligible for care facilities about communication Comparing a communication Level of inclusion if they were Intervention Arm - Home, training workshop led by a diagnosed with progressive Prognosis: Patient Functional Status: speech therapist with an Randomization - residential care facilities Intervention Arm: Patients and dementia. Not described Not described informational booklet about communication for dementia families/caregivers Communication only Caregivers were eligible for Multidisciplinarity and Clinical training workshop led caregivers, the study inclusion if they were informal Caregiver Symptoms: Disciplines: by a speech therapist demonstrated a significant Study Duration - careers, had some form of improvement in the None described Control Arm - No multidisciplinary Six weeks verbal communication with the intervention group in the team; clinical disciplines not care recipient, and were Intervention number of communication Risk of Bias - 1 specified. willing to participate in the Duration: breakdowns accurately study. Intervention Arm - No One week identified in 10 scenes of a multidisciplinary team; speech video depicting an interaction therapist. between a spousal caregiver Number of Caregiver and a dementia patient. The Participants: study did not demonstrate Screened - 53 significant changes in communication problems or Enrolled - 45 caregiver stress. Completed - Not described

275

Evidence Table 2: Outcomes

Done et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 72 2001 EU None described Outcomes Study Arms - 2 Communication or Care Planning: Comparing a communication Level of training workshop led by a Caregiver Outcomes: Number of communication breakdowns speech therapist with an Randomization - accurately identified in 10 scenes of a Patients and Communication or Care Planning - informational booklet about video depicting an interaction between communication for dementia families/caregivers AACS, communication problems a spousal caregiver and a dementia caregivers, the study only (TACI) patient: baseline, control, mean 1.8 demonstrated a significant Study Duration - Other QOL-Relevant Outcomes - (95% CI 0.6-3.0); intervention, mean improvement in the Six weeks RSS 1.7 (95% CI 0.7-2.7); six weeks, intervention group in the Risk of Bias - 1 control, mean 3.4 (SD 1.8-5.0); number of communication intervention, mean 6.4 (SD 4.6-8.2); breakdowns accurately group by time p=.005. identified in 10 scenes of a video depicting an interaction between a spousal caregiver Caregiver Outcomes: and a dementia patient. The N/A study did not demonstrate significant changes in communication problems or caregiver stress.

276

Evidence Table 1: Populations and Interventions

Eisdorfer et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 70 al., 2003 US Criteria: Dementia None described Control Arm - Home Telephone-based Outcomes Study Arms - 3 Patients were eligible for Intervention Arm 1 - Home minimal support Comparing a family therapy Level of inclusion if they had a medical intervention with computer- diagnosis of probable Prognosis: Patient Functional Status: Intervention Arm 2 - Home telephone support for Randomization - Intervention Arm 1: Families or Alzheimer’s disease and Not described Katz ADL scale (0-6 scale, caregivers of dementia related disorders or had an patients with family therapy caregivers lower is better): all Multidisciplinarity and Clinical SET, a structural MMSE score of less than 24, alone and minimal support, participants, at least 1. Disciplines: family therapy Study Duration - 18 and had at least one ADL the study demonstrated a intervention focused months limitation or two IADL Control Arm - No multidisciplinary on restructuring significant improvement in Risk of Bias - 0 dependencies. Caregiver Symptoms: team; clinical disciplines not family interactions caregiver depression at six Depression (0-60 scale, lower specified. that may be linked to months but not 18 months in Caregivers were eligible for is better): minimal support, Intervention Arm 1 - No caregiver burden the family therapy with inclusion if they lived with the computer-telephone support Cuban-American participants, multidisciplinary team; therapist. patient and provided care for mean 17.70 (SD 8.9); minimal group versus family therapy at least four hours per day for Intervention Arm 2 - No Intervention Arm 2: alone and minimal support. support, white non-Hispanic multidisciplinary team; therapist. at least six months. They were participants, mean 18.57 (SD SET plus CTIS, an Family therapy with computer- excluded if their care 11.9); SET, Cuban-American information network telephone support was also recipients had a terminal or participants, mean 21.77 (SD that connects more effective than family severe illness or disability that 14.3); SET, white non- caregivers with their therapy alone and minimal would prohibit them from Hispanic participants, mean family and with support in improving participating in the 14.91 (SD 10.7); SET plus supportive resources depression for husband interventions, they were CTIS, Cuban-American outside of the home caregivers than for wife or involved in another caregiver daughter caregivers, and for participants, mean 19.31 (SD intervention study, had an 9.3); SET plus CTIS, white Cuban-American caregivers acute illness that would non-Hispanic participants, Intervention than for white non-Hispanic prevent them from mean 15.67 (SD 11.8); p=.25. Duration: caregivers. participating in the study, or 12 months were not planning to reside in the Miami area for at least six months.

Number of Patient Participants and Decedents: Screened - Not described Enrolled - 225 Completed - 154 Decedents - Not described

Number of Caregiver Participants: Screened - Not described Enrolled - 225 Completed - 154

277

Evidence Table 2: Outcomes

Eisdorfer et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 70 al., 2003 US None described N/A Outcomes Study Arms - 3 Comparing a family therapy intervention with computer- Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - telephone support for Families or Depression - CES-D caregivers of dementia patients with family therapy caregivers Depression: alone and minimal support, Study Duration - 18 Depression (0-60 scale, lower is the study demonstrated a months better): six months, SET plus CTIS vs. significant improvement in Risk of Bias - 0 SET and minimal support, group by caregiver depression at six time p=.036, group by ethnicity p=.035, months but not 18 months in group by relationship p=.003; 18 the family therapy with months, SET plus CTIS vs. SET and computer-telephone support minimal support, group by time p=.15, group versus family therapy group by time by ethnicity p=.001, alone and minimal support. group by relationship p=.209. Family therapy with computer- telephone support was also more effective than family therapy alone and minimal support in improving depression for husband caregivers than for wife or daughter caregivers, and for Cuban-American caregivers than for white non-Hispanic caregivers.

278

Evidence Table 1: Populations and Interventions

Hepburn et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 137 al., 2005 US Criteria: Dementia None described Control Arm - Not specified Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Easily accessible Comparing a Level of inclusion if they were community settings psychoeducational diagnosed with a dementing Prognosis: Patient Functional Status: Intervention Arm: intervention designed to Randomization - Patients and illness and were community Not described Not described Partners in strengthen the skills, dwelling. They were excluded knowledge, and attitude of families/caregivers Multidisciplinarity and Clinical Caregiving Program, only if they had serious psychiatric Disciplines: a psychoeducational dementia family caregivers, comorbidities. Caregiver Symptoms: with usual care, the study Control Arm - No multidisciplinary intervention designed Study Duration - 12 demonstrated significant None described team; clinical disciplines not to strengthen the months Caregivers were eligible for improvements in the specified. skills, knowledge, and Risk of Bias - 1 inclusion if they were giving attitude of dementia intervention group in caregiver care to a community-dwelling Intervention Arm - No family caregivers distress and beliefs and relative diagnosed with a multidisciplinary team; clinical attitudes about caregiving at

dementing illness and disciplines not specified. six months but not 12 months, expected that the care Intervention and a significant improvement recipient would remain in the Duration: in caregiver competence at 12 community and not be placed Six weeks months but not six months. in a nursing home for the one The study did not demonstrate year study duration. They significant changes in were excluded if they had caregiver burden, role serious medical or mental captivity, or relational health problems. deprivation.

Number of Patient Participants and Decedents: Screened - 233 Enrolled - 215 Completed - 131 Decedents - 7

Number of Caregiver Participants: Screened - 233 Enrolled - 215 Completed - 131

279

Evidence Table 2: Outcomes

Hepburn et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 137 al., 2005 US None described N/A Outcomes Study Arms - 2 Comparing a psychoeducational Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - intervention designed to Patients and Other QOL-Relevant Outcomes - strengthen the skills, knowledge, and attitude of families/caregivers Undescribed caregiver competence Other QOL-Relevant Outcomes: only scale, caregiving beliefs and dementia family caregivers, attitudes (BACS), study-specific Caregiver competence (scale not with usual care, the study Study Duration - 12 caregiver distress scale, ZBI, described, higher is better): 12 months demonstrated significant months undescribed caregiver role captivity vs. baseline, control, mean +0.144 (SD improvements in the Risk of Bias - 1 scale, undescribed caregiver 0.44); intervention, mean +0.290 (SD intervention group in caregiver relational deprivation scale 0.52); p=.038. distress and beliefs and Caregiving beliefs and attitudes (scale attitudes about caregiving at not described, lower is better): six six months but not 12 months, months vs. baseline, control, mean and a significant improvement +2.03 (SD 3.73); intervention, mean - in caregiver competence at 12 2.87 (SD 4.42); p=.016. months but not six months. The study did not demonstrate Caregiver distress (scale not significant changes in described, lower is better): six months caregiver burden, role vs. baseline, control, mean +3.54 (SD captivity, or relational 7.59); intervention, mean +0.525 (SD deprivation. 7.46); p=.035.

280

Evidence Table 1: Populations and Interventions

Magai et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 138 2002 US Criteria: Dementia Symptomatology (combined Control Arm - Nursing home Usual care Outcomes Study Arms - 3 Patients were eligible for BEHAVE-AD, CMAI, and CDS Intervention Arm 1- Nursing home Comparing a training inclusion if they were nursing scores) (scale not described, intervention for nursing home Level of Prognosis: Intervention Arm 2 - Nursing home Intervention Arm 1: Randomization - home residents with dementia lower is better): control, mean staff caregivers of dementia Sites identified via Minimum Data Not described 40.6 (SD 7.8); training Training sessions for patients that focused on Set charts and the MMSE. placebo, mean 25.2 (SD 5.2); patient affect and nonverbal Multidisciplinarity and Clinical nursing home staff Study Duration - 12 training intervention, mean sensitivity with a (1) training Disciplines: caregivers that weeks 83.7 (SD 51.2). addressed the control that did not have that Number of Patient Control Arm - No multidisciplinary Risk of Bias - 6 Psychological symptoms cognitive and focus and (2) usual care, the Participants and Decedents: team; nursing home staff (scale not described, lower is behavioral aspects of study did not demonstrate any caregivers. Patients Screened - 102 better): control, mean 69.5 dementia significant changes in patient overall symptomatology or Enrolled - 99 (SD 11.4); training placebo, Intervention Arm 1 - No mean 59.3 (SD 7.1); training multidisciplinary team; nursing psychological symptoms. Completed - 91 Intervention Arm 2: intervention, mean 65.1 (SD home staff caregivers. Decedents - 6 Training sessions for 4.0). Intervention Arm 2 - No Caregiver QOL-Relevant nursing home staff multidisciplinary team; nursing Outcomes home staff caregivers. caregivers that Patient Functional Status: addressed the The study did not demonstrate Not described cognitive and a significant change in behavioral aspects of caregiver psychological dementia with symptoms. Caregiver Symptoms: specific attention to None described patient affect and nonverbal sensitivity

Intervention Duration: Two weeks

281

Evidence Table 2: Outcomes

Magai et al., Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 138 2002 US Other Symptoms - Symptomatology Outcomes Study Arms - 3 (combined BEHAVE-AD, CMAI, and Comparing a training Level of CDS scores), psychological intervention for nursing home Randomization - symptoms (BSI) staff caregivers of dementia Sites patients that focused on patient affect and nonverbal Study Duration - 12 Caregiver Outcomes: sensitivity with a (1) training weeks Other Symptoms - Psychological control that did not have that Risk of Bias - 6 symptoms (BSI) focus and (2) usual care, the study did not demonstrate any significant changes in patient overall symptomatology or psychological symptoms.

Caregiver QOL-Relevant Outcomes The study did not demonstrate a significant change in caregiver psychological symptoms.

282

Evidence Table 1: Populations and Interventions

Pillemer et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 139 al., 2002 US Criteria: Dementia None described Control Arm - Not described Usual care Outcomes Study Arms - 2 Caregivers were eligible for Intervention Arm - Not described Comparing a peer support inclusion if they were the intervention for caregivers of Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - primary caregiver to a dementia patients with usual Families or diagnosed Alzheimer’s Not described Need assistance with bathing: Multidisciplinarity and Clinical Peer support care, the study did not caregivers disease patient who did not control, 57.4%; intervention, Disciplines: intervention for demonstrate any significant reside in a nursing home. changes in caregiver 56.6%; p=0.78. Control Arm - No multidisciplinary dementia caregivers Study Duration - depression or self-esteem. Six months Need assistance with team; clinical disciplines not dressing: control, 47.5%; specified. Risk of Bias - 0 Number of Caregiver Intervention Participants: intervention, 50.0%; p=0.54. Intervention Arm - No Duration: Screened - 294 Need assistance with eating: multidisciplinary team; clinical Eight weeks control, 36.0; intervention, disciplines not specified. Enrolled - 147 29.6%; p=0.20. Completed - 115 Caregiver Symptoms: Depression (0-60 scale, lower is better): control, mean 18.8; treatment, mean 18.8; p=0.98.

283

Evidence Table 2: Outcomes

Pillemer et Study Location - Patient Outcomes: None None N/A Caregiver QOL-Relevant 139 al., 2002 US None described Outcomes Study Arms - 2 Comparing a peer support intervention for caregivers of Level of Caregiver Outcomes: Randomization - dementia patients with usual Families or Depression - CES-D care, the study did not demonstrate any significant caregivers Other QOL-Relevant Outcomes - changes in caregiver Rosenberg Self-Esteem Scale Study Duration - depression or self-esteem. Six months Risk of Bias - 0

284

Evidence Table 1: Populations and Interventions

Ream et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 36 2006 EU Criteria: Cancer (mixed cancer) Fatigue (0-100 scale, lower is Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for better): control, mean 42.6 Intervention Arm - Home Comparing a fatigue education inclusion if they were between (SD 28.8); intervention, mean and monitoring and Level of Prognosis: Intervention Arm: Randomization - 18 and 70 years old; had been 38.8 (SD 28.9); p=.51. psychological support Patients diagnosed with non-Hodgkin’s Not described Multidisciplinarity and Clinical Intervention that intervention with usual care for lymphoma or gastrointestinal, Disciplines: provided assessment cancer patients undergoing Study Duration - nonsmall cell lung, colorectal, Patient Functional Status: chemotherapy, the study Control Arm - No multidisciplinary and monitoring of Four months breast, or unknown primary demonstrated significant Not described team; clinical disciplines not fatigue, fatigue cancer; were chemotherapy- improvements in the Risk of Bias - 3 specified. education, coaching naive; were due to commence in self-care, and intervention group in fatigue, their first cycle of cytotoxic Caregiver Symptoms: Intervention Arm 1 - No emotional support distress caused by fatigue, chemotherapy; and could multidisciplinary team; nurses. depression, anxiety, vitality, None described understand, speak, read, and mental health, coping, and write English. They were Intervention adaptive coping (more humor excluded if they were being Duration: and less behavioral treated for psychiatric illness. Three months disengagement). The study did not demonstrate a significant change in extent of Number of Patient fatigue. Participants and Decedents: Screened - 145 Enrolled - 103 Completed - 86 Decedents - Not described

285

Evidence Table 2: Outcomes

Ream et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 36 2006 EU Depression - HADS depression Outcomes Study Arms - 2 subscale Depression: Comparing a fatigue education Level of Anxiety - HADS anxiety subscale Depression (scale not described, lower is better): and monitoring and baseline, control, mean rank 48.6; intervention, Randomization - psychological support Other Symptoms - Study-specific mean rank 40.2; four months, control, mean rank intervention with usual care for Patients 48.3; intervention, mean rank 37.6; p=.02. fatigue, distress caused by fatigue, cancer patients undergoing Study Duration - and extent of fatigue VASs chemotherapy, the study Four months Other QOL-Relevant Outcomes - Anxiety: demonstrated significant Risk of Bias - 3 SF-36 vitality and mental health Anxiety (scale not described, lower is better): improvements in the subscales, COPE behavioral baseline, control, mean rank 44.5; intervention, intervention group in fatigue, mean rank 44.4; four months, control, mean rank disengagement and humor 47.9; intervention, mean rank 38.0; p=.03. distress caused by fatigue, subscales, study-specific coping depression, anxiety, vitality,

VAS mental health, coping, and Other Symptoms: adaptive coping (more humor

Fatigue (0-100 scale, lower is better): baseline, and less behavioral Caregiver Outcomes: control, mean 42.6 (SD 28.8); intervention, mean disengagement). The study 38.8 (SD 28.9); four months, control, mean 41.6 None described (SD 29.4); intervention, mean 30.6 (SD 27.7); did not demonstrate a p=.04. significant change in extent of Distress caused by fatigue (0-100 scale, lower is fatigue. better): baseline, control, mean 33.3 (SD 33.3); intervention, mean 28.4 (SD 30.1); four months, control, mean 38.4 (SD 31.0); intervention, mean 24.0 (SD 27.1); p=.01.

Other QOL-Relevant Outcomes: Vitality (0-100 scale, higher is better): baseline, control, mean rank 42.8; intervention, mean rank 46.3; four months, control, mean rank 38.2; intervention, mean rank 47.1; p=.05. Mental health (0-100 scale, higher is better): baseline, control, mean rank 43.8; intervention, mean rank 45.3; four months, control, mean rank 36.9; intervention, mean rank 48.4; p=.02. Behavioral disengagement coping strategy (scale not described, lower is better): baseline, control, mean rank 41.4; intervention, mean rank 44.7; four months, control, mean rank 46.3; intervention, mean rank 39.7; p=.02. Humor coping strategy (scale not described, higher is better): baseline, control, mean rank 40.5; intervention, mean rank 48.7; four months, control, mean rank 35.2; intervention, mean rank 51.0; p<.01. Coping (scale not described, lower is better): baseline, control, mean rank 24.0; intervention, mean rank 18.2; four months, control, mean rank 18.3; intervention, mean rank 11.9; p=.05.

Caregiver Outcomes: N/A

286

Evidence Table 1: Populations and Interventions

Teri et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 64 2005 US Criteria: Dementia None described Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a training inclusion if they had a intervention for dementia Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of probable or caregivers that focused on Patients and possible Alzheimer’s disease Not described Not described Multidisciplinarity and Clinical Training for dementia reducing mood and behavioral from their primary physician problems in patients with families/caregivers Disciplines: caregivers that and had three or more Alzheimer’s disease, with only Control Arm - No multidisciplinary focused on reducing agitated or depressed Caregiver Symptoms: mood and behavior usual care, the study Study Duration - team; clinical disciplines not behavior problems reported by Depression (0-60 scale, lower problems in patients demonstrated significant Six months specified. their caregivers that occurred is better): control, mean 13.2 with Alzheimer’s improvements in the Risk of Bias - 2 three or more times during the (SD 8.5); intervention, mean Intervention Arm - No disease intervention group in past week at study entry. multidisciplinary team; social caregiver-reported patient 14.8 (SD 9.1). workers, consultants with masters HRQOL and patient memory- Caregivers were eligible for Depression (0-50 scale, lower degrees or equivalents in Intervention related problem behaviors. inclusion if they were a is better): control, mean 7.6 counseling or psychology. Duration: (SD 5.0); intervention, mean spouse or adult relative caring Six months for a dementia patient in the 6.9 (SD 4.1). Caregiver QOL-Relevant home. Outcomes The study demonstrated significant improvements in Number of Patient the intervention group in Participants and Decedents: caregiver depression, burden, Screened - Not described reactivity to patient behavior Enrolled - 95 problems, and sleep quality, quantity, and frequency of Completed - 66 sleep problems. Decedents - Not described

Number of Caregiver Participants: Screened - Not described Enrolled - 95 Completed - 66

287

Evidence Table 2: Outcomes

Teri et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 64 2005 US HRQOL - QOL-AD Outcomes HRQOL: Study Arms - 2 Other QOL-Relevant Outcomes - Comparing a training Caregiver-reported patient QOL (13-52 scale, higher intervention for dementia Level of RMBPC memory subscale is better): baseline, control, mean 28.3 (SD 4.9); intervention, mean 27.8 (SD 5.5); two months, caregivers that focused on Randomization - Patients and control, mean 28.4 (SD 4.9); intervention, mean 29.4 reducing mood and behavioral (SD 5.2); six months, control, mean 28.2 (SD 4.6); problems in patients with families/caregivers Caregiver Outcomes: intervention, mean 28.4 (SD 5.4); p=.031. Alzheimer’s disease, with only Depression - CES-D, HAM-D usual care, the study Study Duration - Other QOL-Relevant Outcomes - Other QOL-Relevant Outcomes: demonstrated significant Six months SCB subjective burden subscale, Memory-related problem behaviors (0-28 scale, lower improvements in the is better): baseline, control, mean 3.0 (SD 0.8); Risk of Bias - 2 RMBPC caregiver reaction subscale, intervention, mean 3.0 (SD 0.7); two months, control, intervention group in Caregiver Sleep Questionnaire mean 3.0 (SD 0.9); intervention, mean 2.8 (SD 0.8); caregiver-reported patient six months, control, mean 3.1 (SD 1.0); intervention, HRQOL and patient memory- mean 2.8 (SD 0.8); p=.031. related problem behaviors.

Caregiver Outcomes: Caregiver QOL-Relevant Depression: Outcomes Depression (0-60 scale, lower is better): baseline, The study demonstrated control, mean 13.2 (SD 8.5); intervention, mean 14.8 significant improvements in (SD 9.1); two months, control, mean 13.6 (SD 9.3); intervention, mean 12.4 (SD 7.6); six months, control, the intervention group in mean 15.8 (SD 10.5); intervention, mean 12.5 (SD caregiver depression, burden, 7.7); p=.023. reactivity to patient behavior Depression (0-50 scale, lower is better): baseline, problems, and sleep quality, control, mean 7.6 (SD 5.0); intervention, mean 6.9 (SD 4.1); two months, control, mean 7.8 (SD 5.3); quantity, and frequency of intervention, mean 6.3 (SD 4.5); six months, control, sleep problems. mean 8.5 (SD 5.7); intervention, mean 6.7 (SD 3.9); p=.041.

Other QOL-Relevant Outcomes: Caregiver burden (scale not described, lower is better): baseline, control, mean 23.4 (SD 12.2); intervention, mean 24.7 (SD 12.4); two months, control, mean 23.1 (SD 14.1); intervention, mean 20.1 (SD 10.6); six months, control, mean 25.8 (SD 13.7); intervention, mean 21.4 (SD 12.5); p=.029. Caregiver reactivity to patient behavior problems (scale not described, lower is better): baseline, control, mean 25.0 (SD 14.7); intervention, mean 28.1 (SD 15.9); two months, control, mean 23.3 (SD 16.0); intervention, mean 22.3 (SD 15.8); six months, control, mean 23.4 (SD 14.5); intervention, mean 21.9 (SD 15.6); p=.037. Sleep quality, quantity, and frequency of sleep problems (scale not described, lower is better): baseline, control, mean 10.5 (SD 3.9); intervention, mean 11.3 (SD 3.9); two months, control, mean 10.1 (SD 4.2); intervention, mean 9.9 (SD 3.6); six months, control, mean 9.1 (SD 5.2); intervention, mean 9.2 (SD 4.6); p=.033.

288

Evidence Table 1: Populations and Interventions

Wilkie et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 140 2010 US Criteria: Cancer (lung) Months since noticed pain: Control Arm - Not described Usual care Outcomes Study Arms - 2 Patients were eligible for control, mean 8.0 (SD 17.5); Intervention Arm - Not described Comparing an intervention inclusion if they had a intervention, mean 6.2 (SD that coached lung cancer Level of Prognosis: Intervention Arm: Randomization - diagnosis of small cell non- 10.8); p=.46. patients on sensory self- Patients small cell lung cancer, had Not described Multidisciplinarity and Clinical Decision support, monitoring and reporting of pain related to lung cancer or Disciplines: coaching intervention pain with usual care, the study Study Duration - to anticancer therapies during Patient Functional Status: did not demonstrate any Control Arm - No multidisciplinary involving sensory Four weeks the week prior to enrolling, significant changes in pain Not described team; clinical disciplines not self-monitoring and and spoke and read English. intensity, pain quality, pain Risk of Bias - 5 specified. reporting of pain They were excluded if they catastrophizing coping,

had pain for less than two Caregiver Symptoms: Intervention Arm - No anxiety, or depression. multidisciplinary team; clinical Decision support months after thoracotomy, had None described a score of 20 or less on the disciplines not specified. elements: Drugs MMSE, or were physically Video-based and unable to participate. printed decision aids The study did not demonstrate a significant change in the

adequacy of analgesics Number of Patient Intervention prescriptions. Participants and Decedents: Duration: Screened - 296 Three weeks Enrolled - 215 Completed - 151 Decedents - 22

289

Evidence Table 2: Outcomes

Wilkie et al., Study Location - Patient Outcomes: None Drugs - Adequacy of analgesics None Patient QOL-Relevant 140 2010 US Pain - Pain intensity VAS, pain prescriptions Outcomes Study Arms - 2 quality (MPQ, PRI), pain Comparing an intervention Level of catastrophizing coping (CSQ) that coached lung cancer Randomization - Depression - CES-D patients on sensory self- Patients monitoring and reporting of Anxiety - STAI pain with usual care, the study Study Duration - did not demonstrate any Four weeks significant changes in pain Risk of Bias - 5 intensity, pain quality, pain catastrophizing coping, anxiety, or depression.

Drugs The study did not demonstrate a significant change in the adequacy of analgesics prescriptions.

290

Evidence Table 1: Populations and Interventions

Williams et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Settings of Care: Control Arm: Patient QOL-Relevant 141 al., 2001 EU Criteria: Cancer (mixed cancer) Fatigue (0-100 scale, lower is Control Arm - Not described Usual care Outcomes better): control, mean 48.67 (SD Study Arms - 2 Patients were eligible for Intervention Arm - Not described Comparing a patient-held inclusion if they had a 28.54); intervention, mean 46.04 record intervention with usual Level of Prognosis: (SD 28.46); p=.30. Intervention Arm: Randomization - diagnosis of cancer and were care for cancer patients, the Patients at least 18 years old. They Not described Nausea and vomiting (0-100 Multidisciplinarity and Clinical Patient-held record study did not demonstrate any were excluded if they had scale, lower is better): control, significant changes in global Disciplines: Study Duration - basal cell carcinoma of the mean 14.20 (SD 23.75); HRQOL, physical functioning, Control Arm - No multidisciplinary Six months skin, were not able to be intervention, mean 10.09 (SD Intervention pain, dyspnea, fatigue, team; clinical disciplines not responsible for a patient-held 19.49); p=.03. Duration: anorexia, nausea and Risk of Bias - 1 specified. record and did not have Pain (0-100 scale, lower is Six months vomiting, constipation, cancer who was able to take better): control, mean 40.07 (SD Intervention Arm - No diarrhea, or sleep this responsibility, or were 33.50); intervention, mean 37.52 multidisciplinary team; clinical disturbances. (SD 30.90); p=.38. disciplines not specified. registered at one of the six general practices involved in a Dyspnea (0-100 scale, lower is pilot project. better): control, mean 28.93 (SD Health Care Use 33.31); intervention, mean 28.02 The study did not demonstrate (SD 33.17); p=.76. any significant changes in Number of Patient Sleep disturbance (0-100 scale, outpatient visits, GP home Participants and Decedents: lower is better): control, mean visits, district nurse visits, Screened -1148 37.20 (SD 34.81); intervention, specialist nurse visits, hospital mean 36.39 (SD 34.70); p=.79. Enrolled - 501 days, or day care attendance. Anorexia (0-100 scale, lower is Completed - 344 better): control, mean 28.00 (SD Decedents - 104 35.19); intervention, mean 24.30 Procedures (SD 32.63); p=.22. The study did not demonstrate Constipation (0-100 scale, lower a significant change in surgery is better): control, mean 25.87 visits. (SD 33.23); intervention, mean 24.83 (SD 30.53); p=.72. Diarrhea (0-100 scale, lower is better): control, mean 6.93 (SD 18.56); intervention, mean 7.70 (SD 19.39); p=.65.

Patient Functional Status: EORTC QLQ-C30 physical functioning scale (0-100 scale, higher is better): control, mean 61.20 (SD 29.62); intervention, mean 62.55 (SD 28.84); p=.61.

Caregiver Symptoms: None described

291

Evidence Table 2: Outcomes

Williams et Study Location - Patient Outcomes: None Health Care Use - Inpatient None Patient QOL-Relevant 141 al., 2001 EU Pain - EORTC QLQ-C30 pain scale hospital, outpatient visits, home Outcomes Study Arms - 2 health care, nurse visits, day Comparing a patient-held Dyspnea - EORTC QLQ-C30 care attendance Level of dyspnea single item record intervention with usual Procedures - Surgery visits care for cancer patients, the Randomization - Other Symptoms - EORTC QLQ-C30 Patients study did not demonstrate any fatigue and nausea and vomiting significant changes in global Study Duration - scales and sleep disturbance, HRQOL, physical functioning, Six months constipation, diarrhea, and anorexia pain, dyspnea, fatigue, Risk of Bias - 1 single items anorexia, nausea and Functional Status - EORTC QLQ- vomiting, constipation, C30 physical functioning scale diarrhea, or sleep disturbances. HRQOL - EORTC QLQ-C30 global QOL scale Health Care Use Caregiver Outcomes: The study did not demonstrate any significant changes in None described outpatient visits, GP home visits, district nurse visits, specialist nurse visits, hospital days, or day care attendance.

Procedures The study did not demonstrate a significant change in surgery visits.

292

Evidence Table 1: Populations and Interventions

Wright et al., Study Location - Inclusion Exclusion Criteria: Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 28 2001 US Patients were eligible for Dementia Agitation (scale not described, Control Arm - Home Usual care Outcomes Study Arms - 2 inclusion if they had lower is better): control, mean Intervention Arm - Home Comparing a nurse education Alzheimer’s disease. 74.09 (SD 21.42); and counseling program for Level of Prognosis: Intervention Arm: Randomization - intervention, mean 65.90 (SD caregivers of patients with Families or Caregivers were eligible for Not described 21.92); p=.13. Multidisciplinarity and Clinical Nurse education and Alzheimer’s disease, with caregivers inclusion if they identified as Disciplines: counseling program usual care, the study did not the primary caregiver for an demonstrate a significant Control Arm - No multidisciplinary Study Duration - 12 elder with Alzheimer’s Patient Functional Status: change in patient agitation. team; clinical disciplines not months disease, lived in the same Intervention None described specified. Risk of Bias - 2 household as the Alzheimer’s Duration: Caregiver QOL-Relevant disease patient, and lived Intervention Arm - No 12 months within 80 miles of the hospital. Caregiver Symptoms: multidisciplinary team; nurses. Outcomes Depression (0-60 scale, lower The study did not demonstrate is better): control, mean 12.04 any significant changes in Number of Caregiver (SD 8.70); intervention, mean caregiver depression, stress, Participants: 13.30 (SD 11.95); p=.59. or physical health. Screened - 102 Enrolled - 93 Health Care Use Completed - Not described The study did not demonstrate a significant change in rates of institutionalization.

293

Evidence Table 2: Outcomes

Wright et al., Study Location - Patient Outcomes: None Health Care Use - None Patient QOL-Relevant 28 2001 US Other Symptoms - CMAI Institutionalization Outcomes Study Arms - 2 Comparing a nurse education and counseling program for Level of Caregiver Outcomes: Randomization - caregivers of patients with Families or Depression - CES-D Alzheimer’s disease, with usual care, the study did not caregivers Other QOL-Relevant Outcomes - demonstrate a significant Stress (Caregiving Hassle Scale), Study Duration - 12 change in patient agitation. months physical health (MAI)

Risk of Bias - 2 Caregiver QOL-Relevant Outcomes The study did not demonstrate any significant changes in caregiver depression, stress, or physical health.

Health Care Use The study did not demonstrate a significant change in rates of institutionalization.

294

Evidence Table 1: Populations and Interventions

Higginson et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 60 al., 2009 EU Criteria: MS Pain (0-4 scale, lower is Control Arm - Outpatient clinics, Usual care Outcomes Study Arms - 2 Patients were eligible for better): control, mean 1.28 home, hospital Comparing a palliative care inclusion if they were in South (SD 1.24); intervention, mean team intervention with usual Level of Prognosis: Intervention Arm - Outpatient Intervention Arm: Randomization - East London, were living with 1.69 (SD 1.35). Clinics, home, hospital, nursing care for patients with MS, the Patients MS, and were deemed by Identified by clinicians as homes Palliative care team study demonstrated a clinicians to have one or more significant improvement in the poor prognosis Study Duration - 24 of unresolved symptoms, Patient Functional Status: intervention group in pain. The weeks psychosocial concerns, end- UNDS score (0-60 scale, Multidisciplinary Team and Palliative care team study did not demonstrate a Risk of Bias - 4 of-life issues, progressive lower is better): control, mean Clinical Disciplines: elements: significant change in a illness, or complex needs (i.e., measure of anxiety, patient 30 (SD 9.2); intervention, Control Arm - No multidisciplinary Inpatient consultation, palliative care needs). They and carer concerns, and mean 28 (SD 8.9); p>.05. team; physicians, nurses; outpatient were excluded if they had very practical needs. EDSS score (0-10 scale, rehabilitation, neurological, and consultation urgent needs or were lower is better): control, mean social services. deteriorating rapidly when 7.9 (SD 1); intervention, mean immediate referral to the Intervention Arm - Palliative care Intervention Caregiver QOL-Relevant 7.7 (SD 1); p>.05. service was offered. team; physicians, nurses, Duration: Outcomes psychosocial worker, administrator; 12 weeks The study demonstrated a Caregiver Symptoms: rehabilitation, neurological, and significant improvement in the Number of Patient social services. intervention group in caregiver None described Participants and Decedents: burden. Screened - 69 Enrolled - 52 Health Care Use Completed - 46 The study demonstrated Decedents - 3 unclear changes in the use of nurses; GPs; specialists at home or in the hospital; occupational therapists or physiotherapists; dieticians, chiropodists, or dentists; speech therapists; social services; day centers; inpatient care; and respite care.

295

Evidence Table 2: Outcomes

Higginson et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient None Patient QOL-Relevant 60 al., 2009 EU Pain - POS-8 pain item hospital, outpatient visits, home Outcomes Study Arms - 2 health care; use of nurses, Comparing a palliative care Other QOL-Relevant Outcomes - Pain: occupational therapists or Level of Anxiety, patient and carer concerns, team intervention with usual Pain: (0-4 scale, lower is better): physiotherapists, dieticians, care for patients with MS, the Randomization - and practical needs (POS-8) baseline, control, mean 1.28 (SD 1.24); chiropodists, or dentists, speech Patients study demonstrated a intervention, mean 1.69 (SD 1.35); six therapists, social services, day significant improvement in the centers, and respite care Study Duration - 24 Caregiver Outcomes: weeks vs. baseline, control, mean intervention group in pain. The weeks +0.09 (95% CI -0.36 to 0.54); Health Care Costs - Total costs study did not demonstrate a Other QOL-Relevant Outcomes - ZBI Risk of Bias - 4 intervention, mean -0.23 (95% CI -0.66 significant change in a to 0.20); 12 weeks vs. baseline, measure of anxiety, patient control, mean +0.30 (95% CI -0.16 to and carer concerns, and 0.76); intervention, mean -0.46 (95% practical needs. CI -0.98 to 0.05); p=.028.

Caregiver QOL-Relevant Caregiver Outcomes: Outcomes The study demonstrated a Other QOL-Relevant Outcomes: significant improvement in the intervention group in caregiver Caregiver burden (0-48 scale, lower is burden. better): baseline, control, mean 22.6 (SD 6.2); intervention, mean 20.5 (SD 10.7); six weeks vs. baseline, control, Health Care Use mean -1.13 (95% CI -3.41 to 1.14); intervention, mean +1.10 (95% CI - The study demonstrated 3.43 to 5.63); 12 weeks vs. baseline, unclear changes in the use of control, mean +1.58 (95% CI -0.51 to nurses; GPs; specialists at 3.67); intervention, mean -2.88 (95% home or in the hospital; CI -5.99 to 0.24); p=.011. occupational therapists or physiotherapists; dieticians, chiropodists, or dentists; speech therapists; social services; day centers; inpatient care; and respite care.

296

Evidence Table 1: Populations and Interventions

Mills et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 95 2009 EU Criteria: Cancer (lung) None described Control Arm - Not specified Usual care Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Home Comparing a QOL diary with inclusion if they had a clinical usual care for patients with Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - diagnosis of inoperable lung lung cancer, the study Patients cancer, had an ECOG Not described FACT-L physical well-being Multidisciplinary Team and Weekly completion of demonstrated mixed effects performance status of 0 to 2, subscale (scale not described, Clinical Disciplines: a QOL questionnaire on HRQOL (either worse in Study Duration - 16 and were able to give written the intervention group or no higher is better): control, mean Control Arm - No multidisciplinary in a diary format weeks informed consent and significant difference between 19.3 (SD 6.8); intervention, team; clinical disciplines not complete questionnaires. groups). The study did not Risk of Bias - 2 mean 19.4 (SD 6.4); p>.05. specified. Intervention demonstrate significant FACT-L functional well-being Intervention Arm - No Duration: changes in lung cancer subscale (scale not described, Number of Patient multidisciplinary team; clinical symptoms, perceptions that higher is better): control, mean 16 weeks Participants and Decedents: disciplines not specified. care met needs and 14.4 (SD 6.6); intervention, expectations, or satisfaction Screened - 498 mean 14.6 (SD 7.2); p>.05. with involvement in decision- Enrolled - 115 ECOG performance status (0- making, information given, Completed - 57 5 scale, lower is better): emotional support, or overall Decedents -27 control, 0, 30%; 1, 53%; 2, quality of care. 17%; intervention, 0, 23%; 1,

56%; 2, 21%; p>.05.

Caregiver Symptoms: None described

297

Evidence Table 2: Outcomes

Mills et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 95 2009 EU Other Symptoms - FACT-L LCS Outcomes Study Arms - 2 Experience or Satisfaction - HRQOL: Comparing a QOL diary with usual care for patients with Level of Unnamed perception that care met QOL (0-108 scale, higher is better): Randomization - needs and expectations, and lung cancer, the study baseline, control, mean 90.5 (SD 20.9); demonstrated mixed effects Patients satisfaction with involvement in intervention, mean 93.8 (SD 21.8); 16 decision-making, information given, on HRQOL (either worse in Study Duration - 16 weeks vs. baseline, control, mean +3.5 the intervention group or no weeks emotional support, and overall (SD 18.4); intervention, mean -6.3 (SD quality of care items significant difference between Risk of Bias - 2 14.9); p=.04. groups). The study did not HRQOL - FACT-L, FACT-G, TOI QOL (scale not described, higher is demonstrate significant (sum of FACT-L, LCS, and physical better): baseline, control, mean 73.3 changes in lung cancer and functional well-being subscales), (SD 17.4); intervention, mean 75.2 (SD symptoms, perceptions that PQLI 17.5); 16 weeks vs. baseline, control, care met needs and mean +0.2 (SD 15.7); intervention, expectations, or satisfaction with involvement in decision- Caregiver Outcomes: mean -6.6 (SD 12.5); p=.04. making, information given,

None described emotional support, or overall Caregiver Outcomes: quality of care. N/A

298

Evidence Table 1: Populations and Interventions

Pantilat et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 118 al., 2010 US Criteria: Cancer (mixed cancer), Depression: (0-15 scale, lower Control Arm - Hospital Usual care, packet of Outcomes Study Arms - 2 Patients were eligible for CHF, COPD, cirrhosis is better): all participants, 21% Intervention Arm - Hospital information on diet Comparing a physician- inclusion if they were at least had a score of six or higher. and exercise focused palliative medicine Level of Randomization - 65 years old; had heart failure, Pain (0-10 scale, lower is consultation, with usual care cancer, COPD, or cirrhosis; Prognosis: Multidisciplinary Team and and a packet of information on Patients better): control, mean 3.5 Intervention Arm: were able to give informed Not described (range 2.4-4.8); intervention, Clinical Disciplines: diet and exercise, for Study Duration - consent; and spoke English. hospitalized patients, the mean 4.9 (range 3.8-6.0). Control Arm - No multidisciplinary Palliative care team Two weeks after study did not demonstrate team; physicians. hospital discharge Dyspnea (0-10 scale, lower is significant changes in pain, Risk of Bias - 2 Number of Patient better): control, mean 3.0 Intervention Arm - No Palliative care team dyspnea, anxiety, Participants and Decedents: (range 1.8-4.2); intervention, multidisciplinary team; physicians, elements: communication about religion, mean 4.4 (range 3.3-5.5). Screened - Not provided chaplains, pharmacists. Inpatient consultation or communication with the Anxiety (0-10 scale, lower is health care team. Enrolled - 107 better): control, mean 3.8 Completed - 81 (range 2.7-5.0); intervention, Intervention Decedents - Not provided mean 5.5 (range 4.2-5.5). Duration: Length of hospital stay (study mean, 5.5 Patient Functional Status: days) Katz ADL scale (0-6 scale, lower is better): all participants, 71% had one or more dependencies.

Caregiver Symptoms: None described

299

Evidence Table 2: Outcomes

Pantilat et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 118 al., 2010 US Pain - Worst pain in past 24 hours Outcomes Study Arms - 2 (10-point NRS) Comparing a physician- Level of Dyspnea - Worst dyspnea in past 24 focused palliative medicine Randomization - hours (10-point NRS) consultation, with usual care Patients and a packet of information on Anxiety - Worst anxiety in past 24 diet and exercise, for Study Duration - hours (10-point NRS) hospitalized patients, the Two weeks after Existential or Spiritual Well-Being - study did not demonstrate hospital discharge Study-specific communication about significant changes in pain, Risk of Bias - 2 religion items dyspnea, anxiety, communication about religion, Communication or Care Planning - or communication with the Study-specific communication with health care team. health care team items

Caregiver Outcomes: None described

300

Evidence Table 1: Populations and Interventions

Steinhauser Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant et al., US Criteria: Cancer (type not None described Control Arm - Not described Usual care Outcomes 2008142 Study Arms - 3 Patients were eligible for described), heart disease, Intervention Arm 1 - Not described Comparing discussion of inclusion if their physicians lung disease, ESRD, preparation for end of life and Level of Patient Functional Status: Intervention Arm 2 - Not described Intervention Arm 1: Randomization - estimated that they had a neurodegenerative life completion issues with a Patients prognosis of less than six disease Rosow-Breslau ADL scale Listening to relaxation facilitator, with listening to a months to live and lived within relaxation tape and usual (scale not described, higher is Multidisciplinary Team and tape Study Duration - 35 miles of Durham, North care, the study did not better): control, mean 28.4 Clinical Disciplines: Five weeks Carolina. They were excluded Prognosis: (SD 7.3); relaxation control, demonstrate any significant Risk of Bias - 2 if they unable to speak or Identified by clinicians as mean 26.3 (SD 6.3); Control Arm - No multidisciplinary Intervention Arm 2: changes in depression, severely cognitively impaired. poor prognosis, referred intervention, mean 26.1 (SD team; clinical disciplines not Discussion of anxiety, functional status, specified. to or receiving palliative 5.3). preparation for end of preparation for the end of life, care, referred to or Intervention Arm 1 - No life and life or life completion. Number of Patient receiving hospice multidisciplinary team; clinical completion issues Participants and Decedents: Caregiver Symptoms: disciplines not specified. with facilitator Screened - 209 None described Intervention Arm 2 - No Enrolled - 82 multidisciplinary team; clinical Intervention Completed - 42 disciplines not specified. Duration: Decedents - Not provided Three weeks

301

Evidence Table 2: Outcomes

Steinhauser Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant et al., US Depression: CES-D Outcomes 2008142 Study Arms - 3 Anxiety: POMS anxiety subscale Comparing discussion of Level of preparation for end of life and Functional Status: Rosow-Breslau life completion issues with a Randomization - ADL scale Patients facilitator, with listening to a Other QOL-Relevant Outcomes: relaxation tape and usual Study Duration - Preparation for the end of life care, the study did not Five weeks (QUAL-E), life completion (QUAL-E) demonstrate any significant changes in depression, Risk of Bias - 2 anxiety, functional status, Caregiver Outcomes: preparation for the end of life, None described or life completion.

302

Evidence Table 1: Populations and Interventions

Azoulay et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 143 al., 2002 EU Criteria: Acute respiratory failure, None described Control Arm - Hospital Usual care Outcomes Study Arms - 2 Patients were eligible for acute renal failure, shock Intervention Arm - Hospital Comparing a family inclusion if they had an coma information leaflet on ICU care Level of Patient Functional Status: Intervention Arm: Randomization - expected length ICU stay of at and communication with ICU Sites least 48 hours. Not described Multidisciplinary Team and Family information staff with usual care for ICU Prognosis: patients, the study Clinical Disciplines: leaflet that provided Study Duration - demonstrated a significant Not described Control Arm - No multidisciplinary general information Five days Number of Patient Caregiver Symptoms: improvement in the team; clinical disciplines not about ICU care and Participants and Decedents: intervention group in caregiver Risk of Bias - 3 None described specified. encouraged Screened - 204 communication comprehension of patient Intervention Arm - No diagnosis, prognosis, and Enrolled - 175 between families and multidisciplinary team; clinical ICU staff treatment. The study did not Completed - Not provided disciplines not specified. demonstrate a significant

Decedents - 44 change in caregiver Intervention satisfaction. Duration:

Number of Caregiver Five days Participants: Screened - 204 Enrolled - 175 Completed - 175

303

Evidence Table 2: Outcomes

Azoulay et Study Location - Patient Outcomes: Patient Outcomes: None N/A Caregiver QOL-Relevant 143 al., 2002 EU None described N/A Outcomes Study Arms - 2 Comparing a family information leaflet on ICU care Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - and communication with ICU Sites Experience or Satisfaction - CCFNI Other QOL-Relevant Outcomes: staff with usual care for ICU patients, the study Other QOL-Relevant Outcomes - Poor comprehension of patient Study Duration - demonstrated a significant Study-specific poor comprehension diagnosis, prognosis, and treatment: Five days improvement in the of patient diagnosis, prognosis, and five days, control, 40.9%; intervention, intervention group in caregiver Risk of Bias - 3 treatment measure 11.5%; p<.001. comprehension of patient diagnosis, prognosis, and treatment. The study did not demonstrate a significant change in caregiver satisfaction.

304

Evidence Table 1: Populations and Interventions

Moore et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant Outcomes 200225 EU Criteria: Comparing a nursing-led case Cancer (lung), cardiac Pain (0-100 scale, lower is Control Arm - Outpatient clinics Usual care management intervention with usual care Study Arms - 2 Patients were eligible for disease, hypertension, better): control, median 16.7 Intervention Arm - Outpatient clinics for lung cancer patients, the study inclusion if they had lung COPD, pleural effusion, (IQR 0.0-33.3); intervention, demonstrated significant improvements in Level of Intervention Arm: the intervention group in dyspnea at three Randomization - cancer, had completed their gastrointestinal disease, median 16.7 (IQR 0.0-33.3). months but not six or 12 months and initial anticancer treatment, arthritis Multidisciplinary Team and Palliative case emotional functioning at 12 months but Patients Fatigue (0-100 scale, lower is not three or six months. The study also and were expected to survive Clinical Disciplines: management Study Duration - 12 better): control, median 33.3 demonstrated significant improvements in for at least three months. (IQR 22.2-47.2); intervention, satisfaction with organization of care, months Prognosis: Control Arm - No multidisciplinary information and advice, and the personal mean 33.3 (IQR 22.2-55.6). team; clinical disciplines not Palliative case experience of care at three, six, and 12 Risk of Bias - 2 Not described specified. management months; satisfaction with overall support Number of Patient Anorexia (0-100 scale, lower at three and six months but not 12 Participants and Decedents: is better): control, median 0.0 Intervention Arm - No elements: months; and satisfaction with care at (IQR 0.0-33.3); intervention, three months but not six or 12 months. Screened - 271 multidisciplinary team; nurses. Nurse-led case The study did not demonstrate significant median 0.0 (IQR 0.0-33.3). management changes in pain, fatigue, appetite, cough, Enrolled - 203 Dyspnea (0-100 scale, lower physical functioning, or HRQOL.

Completed - 60 is better): control, median 25.0 Intervention Health Care Use Decedents - 141 (IQR 16.7-50.0); intervention, Duration: The study demonstrated a significant median 25.0 (IQR 16.7-41.7). reduction in the intervention group in 12 months Cough (0-100 scale, lower is consultations with hospital doctors and a significant increase in radiotherapy over better): control, median 33.3 the first three months of the intervention, (IQR 33.3-41.7); intervention, and a significant reduction in radiographs taken over the first six months of the median 33.3 (IQR 0.0-33.3). intervention. The study did not demonstrate significant changes in admissions to hospital or hospice; GP Patient Functional Status: visits; GP home visits; district nurse visits; Macmillan nurse or home care team EORTC-QLQ C30 physical visits; social services used; chemotherapy; scans; physiotherapist, functioning subscale (0-100 occupational therapist, or other therapist scale, higher is better): visits; or consultations with other medical control, median 86.7 (IQR staff, specialists, or therapists. 86.7-93.3); intervention, median 86.7 (IQR 86.7-93.3). Hospice Referral or Use The study did not demonstrate a significant change in admissions to hospital or hospice. Caregiver Symptoms:

None described Site of Death The study demonstrated a significant increase in the likelihood of dying at home versus in a hospital or a hospice.

Drugs The study did not demonstrate a significant change in drugs or new drugs.

Procedures The study did not demonstrate a significant change in surgeries.

Health Care Costs The study did not demonstrate a significant change in overall costs of care.

305

Evidence Table 2: Outcomes

Moore et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant Outcomes 25 2002 EU Pain - EORTC QLQ-C30 pain hospital; outpatient visits; home Consultations with hospital Comparing a nursing-led case Dyspnea: health care; admissions to management intervention with usual care Study Arms - 2 subscale doctor over first three months of for lung cancer patients, the study Dyspnea (0-100 scale, lower is better): baseline, control, hospital or hospice; district intervention: control, 0, 7.6%; 1, demonstrated significant improvements in Level of Dyspnea - EORTC QLQ-C30 median 25.0 (IQR 16.7-50.0); intervention, median 25.0 the intervention group in dyspnea at three (IQR 16.7-41.7); three months, control, median 33.3 (IQR nurse visits; social services 41.8%; 2, 21.5%; 3, 17.7%, 4 or months but not six or 12 months and Randomization - dyspnea subscale 25.0-58.3); intervention, median 25.0 (IQR 16.7-41.7); used; chemotherapy; p=.03. more, 11.4%; intervention, 0, emotional functioning at 12 months but Patients Other Symptoms - EORTC QLQ-C30 radiotherapy; radiographs taken; not three or six months. The study also 40.8%; 1, 26.3%; 2, 11.8%; 3, Study Duration - 12 fatigue, appetite, and cough scans; physiotherapist, demonstrated significant improvements in Experience or Satisfaction: 10.5%; 4 or more, 10.5%; satisfaction with organization of care, months subscales occupational therapist, or other p=.004. information and advice, and the personal Satisfaction with organization of care (0-100 scale, higher therapist visits; consultations experience of care at three, six, and 12 Risk of Bias - 2 Experience or Satisfaction - is better): baseline, control, median 71.9 (IQR 65.6-81.3); months; satisfaction with overall support intervention, median 68.8 (IQR 62.5-82.8); three months, with other medical staff, Radiotherapy over first three Unnamed satisfaction scale at three and six months but not 12 control, median 71.9 (IQR 65.6-78.1); intervention, median specialists, or therapists months of intervention: control, 81.3 (IQR 75.0-93.8); six months, control, median 75.0 months; and satisfaction with care at Functional Status - EORTC QLQ- 3.8%; intervention, 17.1%. three months but not six or 12 months. (IQR 68.8-78.1); intervention, median 83.3 (IQR 75.0- Hospice Referral or Use - C30 physical functioning subscale 93.8); 12 months, control, median 75.0 (IQR 70.3-83.3); p=.01. The study did not demonstrate significant intervention, median 81.3 (IQR 75.0-96.9); three months, Admission to hospital or hospice changes in pain, fatigue, appetite, cough, HRQOL - EORTC QLQ-C30 global intervention vs. control, p<.001; six months, intervention Radiographs taken: over first physical functioning, or HRQOL. vs. control, p<.001; 12 months, intervention vs. control, Site of Death - Died at home vs. health status/QOL subscale p=.01. three months of intervention, in a hospital or a hospice control, 0, 16.5%; 1, 53.2%; 2 or Health Care Use Other QOL-Relevant Outcomes - Satisfaction with information and advice (0-100 scale, higher is better): baseline, control, median 69.4 (IQR 62.5- Drugs - Drugs, new drugs more, 30.4%; intervention, 0, The study demonstrated a significant EORTC QLQ-C30 emotional 77.1); intervention, median 72.9 (IQR 63.0-81.3); three 38.2%; 1, 34.2%; 2, 27.6%; reduction in the intervention group in functioning subscale months, control, median 68.8 (IQR 58.3-75.0); Procedures - Surgeries consultations with hospital doctors and a intervention, median 77.1 (IQR 69.8-89.6); six months, between three and six months of significant increase in radiotherapy over control, median 66.7 (IQR 58.0-75.0); intervention, median Health Care Costs - Overall the first three months of the intervention, 75.0 (IQR 67.9-85.4); 12 months, control, median 68.8 intervention, 1 or more, control, and a significant reduction in radiographs (IQR 64.6-77.1); intervention, median 75.0 (IQR 70.8- costs of care 79%; intervention, 58%; over Caregiver Outcomes: 91.7); three months, intervention vs. control, p<.001; six taken over the first six months of the months, intervention vs. control, p<.001; 12 months, first three months of intervention. The study did not None intervention vs. control, p=.01. intervention, intervention vs. demonstrate significant changes in admissions to hospital or hospice; GP Satisfaction with personal experience of care (0-100 scale, control, p=.04; between three visits; GP home visits; district nurse visits; higher is better): baseline, control, median 75.0 (IQR 72.7- 86.4); intervention, median 77.3 (IQR 72.7-88.6); three and six months of intervention, Macmillan nurse or home care team months, control, median 75.0 (IQR 68.2-80.1); intervention vs. control, p=.03. visits; social services used; intervention, median 77.3 (IQR 75.0-95.5); six months, chemotherapy; scans; physiotherapist, control, median 75.0 (IQR 68.2-77.3); intervention, median occupational therapist, or other therapist 79.5 (IQR 72.7-97.7); 12 months, control, median 75.0 visits; or consultations with other medical (IQR 70.2-87.5); intervention, median 79.5 (IQR 75.0- Site of Death: staff, specialists, or therapists. 100.0); three months, intervention vs. control, p=.002; six months, intervention vs. control, p=.001; 12 months, Died at home vs. in a hospital or Hospice Referral or Use intervention vs. control, p=.03. a hospice: control, 23%; Satisfaction with care (0-100 scale, higher is better): The study did not demonstrate a baseline, control, median 75.0 (IQR 59.1-86.4); intervention, 40%; p=.04. significant change in admissions to intervention, median 79.5 (IQR 63.6-99.4); three months, hospital or hospice. control, median 70.0 (IQR 51.1-79.5); intervention, median 78.4 (IQR 61.6-100.0); three months, intervention vs. control, p=.005. Site of Death Satisfaction with overall support (0-100 scale, higher is The study demonstrated a significant better): baseline, control, median 90.0 (IQR 74.8-98.3); increase in the likelihood of dying at intervention, median 90.0 (IQR 72.0-100.0); three months, home versus in a hospital or a hospice. control, median 78.0 (IQR 57.0-94.0); intervention, median 93.0 (IQR 80.0-100.0); six months, control, median 83.0 (IQR 64.5-96.3); intervention, median 89.0 (IQR 82.8- Drugs 98.3); three months, intervention vs. control, p=.002; six months, intervention vs. control, p=.04. The study did not demonstrate a significant change in drugs or new drugs.

Other QOL-Relevant Outcomes: Procedures Emotional functioning (0-100 scale, higher is better): baseline, control, median 79.2 (IQR 64.6-91.7); The study did not demonstrate a intervention, median 75.0 (IQR 58.3-95.8); 12 months, significant change in surgeries.

control, median 66.7 (IQR 54.2-87.5); intervention, median 91.7 (IQR 66.7-100.0); p=.03. Health Care Costs

The study did not demonstrate a Caregiver Outcomes: significant change in overall costs of care. N/A

306

Evidence Table 1: Populations and Interventions

Allen et al., Study Location - Inclusion and Exclusion Criteria: Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 103 2014 US Patients were eligible for inclusion Cancer (type not Number of physical symptoms Control Arm - Home, residential Supportive telephone Outcomes Study Arms - 2 if they were at least 55 years old, described), heart disease, (assessed by MSAS-SF, scale care facilities calls Comparing a reminiscence were living in the community or Level of assisted living, had an advanced diabetes, arthritis not described, lower is better): Intervention Arm - Home, and creative activity Randomization - illness or combination of chronic control, mean 8.30 (SD 2.8); residential care facilities intervention aimed at illnesses, had a score of at least intervention, mean 9.27 (SD Intervention Arm: alleviating distress with Patients and Prognosis: families/caregivers three on the VES-13, had no more 5.2); p>.05. Reminiscence and supportive telephone calls for than mild cognitive impairment as Referred to or receiving Multidisciplinary Team and creative activity palliative care patients and only measured by a score of 17 or Number of emotional palliative care symptoms (assessed by Clinical Disciplines: intervention aimed at their caregivers, the study Study Duration - 30 greater on the TICS-m, received demonstrated significant an average of four hours per week MSAS-SF, scale not Control Arm - No multidisciplinary alleviating distress weeks improvements in the number of care from a caregiver, and read described, lower is better): team; clinical disciplines not and delivered by of emotional symptoms, Risk of Bias - 2 and spoke English. control, mean 1.39 (SD 1.3); specified. trained retired senior forgiveness of self or others intervention, mean 2.22 (SD volunteers Intervention Arm - No and/or experienced God’s 1.7); p>.05. Caregivers were eligible for multidisciplinary team; clinical forgiveness, and negative inclusion if they were identified by Depression (0-60 scale, lower disciplines not specified. Intervention religious coping at 17 weeks an eligible palliative care patient, is better): control, mean 11.65 were at least 19 years old Duration: but not at 30 weeks, and (SD 8.0); intervention, mean (Alabama’s age of majority), were 17 weeks bother from emotional providing an average of four hours 14.27 (SD 8.8); p>.05. symptoms at 17 and 30 per week of care for the patient, weeks. The study did not were cognitively intact as demonstrate significant Patient Functional Status: measured by a score of 28 or changes in the number of greater on the TICS-m, lived nearby, and read and spoke VES-13 score (0-10 scale, physical symptoms, bother English. lower is better): all from physical symptoms, participants, mean 7.29 (SD functional status, depression,

1.95); intervention higher than positive or negative affect, Dyads were excluded if the patient control, p=.008. daily spiritual experiences, were receiving hospice care, had a religious meaning, meaning in nursing home admission planned life, or motivation and within three months, or if either Caregiver Symptoms: individual had schizophrenia or approach to life. bipolar disorder. Number of physical symptoms (assessed by MSAS-SF, scale not described, lower is better): Caregiver QOL-Relevant Number of Patient Participants Outcomes and Decedents: control, mean 9.57 (SD 3.2); intervention, mean 9.86 (SD The study demonstrated a Screened - 76 4.2); p>.05. significant improvement in Enrolled - 45 Number of emotional meaning in life at 17 weeks Completed - 24 symptoms (assessed by but not at 30 weeks. The study Decedents - Not provided MSAS-SF, scale not did not demonstrate significant changes in number of physical described, lower is better): control, mean 2.43 (SD 1.2); symptoms, bother from Number of Caregiver physical symptoms, number of Participants: intervention, mean 1.86 (SD 1.4); p>.05. emotional symptoms, bother Screened - 76 from emotional symptoms, Depression (0-60 scale, lower Enrolled - 45 depression, positive or is better): control, mean 9.70 negative affect, religious Completed - 24 (SD 7.2); intervention, mean meaning, motivation and 9.68 (SD 7.6); p>.05. approach to life, caregiver stress, or positive aspects of caregiving.

307

Evidence Table 2: Outcomes

Allen et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 103 2014 US Depression - CES-D Outcomes Study Arms - 2 Other Symptoms - Number of Other Symptoms: Comparing a reminiscence Level of physical and emotional symptoms and creative activity Number of emotional symptoms (scale intervention aimed at Randomization - and bothered by physical and not described, lower is better): Patients and emotional symptoms (MSAS-SF) alleviating distress with baseline, control, mean 2.43 (SD 1.2); supportive telephone calls for families/caregivers Existential or Spiritual Well-Being - intervention, mean 1.86 (SD 1.4); 17 only palliative care patients and BMMRS forgiveness, abandonment, weeks, control, mean 2.00 (SD 1.2); their caregivers, the study Study Duration - 30 meaning, and daily spiritual intervention, mean 1.80 (SD 1.1); demonstrated significant weeks experiences subscales p=.02. improvements in the number Risk of Bias - 2 Functional Status - VES-13 Bothered by emotional symptoms of emotional symptoms, forgiveness of self or others Other QOL-Relevant Outcomes - (scale not described, lower is better): and/or experienced God’s Positive and negative affect baseline, control, mean 0.64 (SD 0.7); forgiveness, and negative (PANAS), motivation and approach intervention, mean 1.15 (SD 0.9); 17 religious coping at 17 weeks to life (Future Time Perspective weeks, control, mean 0.97 (SD 0.7); but not at 30 weeks, and Scale), Meaning in Life Scale intervention, mean 1.0 (SD 0.8); 17 weeks vs. baseline, p=.04; 30 weeks bother from emotional vs. baseline, p=.024. symptoms at 17 and 30 weeks. The study did not Caregiver Outcomes: demonstrate significant Depression - CES-D Existential or Spiritual Well-Being: changes in the number of Other Symptoms - Number of Forgiveness of self or others and/or physical symptoms, bother physical and emotional symptoms experienced God’s forgiveness (scale from physical symptoms, and bothered by physical and not described, higher is better): functional status, depression, emotional symptoms (MSAS-SF) baseline, control, mean 3.54 (SD 0.7); positive or negative affect, daily spiritual experiences, Existential or Spiritual Well-Being - intervention, mean 3.59 (SD 0.3); 17 religious meaning, meaning in BMMRS meaning subscale weeks, control, mean 3.50 (SD 0.7); intervention, mean 3.80 (SD 0.3); life, or motivation and Other QOL-Relevant Outcomes - p=.031. approach to life. Positive and negative affect

(PANAS), motivation and approach Negative religious coping (scale not to life (Future Time Perspective described, lower is better): baseline, Caregiver QOL-Relevant Scale), Meaning in Life Scale, control, mean 3.96 (SD 0.2); Outcomes intervention, mean 3.73 (SD 0.6); 17 Caregiver Stressors Scale-Revised, The study demonstrated a weeks, control, mean 3.83 (SD 0.5); Positive Aspects of Caregiving scale significant improvement in intervention, mean 3.80 (SD 0.4); meaning in life at 17 weeks p=.037. but not at 30 weeks. The study did not demonstrate significant Caregiver Outcomes: changes in number of physical symptoms, bother from Other QOL-Relevant Outcomes: physical symptoms, number of Meaning in life (scale not described, emotional symptoms, bother higher is better): baseline, control, from emotional symptoms, mean 3.53 (SD 0.4); intervention, depression, positive or mean 3.42 (SD 0.4); 17 weeks, control, negative affect, religious mean 3.53 (SD 0.4); intervention, meaning, motivation and mean 3.64 (SD 0.4); p=.02. approach to life, caregiver stress, or positive aspects of caregiving.

308

Evidence Table 1: Populations and Interventions

Berry et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 80 2014 US Criteria: Cancer (mixed cancer) Symptom distress (15-75 Control Arm - Outpatient clinics, Computerized Outcomes Study Arms - 2 Patients were eligible for scale, lower is better): control, home symptoms and QOL Comparing continuously inclusion if they had a mean 24.1 (SD 6.8); screening at four available computerized Level of Prognosis: Intervention Arm - Outpatient Randomization - diagnosis of cancer, were intervention, mean 24.3 (SD clinics, home predetermined time symptoms and QOL ambulatory, were at least 18 Not described 6.7). points screening, targeted education, Patients years old, were starting a new communication coaching, and Study Duration - therapeutic regimen, and Multidisciplinary Team and the opportunity to track/graph Four weeks after spoke and read English. Patient Functional Status: Clinical Disciplines: Intervention Arm: symptoms and QOL over time, end of treatment with computerized screening Not described Control Arm - No multidisciplinary Continuously at four predetermined time Risk of Bias - 3 team; physicians. available Number of Patient computerized points, the study Participants and Decedents: Caregiver Symptoms: Intervention Arm - No symptoms and QOL demonstrated a significant Screened - 2,234 None described multidisciplinary team; physicians. screening, targeted improvement in the intervention group in symptom Enrolled - 779 education, communication distress. Completed - 495 coaching, and the Decedents - 18 opportunity to track/graph

symptoms and QOL over time

Intervention Duration: Five days

309

Evidence Table 2: Outcomes

Berry et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 80 2014 US Other Symptoms - SDS-15 Other Symptoms: Outcomes Study Arms - 2 Symptom distress (15-75 scale, lower Comparing continuously available computerized Level of Caregiver Outcomes: is better): baseline, control, mean 24.1 Randomization - (SD 6.8); intervention, mean 24.3 (SD symptoms and QOL Patients None described 6.7); four weeks, control, mean 26.6 screening, targeted education, communication coaching, and Study Duration - (SD 7.5); intervention, mean 26.6 (SD 7.7); four weeks after end of treatment, the opportunity to track/graph Four weeks after symptoms and QOL over time, end of treatment control, mean 25.4 (SD 7.9); intervention, mean 24.2 (SD 6.7); with computerized screening Risk of Bias - 3 p=.02. at four predetermined time points, the study demonstrated a significant Caregiver Outcomes: improvement in the intervention group in symptom N/A distress.

310

Evidence Table 1: Populations and Interventions

Bruera et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 38 Fatigue (FACIT-F fatigue subscale, 0-52 scale, higher is 2013 US Criteria: better): placebo plus control telephone intervention, Outcomes Cancer (mixed cancer) median 21.50 (IQR 13.25-27.75); placebo plus nurse Control Arm - Home Placebo plus control Study Arms - 4 Patients were included if they telephone intervention, median 20.00 (IQR 12.00-25.00); telephone Comparing (1) placebo plus methylphenidate plus control telephone intervention, Intervention Arm 1 - Home had a diagnosis of advanced median 20.00 (SD 15.00-29.00); methylphenidate plus intervention, which telephone-based symptom Level of Prognosis: nurse telephone intervention, median 21.50 (IQR 15.00- Intervention Arm 2 - Home Randomization - cancer; a fatigue score of at 27.00); p=.87. consisted of symptom and medication assessment; least 4 on the ESAS (0-10 Fatigue (ESAS fatigue item, 0-10 scale, lower is better): and medication (2) placebo plus nurse Patients Not described placebo plus control telephone intervention, median 6.00 Intervention Arm 3 - Home scale); a normal score (at (IQR 5.00-7.00); placebo plus nurse telephone assessment telephone-based symptom intervention, median 6.00 (IQR 5.00-7.25; Study Duration - 15 least 24 of 30) on the MMSE; methylphenidate plus control telephone intervention, and medication assessment, median 6.00 (SD 5.00-7.25); methylphenidate plus nurse days a hemoglobin level of at least telephone intervention, median 7.00 (IQR 5.00-8.00); Multidisciplinary Team and psychosocial support, and 8 g/dL within two weeks of p=.56. Clinical Disciplines: Intervention Arm 1: education; (3) Risk of Bias - 4 Fatigue (ESAS fatigue item, 0-10 scale, lower is better): enrollment; no history of both control telephone intervention groups, median 6.00 Placebo plus nurse methylphenidate telephone- (IQR 5.00-8.00); both nurse telephone intervention Control Arm - No multidisciplinary tachycardia, arrhythmia, groups, median 6.00 (IQR 5.00-8.00). team; clinical disciplines not telephone based symptom and uncontrolled hypertension, Depression (HADS depression subscale, 0-21 scale, intervention, which medication assessment); and lower is better): placebo plus control telephone specified. glaucoma, severe anxiety intervention, median 7.00 (IQR 4.00-10.00); placebo plus consisted of symptom (4) methylphenidate plus nurse telephone intervention, median 6.00 (IQR 4.00- Intervention Arm 1 - No disorders, major depression, 9.00; methylphenidate plus control telephone and medication nurse telephone-based intervention, median 6.00 (SD 4.00-10.00); multidisciplinary team; nurses. or substance abuse; and no methylphenidate plus nurse telephone intervention, assessment, symptom and medication current treatment with MAOIs, median 8.00 (IQR 4.50-10.00); p=.47. Intervention Arm 2 - No psychosocial support, assessment, psychosocial Depression (ESAS depression item, 0-10 scale, lower is TCAs, clonidine, warfarin, or better): both control telephone intervention groups, multidisciplinary team; clinical and education support, and education); the erythropoietin. Patients were median 1.00 (IQR 0.00-3.25); both nurse telephone disciplines not specified. study demonstrated significant intervention groups, median 1.00 (IQR 0.00-4.00). excluded if they were improvements in anxiety, Anxiety (HADS anxiety subscale, 0-21 scale, lower is Intervention Arm 3 - No pregnant or lactating women. better): placebo plus control telephone intervention, Intervention Arm 2: nausea, drowsiness, and median 7.00 (IQR 4.00-10.00); placebo plus nurse multidisciplinary team; nurses. telephone intervention, median 6.00 (IQR 2.75-10.25; anorexia in both nurse methylphenidate plus control telephone intervention, Methylphenidate plus median 7.00 (SD 5.00-9.00); methylphenidate plus nurse telephone intervention groups telephone intervention, median 6.00 (IQR 3.00-9.00); control telephone Number of Patient p=.67. intervention as compared with both control Participants and Decedents: Anxiety (ESAS anxiety item, 0-10 scale, lower is better): telephone intervention groups. both control telephone intervention groups, median 2.00 (IQR 0.00-4.00); both nurse telephone intervention The study did not demonstrate Screened - 333 groups, median 2.00 (IQR 0.00-5.00). Intervention Arm 3: significant changes in fatigue, Enrolled - 190 Pain (ESAS pain item, 0-10 scale, lower is better): both control telephone intervention groups, median 3.00 (IQR pain, dyspnea, or depression 0.00-6.00); both nurse telephone intervention groups, Methylphenidate plus Completed - 141 median 3.00 (IQR 1.00-5.00). in both nurse telephone nurse telephone Nausea (ESAS nausea item, 0-10 scale, lower is better): intervention groups as Decedents - 1 both control telephone intervention groups, median 0.00 intervention (IQR 0.00-3.00); both nurse telephone intervention compared with both control groups, median 1.00 (IQR 0.00-3.00). telephone intervention groups. Drowsiness (ESAS drowsiness item, 0-10 scale, lower is better): both control telephone intervention groups, The study also did not median 4.00 (IQR 0.00-7.00); both nurse telephone Intervention intervention groups, median 4.00 (IQR 2.00-7.00). Duration: demonstrate significant Dyspnea (ESAS dyspnea item, 0-10 scale, lower is changes in fatigue, better): both control telephone intervention groups, 14 days median 1.00 (IQR 0.00-4.00); both nurse telephone depression, or anxiety among intervention groups, median 1.00 (IQR 0.00-4.00). the four groups. Anorexia (ESAS anorexia item, 0-10 scale, lower is better): both control telephone intervention groups, median 4.00 (IQR 1.00-5.00); both nurse telephone intervention groups, median 3.00 (IQR 1.00-5.00).

Patient Functional Status: Not described

Caregiver Symptoms: None described

311

Evidence Table 2: Outcomes

Bruera et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 38 2013 US Pain - ESAS pain item Outcomes Study Arms - 4 Dyspnea - ESAS dyspnea item Anxiety: Comparing (1) placebo plus Level of telephone-based symptom Depression - HADS depression Anxiety (0-10 scale, lower is better): and medication assessment; Randomization - subscale, ESAS depression item control telephone intervention, Patients (2) placebo plus nurse Anxiety - HADS anxiety subscale, baseline, median 2.00 (IQR 0.00-4.00); telephone-based symptom Study Duration - 15 ESAS anxiety item eight days, median 0.00 (IQR 0.00- and medication assessment, days 3.00); 15 days, median 0.00 (IQR 0.00- psychosocial support, and Other Symptoms - FACIT-F; ESAS 3.00); nurse telephone intervention, education; (3) Risk of Bias - 4 fatigue, nausea, drowsiness, and baseline, median 2.00 (IQR 0.00-5.00); methylphenidate telephone- anorexia items eight days, median 0.00 (IQR 0.00- based symptom and 3.00); 15 days, median 0.00 (IQR 0.00- medication assessment); and 3.00); p=.01. Caregiver Outcomes: (4) methylphenidate plus nurse telephone-based None described symptom and medication Other Symptoms: assessment, psychosocial Nausea (0-10 scale, lower is better): support, and education); the control telephone intervention, study demonstrated significant baseline, median 0.00 (IQR 0.00-3.00); improvements in anxiety, eight days, median 0.00 (IQR 0.00- nausea, drowsiness, and 2.00); 15 days, median 0.00 (IQR 0.00- anorexia in both nurse 1.50); nurse telephone intervention, telephone intervention groups baseline, median 1.00 (IQR 0.00-3.00); as compared with both control eight days, median 0.00 (IQR 0.00- telephone intervention groups. 1.00); 15 days, median 0.00 (IQR 0.00- The study did not demonstrate 2.00); p=.01. significant changes in fatigue, Drowsiness (0-10 scale, lower is pain, dyspnea, or depression better): control telephone intervention, in both nurse telephone baseline, median 4.00 (IQR 0.00-7.00); intervention groups as eight days, median 3.00 (IQR 0.00- compared with both control 5.00); 15 days, median 1.00 (IQR 0.00- telephone intervention groups. 5.00); nurse telephone intervention, The study also did not baseline, median 4.00 (IQR 2.00-7.00); demonstrate significant eight days, median 3.00 (IQR 0.00- changes in fatigue, 4.00); 15 days, median 2.00 (IQR 0.00- depression, or anxiety among 3.00); p<.001. the four groups. Anorexia (0-10 scale, lower is better): control telephone intervention, baseline, median 4.00 (IQR 1.00-5.00); eight days, median 3.00 (IQR 0.00- 5.75); 15 days, median 2.00 (IQR 0.00- 6.00); nurse telephone intervention, baseline, median 3.00 (IQR 1.00-5.00); eight days, median 2.00 (IQR 0.00- 4.00); 15 days, median 2.00 (IQR 0.00- 5.00); p=.009.

Caregiver Outcomes: N/A

312

Evidence Table 1: Populations and Interventions

Curtis et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 119 Criteria: 2013 US Cancer (type not None described Control Arm - Not specified Usual education for Outcomes Patients were eligible for Study Arms - 2 described), CHF, COPD, Intervention Arm - Not specified physician and nurse Comparing simulation-based inclusion if they had a high other lung disease, ESLD trainees communication skills Level of likelihood of having a discussion Patient Functional Status: Randomization - education with usual about end-of-life care; had a education for internal medicine Providers median survival of Not described Multidisciplinary Team and Intervention Arm: Prognosis: and nurse practitioner approximately 1 to 2 years as Clinical Disciplines: Study Duration - 10 Simulation-based trainees, the study determined by life-limiting illness Identified by clinicians as Control Arm - No multidisciplinary months Caregiver Symptoms: communication skills demonstrated a significant (e.g., metastatic or stage IV poor prognosis team; clinical disciplines not education worsening in the intervention Risk of Bias - 3 cancer, oxygen-dependent None described specified. COPD, stage III or IV CHF, or group in depressive Intervention Arm - No symptoms. The study did not Child-Pugh class C liver Intervention disease) or comorbidities multidisciplinary team; physicians, demonstrate significant Duration: suggesting severe illness (score nurses. changes in quality of of at least 5 on the CCI); had Eight sessions communication about end-of- documentation of life care, physical status, or communication about end-of-life mental status. care (palliative care consult or DNR order), had an ICU stay of 72 hours or longer, were at least Caregiver QOL-Relevant 80 years old with a hospital stay Outcomes of 72 hours or longer, or were outpatients with at least three The study did not demonstrate visits with the trainee; and significant changes in remembered the trainee well depression or quality of enough to evaluate his or her communication about end-of- communication skills. life care.

Trainees were included if they were internal medicine residents or fellows in pulmonary and critical care, oncology, geriatrics, nephrology, or palliative medicine subspecialties; or were nurse practitioners who were currently enrolled in or had recently completed training programs that included care for adults with life-threatening or chronic illnesses.

Number of Trainee Participants: Screened - 1,068 Enrolled - 472 Completed - 194

313

Evidence Table 2: Outcomes

Curtis et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 119 2013 US Depression - PHQ-8 Outcomes Study Arms - 2 Communication or Care Planning - Depression: Comparing simulation-based Level of QOC questionnaire communication skills Depressive symptoms (0-24 scale, education with usual Randomization - Functional Status - SF-12 PCS lower is better): post-intervention, Providers education for internal medicine Other QOL-Relevant Outcomes - control, mean 8.8 (95% CI 8.4-9.2); and nurse practitioner Study Duration - 10 SF-12 MCS intervention, mean 10.0 (95% CI 9.1- trainees, the study months 10.8); post-intervention vs. baseline, demonstrated a significant

Risk of Bias - 3 adjusted mean +2.2 (95% CI 0.6-3.8). worsening in the intervention Caregiver Outcomes: group in depressive symptoms. The study did not Depression - PHQ-8 Caregiver Outcomes: demonstrate significant Communication or Care Planning - None changes in quality of QOC questionnaire communication about end-of- life care, physical status, or mental status.

Caregiver QOL-Relevant Outcomes The study did not demonstrate significant changes in depression or quality of communication about end-of- life care.

314

Evidence Table 1: Populations and Interventions

Donovan et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 39 al., 2014 US Criteria: Cancer (ovarian) Severity of three symptoms Control Arm - Home Usual care Outcomes Study Arms - 2 Patients were eligible for “noticed most” (0-10 scale, Intervention Arm - Home Comparing a nurse-based inclusion if they had recurrent lower is better): control, mean online symptom education Level of Prognosis: Intervention Arm: Randomization - or persistent ovarian cancer 5.69 (SD 2.60); intervention, intervention with usual care for (disease persisted or recurred Not described mean 6.05 (SD 2.45); p=.56. Multidisciplinary Team and WRITE Symptoms, a women with ovarian cancer, Patients nurse-based online after primary therapy), had Clinical Disciplines: the study did not demonstrate Study Duration - three or more concurrent symptom education any significant changes in the Control Arm - No multidisciplinary Nine weeks symptoms related to cancer or Patient Functional Status: intervention severity of, distress from, team; clinical disciplines not its treatment, were able to consequences of, or Risk of Bias - 4 Not described specified. read and write in English, had controllability of the three Intervention consistent access to the Intervention Arm - No symptoms “noticed most.” Caregiver Symptoms: multidisciplinary team; nurses. Duration: Internet and email (defined as any kind of access through None described Six weeks home or work), and lived in a state in which study nurses had licensure (patients were enrolled in 25 states).

Number of Patient Participants and Decedents: Screened - 271 Enrolled - 65 Completed - 56 Decedents - Not provided

315

Evidence Table 2: Outcomes

Donovan et Study Location - Patient Outcomes: None None N/A Patient QOL-Relevant 39 al., 2014 US Other Symptoms - Severity of, Outcomes Study Arms - 2 distress from, consequences of, and Comparing a nurse-based Level of controllability of three symptoms online symptom education Randomization - “noticed most” (SRQ) intervention with usual care for Patients women with ovarian cancer, the study did not demonstrate Study Duration - Caregiver Outcomes: any significant changes in the Nine weeks None described severity of, distress from, Risk of Bias - 4 consequences of, or controllability of the three symptoms “noticed most.”

316

Evidence Table 1: Populations and Interventions

Epstein et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 144 al., 2013 US Criteria: Cancer (pancreatic, None described Control Arm - Outpatient clinics Decision support Outcomes Study Arms - 2 Patients were eligible for hepatobiliary) Intervention Arm - Outpatient clinics Comparing an educational Level of inclusion if they spoke CPR video with an educational English, were at least 21 Patient Functional Status: Decision support CPR narrative for cancer Randomization - Prognosis: elements: Patients years old, had no prior ECOG performance status: Multidisciplinary Team and patients, the study did not advance directive designation, Identified by clinicians as control, mean 1 (range 0-3); Clinical Disciplines: Educational CPR demonstrate any significant Study Duration - and had progressive exocrine changes in knowledge of poor prognosis intervention, mean 1 (range 0- Control Arm - No multidisciplinary narrative Six months pancreas or hepatobiliary CPR, preference for CPR, 2); p>.05. team; clinical disciplines not cancers (defined as a life preference for mechanical Risk of Bias - 10 specified. Intervention Arm: expectancy of less than one ventilation, rate of advance year as judged by the treating Caregiver Symptoms: Intervention Arm - No Decision support directive documentation oncologist, as well as any of None described multidisciplinary team; clinical and/or documented advance the following: unresectable disciplines not specified. care planning discussions, disease with progression on at Decision support number of advance directives least one course of treatment elements: completed within six months [surgery, radiation, or Video-based decision post-intervention, number of chemotherapy]; untreated aid documented advance care disease unfit for therapy planning discussions

because of comorbidities or completed after one month ECOG performance status Intervention post-intervention, time to greater than 2; and ECOG Duration: completion of advance performance status of 2). One-time intervention directives, or time between They were excluded if they advance directive completion had an SPMSQ score of less and death. than “intact mental functioning” (at least three errors) or had any condition Health Care Use (e.g., concern for The study did not demonstrate psychological distress) that any significant changes in the treating oncologist hospital admissions, hospital deemed would make the study length of stay, or ICU inappropriate for them. admissions.

Number of Patient Site of Death Participants and Decedents: The study did not demonstrate Screened - 82 a significant change in the Enrolled - 57 likelihood of dying in hospice care versus an inpatient Completed - 54 hospital. Decedents - 29 Life-Extending Devices or Procedures The study did not demonstrate any significant changes in rates of CPR or mechanical ventilation.

317

Evidence Table 2: Outcomes

Epstein et Study Location - Patient Outcomes: None Health Care Use - Inpatient None Patient QOL-Relevant 144 al., 2013 US Communication or Care Planning - hospital, ICU Outcomes Study Arms - 2 Study-specific CPR knowledge Site of Death - Died in hospice Comparing an educational Level of questionnaire; study-specific vs. inpatient hospital CPR video with an educational Randomization - preference for CPR, preference for Life-Extending Devices or CPR narrative for cancer Patients mechanical ventilation, advance Procedures - Mechanical patients, the study did not directive documentation and/or ventilation, CPR demonstrate any significant Study Duration - documented advance care planning changes in knowledge of Six months discussions, number of advance CPR, preference for CPR, Risk of Bias - 10 directives completed, number of preference for mechanical documented advance care planning ventilation, rate of advance discussions completed, time to directive documentation completion of advance directives, and/or documented advance and time between advance directive care planning discussions, completion and death measures number of advance directives completed within six months post-intervention, number of Caregiver Outcomes: documented advance care None described planning discussions completed after one month post-intervention, time to completion of advance directives, or time between advance directive completion and death.

Health Care Use The study did not demonstrate any significant changes in hospital admissions, hospital length of stay, or ICU admissions.

Site of Death The study did not demonstrate a significant change in the likelihood of dying in hospice care versus an inpatient hospital.

Life-Extending Devices or Procedures The study did not demonstrate any significant changes in rates of CPR or mechanical ventilation.

318

Evidence Table 1: Populations and Interventions

Gellis et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 83 2014 US Criteria: CHF, hypertension, Depression: (0-50 scale, lower Control Arm - Home Usual care, Outcomes Study Arms - 2 Patients were eligible for COPD, diabetes, arthritis is better): control, mean 17.8 Intervention Arm - Home psychoeducation Comparing a telemonitoring, Level of inclusion if they were at least (SD 6.7); intervention, mean about disease and care management, and 65 years old, had a primary 18.1 (SD 6.9); p>.05. general health depression treatment Randomization - Prognosis: Patients diagnosis of heart failure or Depression: (0-27 scale, lower Multidisciplinary Team and monitoring intervention, with usual care, COPD, were above-average Not described is better): control, mean 15.2 Clinical Disciplines: psychoeducation about Study Duration - 12 users of health services disease, and general health (SD 5.8); intervention, mean Control Arm - No multidisciplinary Intervention Arm: months (defined as at least 10 days in monitoring, for patients with 14.9 (SD 6.4); p>.05. team; nurses. Risk of Bias - 9 the hospital in the past 12 Telemonitoring, CHF and/or COPD, the study months, an ED visit in the last Intervention Arm - No chronic illness and demonstrated significant two months, or at least three Patient Functional Status: multidisciplinary team; nurses. depression care improvements in the home care visits per week), management, and intervention group in Katz ADL scale (0-6 scale, and had a PHQ-2 score of at problem-solving depression and overall mental lower is better): control, mean least three. They were treatment for health. The study did not 3.3 (SD 1.8); intervention, excluded if they had an MMSE comorbid depression demonstrate significant mean 3.5 (SD 1.6); p>.05. score of less than 24 or a changes in overall physical diagnosis of dementia based SF-12 PCS (0-100 scale, health or satisfaction with Intervention on chart review, were unable higher is better): control, mean care. 42.1 (SD 11.5); intervention, Duration: to use a tele-monitoring device because of physical mean 42.3 (SD 12.1); p>.05. Three months Health Care Use disability, or had behavioral problems (e.g., agitation, The study demonstrated a Caregiver Symptoms: delirium, or paranoia) that significant reduction in the would interfere with use of the None described intervention group in ED visits device. in the 12 month period after baseline versus the 12 month period before baseline. The Number of Patient study did not demonstrate Participants and Decedents: significant changes in hospital Screened - 236 days or episodes of care Enrolled - 115 between those times frames. Completed - 94 Decedents - Not provided

319

Evidence Table 2: Outcomes

Gellis et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 83 2014 US Depression - HAM-D, PHQ-9 hospital, ED, episodes of care ED visits: 12 months pre- Outcomes Study Arms - 2 Experience or Satisfaction - Study- Depression: baseline to baseline, control, Comparing a telemonitoring, Level of specific satisfaction with provider mean 0.8 (SD 0.8); intervention, care management, and Depression (0-50 scale, lower is mean 0.7 (SD 9.3); baseline to depression treatment Randomization - instructions, satisfaction with better): baseline, control, mean 17.8 Patients instructions on privacy, care helped 12 months post-baseline, intervention, with usual care, (SD 6.7); intervention, mean 18.1 (SD control, mean 1.4 (SD 1.2); psychoeducation about Study Duration - 12 me stay healthier, better 6.9); three months, control, mean 18.6 understanding of my condition, care intervention, mean 0.6 (SD 1.6); disease, and general health months (SD 5.7); intervention, mean 9.8 (SD p=.03. monitoring, for patients with made my life better, and overall 5.6); six months, control, mean 17.4 Risk of Bias - 9 satisfaction items CHF and/or COPD, the study (SD 6.3); intervention, mean 10.4 (SD demonstrated significant Functional Status - SF-12 PCS 7.1); three months vs. baseline, p=.02; improvements in the Other QOL-Relevant Outcomes - six months vs. baseline, p=.05. intervention group in SF-12 MCS Depression (0-27 scale, lower is depression and overall mental health. The study did not better): baseline, control, mean 15.2 (SD 5.8); intervention, mean 14.9 (SD demonstrate significant Caregiver Outcomes: 6.4); three months, control, mean 13.6 changes in overall physical None described (SD 5.6); intervention, mean 7.4 (SD health or satisfaction with 5.7); six months, control, mean 14.1 care. (SD 5.9); intervention, mean 7.9 (SD 5.3); three months vs. baseline, p=.01; six months vs. baseline, p=.05. Health Care Use The study demonstrated a significant reduction in the Other QOL-Relevant Outcomes: intervention group in ED visits Overall mental health (0-100 scale, in the 12 month period after higher is better): baseline, control, baseline versus the 12 month mean 39.4 (SD 8.1); intervention, period before baseline. The mean 39.7 (SD 7.6); three months, study did not demonstrate control, mean 42.8 (SD 20.7); significant changes in hospital intervention, mean 53.6 (SD 21.7); six days or episodes of care months, control, mean 40.3 (SD 27.4); between those times frames. intervention, mean 52.1 (SD 24.3); three months vs. baseline, p=.01; six months vs. baseline, p=.05.

Caregiver Outcomes: N/A

320

Evidence Table 1: Populations and Interventions

Hollingworth Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 93 et al., 2013 EU Criteria: Cancer (mixed cancer) Anxiety/depression: control, Control Arm - Outpatient clinics Usual care Outcomes Study Arms - 2 Patients were eligible for 13.0%; intervention, 9.8%. Intervention Arm - Outpatient clinics Comparing a distress inclusion if they were between screening and needs Level of Prognosis: Tension/anxiety (scale not Intervention Arm: Randomization - 18 and 85 years old, had a described, lower is better): assessment intervention with primary solid tumor diagnosis Not described Multidisciplinary Team and Distress screening, usual care for cancer patients, Patients control, mean 4.8 (SD 5.0); needs assessment, within the previous 12 months, intervention, mean 4.5 (SD Clinical Disciplines: the study demonstrated a Study Duration - 12 were receiving outpatient and action plan significant improvement in the 4.4). Control Arm - No multidisciplinary months external radiotherapy over a intervention group in physical team; clinical disciplines not period of at least two weeks or Depression/dejection (scale functioning. The study did not Risk of Bias - 6 specified. Intervention outpatient chemotherapy of at not described, lower is better): demonstrate significant Duration: least two cycles, and could control, mean 3.4 (SD 5.7); Intervention Arm - No changes in HRQOL, total read and communicate in intervention, mean 3.7 (SD multidisciplinary team; nurses, One-time intervention mood disturbance, English. They were excluded if 5.5). radiographers. depression/dejection, they were diagnosed with Fatigue/inertia (scale not tension/anxiety, fatigue/inertia, ductal carcinoma in situ or described, lower is better): confusion/bewilderment, skin carcinoma or receiving control, mean 5.7 (SD 5.7); vigor/activity, emotional neoadjuvant chemotherapy. intervention, mean 5.6 (SD functioning, social functioning, 5.8). or experience of care. Number of Patient Confusion/bewilderment Participants and Decedents: (scale not described, lower is Health Care Costs better): control, mean 3.2 (SD Screened - 636 3.6); intervention, mean 3.2 The study did not demonstrate Enrolled - 220 (SD 3.4). any significant changes in overall costs of care or in Completed - 209 inpatient hospital, outpatient, Decedents - 7 Patient Functional Status: ED, medication, GP, nurse, psychologist, or other EORTC QLQ-C30 physical community care costs. functioning scale (0-100 scale, higher is better): control, mean 81.4 (SD 19.9); intervention, mean 85.3 (SD 17.0).

Caregiver Symptoms: None described

321

Evidence Table 2: Outcomes

Hollingworth Study Location - Patient Outcomes: Patient Outcomes: Health Care Costs - Total costs None Patient QOL-Relevant 93 et al., 2013 EU Depression - POMS-SF of care, inpatient hospital costs, Outcomes Study Arms - 2 depression/dejection subscale outpatient costs, ED costs, GP Comparing a distress Functional Status: costs, nurse costs, psychologist Level of Anxiety - POMS-SF tension/anxiety screening and needs Physical functioning (0-100 scale, costs, other community care assessment intervention with Randomization - subscale higher is better): baseline, control, costs, medication costs Patients usual care for cancer patients, Other Symptoms - Total mood mean 81.4 (SD 19.9); intervention, the study demonstrated a Study Duration - 12 disturbance (POMS-SF), POMS-SF mean 85.3 (SD 17.0); one month, significant improvement in the months fatigue/inertia and control, mean 80.7 (SD 20.5); intervention group in physical Risk of Bias - 6 confusion/bewilderment subscales intervention, mean 81.9 (SD 20.5); six functioning. The study did not Experience or Satisfaction - Trent months, control, mean 83.8 (SD 18.6); demonstrate significant Patient Views of Cancer Services intervention, mean 84.2 (SD 19.0); 12 changes in HRQOL, total Questionnaire months, control, mean 85.5 (SD 17.8); mood disturbance, intervention, mean 83.8 (SD 19.3); depression/dejection, Functional Status - EORTC QLQ- p=.031. tension/anxiety, fatigue/inertia, C30 physical functioning domain confusion/bewilderment, HRQOL - EORTC QLQ-C30, EQ-5D vigor/activity, emotional Caregiver Outcomes: Other QOL-Relevant Outcomes - functioning, social functioning, POMS-SF vigor/activity subscale, N/A or experience of care. EORTC QLQ-C30 emotional functioning and social functioning Health Care Costs domains The study did not demonstrate

any significant changes in Caregiver Outcomes: overall costs of care or in None described inpatient hospital, outpatient, ED, medication, GP, nurse, psychologist, or other community care costs.

322

Evidence Table 1: Populations and Interventions

Julião et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 52 2014 EU Criteria: Cancer (type not Depression (0-21 scale, lower Control Arm - Hospital Palliative care team Outcomes Study Arms - 2 Patients were eligible for described), ALS, is better): control, median Intervention Arm - Hospital Comparing inpatient palliative inclusion if they were at least trigeminal neuralgia 14.00 (IQR 11.00-16.00); care plus dignity therapy with Level of Palliative care team Randomization - 18 years old, had a life- intervention, median 14.00 inpatient palliative care alone, threatening disease with a (IQR 9.00-16.00); p>.05. Multidisciplinary Team and elements: the study demonstrated Patients Prognosis: prognosis of six months or Anxiety (0-21 scale, lower is Clinical Disciplines: Inpatient consultation significant improvements in Study Duration - 30 less, had no evidence of the intervention group in Identified by clinicians as better): control, median 9.00 Control Arm - Palliative care team; days dementia or delirium (as depression and anxiety. poor prognosis, referred (IQR 5.00-12.00); intervention, physicians, nurses, social workers, determined by chart review or Intervention Arm: Risk of Bias - 4 to or receiving palliative median 10.00 (IQR 6.00- chaplains. clinical consensus of the care 13.00); p>.05. Palliative care team, palliative care team), had an Intervention Arm - Palliative care dignity therapy MMSE score of at least 20, team; physicians, nurses, social

were able to read and speak Patient Functional Status: workers, chaplains. Portuguese, were able to Palliative care team PPS score (0-100 scale, provide written informed elements: higher is better): control, mean consent, and were available 57.6 (SD 18.2); intervention, Inpatient consultation for four to five research mean 56.7 (SD 16.9); p=.82. encounters over a period of one month. Intervention Caregiver Symptoms: Duration: Number of Patient None described Not described Participants and Decedents: Screened - 92 Enrolled - 80 Completed - 36 Decedents - 4

323

Evidence Table 2: Outcomes

Julião et al., Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 52 2014 EU Depression - HADS depression Outcomes Study Arms - 2 subscale Depression: Comparing inpatient palliative Level of Anxiety - HADS anxiety subscale care plus dignity therapy with Depression (0-21 scale, lower is inpatient palliative care alone, Randomization - better): baseline, control, median 14.00 Patients the study demonstrated Caregiver Outcomes: (IQR 11.00-16.00); intervention, significant improvements in Study Duration - 30 median 14.00 (IQR 9.00-16.00); four the intervention group in None described days days, control, median 14.00 (IQR depression and anxiety. Risk of Bias - 4 11.00-17.00); intervention, median 10.00 (IQR 7.00-11.00); 15 days, control, median 15.00 (IQR 10.50- 18.00); intervention, median 10.00 (IQR 8.25-13.75); 30 days, control, median 17.00 (IQR 12.00-18.00); intervention, median 10.00 (IQR 9.00- 13.00); four days vs. baseline, p<.001; 15 days vs. baseline, p=.01; 30 days vs. baseline, p=.043.

Anxiety: Anxiety (0-21 scale, lower is better): control, median 9.00 (IQR 5.00-12.00); intervention, median 10.00 (IQR 6.00- 13.00); four days, control, median 9.00 (IQR 6.00-11.25); intervention, median 5.00 (IQR 3.50-8.00); 15 days, control, median 9.50 (IQR 6.00-15.00); intervention, median 5.00 (IQR 4.00- 8.00); 30 days, control, median 10.00 (IQR 5.50-13.00); intervention, median 4.00 (IQR 3.00-7.00); four days vs. baseline, p<.001; 15 days vs. baseline, p=.001; 30 days vs. baseline, p=.013.

Caregiver Outcomes: N/A

324

Evidence Table 1: Populations and Interventions

Kinley et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 145 2014 EU Criteria: Dementia None described Control Arm - Nursing home Facilitation of GSFCH Outcomes Study Arms - 2 Nursing homes were eligible Intervention Arm - Nursing home program Comparing facilitation of for inclusion if they were implementation in GSFCH program Level of Prognosis: Patient Functional Status: Randomization - registered as a care home nursing homes implementation in nursing Sites with nursing providing care for Not described Not described Multidisciplinary Team and homes and action learning for frail older residents, had a nurse managers, with Clinical Disciplines: Study Duration - manager willing to participate Intervention Arm: facilitation of GSFCH program Control Arm - No multidisciplinary Three years in nine monthly action learning Caregiver Symptoms: implementation alone, the team; nurses, GSFCH facilitators, Facilitation of GSFCH sets (the intervention), and study demonstrated a Risk of Bias - 8 None described research assistant. program were based in one of the five implementation in significant improvement in the Clinical Commissioning Intervention Arm - No nursing homes and intervention group in use of an Groups local to the regional multidisciplinary team; nurses, action learning for end-of-life care plan training center (so that a GSFCH facilitators, research nurse managers documentation pathway. The specific facilitation model assistant. study did not demonstrate

could be provided). Clinical significant changes in the Commissioning Groups are Intervention completion of advance care NHS organizations set up by Duration: plans, documentation of CPR the Health and Social Care Three years preferences in advance care Act 2012 to organize the plans, or use of advance care delivery of NHS services in plans. England. Site of Death Number of Nursing Homes: The study did not demonstrate Screened - 56 a significant change in the Enrolled - 24 likelihood of dying in a nursing home. Completed - 24

325

Evidence Table 2: Outcomes

Kinley et al., Study Location - Patient Outcomes: Patient Outcomes: Site of Death - Death in nursing None Patient QOL-Relevant 145 2014 EU Communication or Care Planning - home Outcomes Study Arms - 2 Study-specific completed advance Communication or Care Planning: Comparing facilitation of Level of care plan, documentation of CPR GSFCH program preference in advance care plan, Used end-of-life care plan implementation in nursing Randomization - documentation pathway (ICP, MP, or Sites used advance care plan, and used homes and action learning for end-of-life care plan documentation LCP): year one, OR 1 (normalized); nurse managers, with Study Duration - pathway (ICP, MP, or LCP) year two, OR 2.026 (95% CI 0.828- facilitation of GSFCH program Three years measures 4.956); year three, OR 6.945 (95% CI implementation alone, the 1.627-29.642); p=.036. Risk of Bias - 8 study demonstrated a significant improvement in the Caregiver Outcomes: intervention group in use of an Caregiver Outcomes: None described end-of-life care plan N/A documentation pathway. The study did not demonstrate significant changes in the completion of advance care plans, documentation of CPR preferences in advance care plans, or use of advance care plans.

Site of Death The study did not demonstrate a significant change in the likelihood of dying in a nursing home.

326

Evidence Table 1: Populations and Interventions

Knapp et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 115 2013 EU Criteria: Dementia None described Control Arm - Not specified Usual care Outcomes Study Arms - 2 Caregivers were eligible for Intervention Arm - Home Comparing a therapy program inclusion if they identified for caregivers of dementia Level of Prognosis: Patient Functional Status: Intervention Arm: Randomization - themselves as the primary patients with usual care, the Families or carer of a family member with Not described Not described Multidisciplinary Team and Therapy program for study demonstrated a dementia not living in 24-hour significant improvement in the caregivers Clinical Disciplines: caregivers of care and provided emotional intervention group in affective Control Arm - No multidisciplinary dementia patients Study Duration - or practical support at least Caregiver Symptoms: symptoms. The study did not team; clinical disciplines not Eight months weekly. They were excluded if demonstrate a significant None described specified. Risk of Bias - 8 they were unable to give Intervention change in HRQOL. Intervention Arm - No Duration: informed consent to the trial, were currently taking part in multidisciplinary team; clinical 14 weeks an RCT in their capacity as a psychologist, psychology graduates Health Care Use carer, lived more than 1.5 with no clinical training. The study had unclear effects hours traveling time from the on outpatient hospital use and researchers’ base, or had use of community-based dementia (as determined by a health care services. score of less than 24 on the

MMSE and clinician assessment). Health Care Costs The study did not demonstrate significant changes in overall, Number of Caregiver outpatient, community, or Participants: other costs. Screened - 472 Enrolled - 260 Completed - 229

327

Evidence Table 2: Outcomes

Knapp et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Outpatient None Caregiver QOL-Relevant 115 2013 EU None described N/A visits, community-based health Outcomes care services Study Arms - 2 Comparing a therapy program Health Care Costs - Overall for caregivers of dementia Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - costs, outpatient costs, patients with usual care, the Families or Other Symptoms - HADS total score community costs, other costs study demonstrated a significant improvement in the caregivers HRQOL - EQ-5D Other Symptoms: intervention group in affective Study Duration - Affective symptoms (0-42 scale, lower symptoms. The study did not Eight months is better): baseline, control, mean 14.8 demonstrate a significant Risk of Bias - 8 (SD 7.4); intervention, mean 13.5 (SD change in HRQOL. 7.3); four months, control, mean 14.3

(SD 7.4); intervention, mean 12.4 (SD 7.4); eight months, control, mean 14.9 Health Care Use (SD 8.0); intervention, mean 12.9 (SD The study had unclear effects 7.9); p=.05. on outpatient hospital use and use of community-based health care services.

Health Care Costs The study did not demonstrate significant changes in overall, outpatient, community, or other costs.

328

Evidence Table 1: Populations and Interventions

Livingston et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Caregiver QOL-Relevant 114 al., 2013 EU Criteria: Dementia Neuropsychiatric symptoms Control Arm - Not specified Usual care Outcomes Study Arms - 2 Caregivers were eligible for (1-144 scale, lower is better): Intervention Arm - Home Comparing a therapy program inclusion if they identified control, mean 26.6 (SD 20.1); for caregivers of dementia Level of Prognosis: Intervention Arm: Randomization - themselves as the primary intervention, mean 24.0 (SD patients with usual care, the Families or carer of a family member with Not described 19.0). Multidisciplinary Team and Therapy program for study demonstrated significant caregivers dementia not living in 24-hour Clinical Disciplines: caregivers of improvements in the care and provided emotional intervention group in Control Arm - No multidisciplinary dementia patients Study Duration - or practical support at least Patient Functional Status: depression and case-level team; clinical disciplines not Eight months weekly. They were excluded if depression, affective Not described specified. they were unable to give Intervention symptoms, and mental health. Risk of Bias - 7 informed consent to the trial, Intervention Arm - No Duration: The study did not demonstrate were currently taking part in Caregiver Symptoms: multidisciplinary team; clinical 14 weeks significant changes in anxiety an RCT in their capacity as a Anxiety (0-21 scale, lower is psychologist, psychology graduates or case-level anxiety, HRQOL, carer, lived more than 1.5 better): control, mean 9.3 (SD with no clinical training. or potentially abusive behavior hours traveling time from the 4.3); intervention, mean 8.1 by caregivers toward patients. researchers’ base, or had (SD 4.4). dementia (as determined by a score of less than 24 on the Case-level anxiety (score of at Health Care Use least 9 on the HADS anxiety MMSE and clinician The study did not demonstrate subscale, 0-21 scale, lower is assessment). a significant change in the rate better): control, 55%; of nursing home admissions. intervention, 49%. Number of Caregiver Depression (0-21 scale, lower Participants: is better): control, mean 5.5 Screened - 472 (SD 3.9); intervention, mean 5.4 (SD 3.8). Enrolled - 260 Case-level depression (score Completed - 229 of at least 9 on the HADS depression subscale, 0-21 scale, lower is better): control, 20%; intervention, 21%.

329

Evidence Table 2: Outcomes

Livingston et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Nursing home None Caregiver QOL-Relevant 114 al., 2013 EU None described N/A Outcomes Study Arms - 2 Comparing a therapy program for caregivers of dementia Level of Caregiver Outcomes: Caregiver Outcomes: Randomization - patients with usual care, the Families or Depression - HADS depression study demonstrated significant improvements in the caregivers subscale Depression: intervention group in Anxiety - HADS anxiety subscale Study Duration - Depression (0-21 scale, lower is depression and case-level Eight months Other Symptoms - HADS total score better): baseline, control, mean 5.5 depression, affective Risk of Bias - 7 HRQOL - QOL-AD (SD 3.9); intervention, mean 5.4 (SD symptoms, and mental health. 3.8); four months, control, mean 5.7 The study did not demonstrate Other QOL-Relevant Outcomes - (SD 4.0); intervention, mean 4.9 (SD significant changes in anxiety Health Status Questionnaire mental 3.9); eight months, control, mean 6.1 or case-level anxiety, HRQOL, health domain, potentially abusive (SD 4.2); intervention, mean 5.2 (SD or potentially abusive behavior behavior by caregivers toward 4.0); p<.05. by caregivers toward patients. patients (MCTS) Case-level depression (score of at least 9 on the HADS depression Health Care Use subscale, 0-21 scale, lower is better): baseline, control, 20%; intervention, The study did not demonstrate 21%; four months, control, 24%; a significant change in the rate intervention, 17%; eight months, of nursing home admissions. control, 32%; intervention, 21%; p<.05.

Other Symptoms: Affective symptoms (0-42 scale, lower is better): baseline, control, mean 14.8 (SD 7.4); intervention, mean 13.5 (SD 7.3); four months, control, mean 14.3 (SD 7.4); intervention, mean 12.4 (SD 7.4); eight months, control, mean 14.9 (SD 8.0); intervention, mean 12.9 (SD 7.9); p=.05.

Other QOL-Relevant Outcomes: Mental health (0-100 scale, higher is better): baseline, control, mean 58.2 (SD 21.7); intervention, mean 58.3 (SD 22.4); four months, control, mean 58.4 (SD 18.0); intervention, mean 62.7 (SD 20.8); eight months, control, mean 58.2 (SD 19.2); intervention, mean 58.6 (SD 22.0); p<.05.

330

Evidence Table 1: Populations and Interventions

Metzelthin et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 146 al., 2013 EU Criteria: Not described Depression (0-21 scale, lower Control Arm - Not specified Usual care Outcomes Study Arms - 2 Patients were eligible for is better): control, mean 6.69 Intervention Arm - Home Comparing a nurse-based inclusion if they were (SD 4.35); intervention, mean formulation of individualized Level of Prognosis: Intervention Arm: Randomization - community-dwelling, were at 6.54 (SD 3.77); p>.05. treatment plans with usual Sites least 70 years old, had a Not described Multidisciplinary Team and Palliative case care for considerably frail Groningen Frailty Indicator Clinical Disciplines: management, nurse- patients, the study did not Study Duration - 24 score of at least five Patient Functional Status: demonstrate any significant Control Arm - No multidisciplinary based formulation of months (indicating considerable changes in functional status, GARS ADL scale (11-44 team; clinical disciplines not individualized frailty), and signed the depression, social Risk of Bias - 5 scale, lower is better): control, specified. treatment plan informed consent form. They participation, social support mean 16.54 (SD 5.35); were excluded if they were intervention, mean 17.97 (SD Intervention Arm - No interactions, or fear of falling. terminally ill, were confined to 6.14); p<.05. multidisciplinary team; physicians, Palliative case bed, had severe cognitive or nurses, occupational therapists, management

psychological impairments, or physical therapists, pharmacists. elements: were unable to communicate Caregiver Symptoms: Nurse-led case in Dutch. None described management

Number of Patient Intervention Participants and Decedents: Duration: Screened - 3,498 Six sessions Enrolled - 346 Completed - 270 Decedents - 25

331

Evidence Table 2: Outcomes

Metzelthin et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 146 al., 2013 EU Depression - HADS depression None Outcomes Study Arms - 2 subscale Comparing a nurse-based Level of Functional Status - GARS ADL scale formulation of individualized Caregiver Outcomes: treatment plans with usual Randomization - Other QOL-Relevant Outcomes - Sites N/A care for considerably frail Social participation (MSPP), social patients, the study did not Study Duration - 24 support interactions (SSL-I12), fear demonstrate any significant months of falling (Short FES-I) changes in functional status, Risk of Bias - 5 depression, social participation, social support Caregiver Outcomes: interactions, or fear of falling. None

332

Evidence Table 1: Populations and Interventions

Mitchell et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 62 al., 2008 Australia Criteria: Not described None described Control Arm - Outpatient clinics Usual care, palliative Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient clinics care team Comparing GP teleconference Level of inclusion if they were referred during a palliative care team to participating palliative care Prognosis: Patient Functional Status: case consultation with usual Randomization - Palliative care team GP/patient dyad units, had a life expectancy of Referred to or receiving Not described Multidisciplinary Team and palliative care, the study at least one month, were elements: demonstrated significant palliative care Clinical Disciplines: Study Duration - capable of giving informed improvements in the Control Arm - Palliative care team; Outpatient Nine weeks consent, and had GPs and Caregiver Symptoms: intervention group in pain, clinical disciplines not specified. consultation principal carers (if present) nausea, and constipation. The Risk of Bias - 4 None described who gave permission for study Intervention Arm - Palliative care study did not demonstrate staff to approach them. They team; physicians. Intervention Arm: significant changes in were excluded if they were too Palliative care team, HRQOL, subjective well-being, ill to participate in an interview GP teleconference degree of worry, or difficulty of approximately 15 minutes during a palliative sleeping. duration. care team case consultation Caregiver QOL-Relevant Number of Patient Outcomes Participants and Decedents: Palliative care team The study demonstrated a Screened - 1,137 elements: significant improvement in the Enrolled - 159 Outpatient intervention group in caregiver consultation burden. Completed - 22

Decedents - 95 Intervention Duration: One-time intervention

333

Evidence Table 2: Outcomes

Mitchell et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 62 al., 2008 Australia Pain - AQEL pain item Outcomes Study Arms - 2 Anxiety - AQEL degree of worry item Pain: Comparing GP teleconference Level of during a palliative care team Other Symptoms - AQEL nausea Pain (scale not described): seven case consultation with usual Randomization - and moving bowels items weeks vs. baseline, mean change 2.61 GP/patient dyad palliative care, the study HRQOL - McGill QOL Questionnaire, in favor of intervention; p=.005. demonstrated significant Study Duration - AQEL improvements in the Nine weeks intervention group in pain, Other QOL-Relevant Outcomes - Other Symptoms: nausea, and constipation. The Risk of Bias - 4 Subjective Well-Being Scale, AQEL Nausea (scale not described): 0 weeks study did not demonstrate difficulty sleeping item vs. baseline, mean change 1.13 in significant changes in favor of control; seven weeks vs. HRQOL, subjective well-being, Caregiver Outcomes: baseline, mean change 2.14 in favor of degree of worry, or difficulty intervention; 0 weeks vs. baseline, sleeping. Other QOL-Relevant Outcomes - p=.004; seven weeks vs. baseline,

Caregiver Reaction Assessment p=.044. Caregiver QOL-Relevant Constipation (scale not described): Outcomes nine weeks vs. baseline, mean change 1.13 in favor of control; nine weeks vs. The study demonstrated a baseline, p=.02. significant improvement in the intervention group in caregiver

burden. Caregiver Outcomes:

Other QOL-Relevant Outcomes: Caregiver burden (scale not described): five weeks, lower in intervention vs. control; p<.05.

334

Evidence Table 1: Populations and Interventions

Nguyen et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 40 al., 2013 US Criteria: CAD, hypertension, Dyspnea (5-35 scale, higher is Control Arm - Outpatient clinics, General health Outcomes Study Arms - 3 Patients were eligible for COPD, arthritis better): general health home education unrelated Comparing a dyspnea self- Level of inclusion if they had a education, mean 23.3 (95% CI Intervention Arm - Outpatient to lung disease management program with diagnosis of COPD; were 21.5-25.0); fDSMP, mean 22.2 Web-based components, a Randomization - Prognosis: clinics, home Patients clinically stable for at least one (95% CI 20.4-23.9); eDSMP, dyspnea self-management month; had an FEV1/FVC Not described mean 23.8 (95% CI 22.1- Intervention Arm 1: program without such Study Duration - 12 ratio of less than 0.70 with 25.5). Multidisciplinary Team and fDSMP, a dyspnea components, and general months FEV1 less than 80% of Clinical Disciplines: self-management health education on topics predicted, or an FEV1/FVC unrelated to lung disease, for Risk of Bias - 3 Control Arm - No multidisciplinary program with ratio less than 0.60 with FEV1 Patient Functional Status: patients with COPD, the study team; clinical disciplines not education and greater than 80% of predicted did not demonstrate any SF-36 PCS (0-100 scale, specified. exercise components or CT-confirmed emphysema; significant changes in higher is better): general had activities limited by health education, mean 33.7 Intervention Arm 1 - No dyspnea, HRQOL, overall dyspnea; could use the (95% CI 30.9-36.6); fDSMP, multidisciplinary team; advanced Intervention Arm 2: physical or mental health, or Internet; and had oxygen mean 34.2 (95% CI 31.4- practice nurses. eDSMP, a dyspnea self-efficacy for managing saturation greater than 85% 37.1); eDSMP, mean 34.5 Intervention Arm 2 - No self-management dyspnea. on room air on at most six (95% CI 31.8-37.3). multidisciplinary team; advanced program with liters/minute of oxygen at the practice nurses. education and end of a six-minute walk test. exercise components They were excluded if they Caregiver Symptoms: that incorporated had active symptomatic illness None described Web-based education (e.g., cancer or heart failure), modules and email had participated in pulmonary and chat contact with rehabilitation in the last six nurses months, or were currently participating in more than two days a week of supervised Intervention exercise. Duration: 12 months Number of Patient Participants and Decedents: Screened - 711 Enrolled - 125 Completed - 110 Decedents - 3

335

Evidence Table 2: Outcomes

Nguyen et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 40 al., 2013 US Dyspnea - CRQ dyspnea subscale None Outcomes Study Arms - 3 Functional Status - SF-36 PCS Comparing a dyspnea self- management program with Level of HRQOL - CRQ Caregiver Outcomes: Randomization - Web-based components, a Patients Other QOL-Relevant Outcomes - N/A dyspnea self-management SF-36 MCS, unnamed self-efficacy program without such Study Duration - 12 for managing dyspnea question components, and general months health education on topics

Risk of Bias - 3 unrelated to lung disease, for Caregiver Outcomes: patients with COPD, the study None described did not demonstrate any significant changes in dyspnea, HRQOL, overall physical or mental health, or self-efficacy for managing dyspnea.

336

Evidence Table 1: Populations and Interventions

Stanton et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 99 al., 2013 US Criteria: Cancer (breast) Depression (0-60 scale, lower Control Arm - Not specified Usual care Outcomes Study Arms - 2 Patients were eligible for is better): control, mean 12 Intervention Arm - Outpatient clinics Comparing a workshop for inclusion if they were at least (SD 10); intervention, mean hands-on creation of a Level of Prognosis: Intervention Arm: Randomization - 18 years old, had a diagnosis 11 (SD 9); p>.05. personal Web site with usual Patients of invasive or metastatic Not described Multidisciplinary Team and Workshop for hands- care for patients with breast breast cancer (any interval Clinical Disciplines: on creation of cancer, the study Study Duration - since diagnosis), could Patient Functional Status: demonstrated significant Control Arm - No multidisciplinary personal Web site to Six months complete the intervention and improvements in the Not described team; clinical disciplines not communicate illness questionnaires in English, had intervention group in Risk of Bias - 5 specified. experience to friends no existing personal Web site, and loved ones depression, vigor, and life and returned the baseline Caregiver Symptoms: Intervention Arm - No appreciation. The study did

questionnaire within one None described multidisciplinary team; clinical not demonstrate significant month of receipt. disciplines not specified. Intervention changes in negative mood, Duration: cancer-related intrusive One-time intervention thoughts and feelings, or Number of Patient relating to others. Participants and Decedents: Screened - 440 Enrolled - 88 Completed - 76 Decedents - Not provided

337

Evidence Table 2: Outcomes

Stanton et Study Location - Patient Outcomes: Patient Outcomes: None N/A Patient QOL-Relevant 99 al., 2013 US Depression - CES-D Outcomes Study Arms - 2 Other QOL-Relevant Outcomes - Depression: Comparing a workshop for Level of POMS vigor and negative mood hands-on creation of a Depression (0-60 scale, lower is personal Web site with usual Randomization - subscales, PTGI life appreciation better): baseline, control, mean 12 (SD Patients and relating to others subscales, care for patients with breast 10); intervention, mean 11 (SD 9); six cancer, the study Study Duration - cancer-related intrusive thoughts months, control, mean 15.5 (SE 1.7); and feelings (IES intrusion subscale) demonstrated significant Six months intervention, mean 9.7 (SE 1.6); improvements in the Risk of Bias - 5 p=.009. intervention group in Caregiver Outcomes: depression, vigor, and life appreciation. The study did None described Other QOL-Relevant Outcomes: not demonstrate significant Vigor (scale not described, higher is changes in negative mood, better): baseline, control, mean 18 (SD cancer-related intrusive 6); intervention, mean 16 (SD 8); six thoughts and feelings, or months, control, mean 14.7 (SE 1.2); relating to others. intervention, mean 18.2 (SE 1.2); p=.03. Life appreciation (scale not described, higher is better): baseline, control, mean 11 (SD 3); intervention, mean 11 (SD 3); six months, control, mean 8.7 (SE 0.5); intervention, mean 9.8 (SE 0.5); p=.03.

Caregiver Outcomes: N/A

338

Evidence Table 1: Populations and Interventions

Stein et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 109 2013 Australia Criteria: Cancer (mixed cancer) Depression (0-21 scale, lower Control Arm - Not specified Usual care Outcomes Study Arms - 2 Patients were eligible for is better): control, mean 5.2 Intervention Arm - Outpatient clinics Comparing a pamphlet on inclusion if they had a (SD 3.5); intervention, mean living with advanced cancer Level of Prognosis: Intervention Arm: Randomization - diagnosis of metastatic 4.3 (SD 2.9); p=.40. and a discussion with a Patients cancer, were not undergoing Identified by clinicians as Anxiety (0-21 scale, lower is Multidisciplinary Team and Other care planning, psychologist to encourage further curative treatment, had poor prognosis better): control, mean 5.3 (SD Clinical Disciplines: pamphlet on living consideration of preferences Study Duration - 12 an estimated life expectancy and values toward the end of 3.0); intervention, mean 4.9 Control Arm - No multidisciplinary with advanced cancer weeks of three to 12 months, were life, with usual care for (SD 3.4); p=.60. team; clinical disciplines not and discussion with aware of their prognosis, and patients with metastatic Risk of Bias - 6 specified. psychologist to were literate in English. encourage cancer, the study Intervention Arm - No Patient Functional Status: consideration of demonstrated a significant multidisciplinary team; psychologist. increase in the intervention Not described preferences and Number of Patient values toward the group in the time between Participants and Decedents: end of life DNR order placement and death. The study did not Screened - 159 Caregiver Symptoms: demonstrate significant Enrolled - 120 None described Intervention changes in the number of Completed - 64 Duration: DNR orders placed, Decedents - 19 One-time intervention depression, or anxiety.

Number of Caregiver Caregiver QOL-Relevant Participants: Outcomes Screened - 99 The study demonstrated a significant improvement in Enrolled - 97 disruption to caregivers’ Completed - 49 schedules. The study did not demonstrate significant changes in caregiver health, self-esteem, or family support.

Site of Death The study demonstrated a significant reduction in the rate of hospital deaths.

339

Evidence Table 2: Outcomes

Stein et al., Study Location - Patient Outcomes: Patient Outcomes: Site of Death - Death in hospital Site of Death: Patient QOL-Relevant 109 2013 US Depression - HADS depression Death in hospital: control, 50%; Outcomes Study Arms - 2 subscale Communication or Care Planning: intervention, 19%; p=.003. Comparing a pamphlet on Level of Anxiety - HADS anxiety subscale living with advanced cancer Time between DNR order placement and a discussion with a Randomization - Communication or Care Planning - and death: control, median 12.5 days; Patients psychologist to encourage Study-specific DNR order placement intervention, median 27.0 days; group consideration of preferences Study Duration - measure, Study-specific time by time p=.03. and values toward the end of Six months between DNR order placement and life, with usual care for death measure patients with metastatic Risk of Bias - 5 Caregiver Outcomes: cancer, the study demonstrated a significant Caregiver Outcomes: Other QOL-Relevant Outcomes: increase in the intervention Other QOL-Relevant Outcomes - group in the time between CRA health, self-esteem, family Schedule disruption (scale not DNR order placement and support, and schedule disruption described, lower is better): baseline, death. The study did not subscales control, mean 11.8 (SD 5.5); demonstrate significant intervention, mean 13.6 (SD 4.6); three changes in the number of weeks, control, mean 12.9 (SD 5.9); DNR orders placed, intervention, mean 12.5 (SD 3.9); 12 depression, or anxiety. weeks, control, mean 13.6 (SD 6.4); intervention, mean 11.9 (SD 6.4); group by time p=.05. Caregiver QOL-Relevant Outcomes The study demonstrated a significant improvement in disruption to caregivers’ schedules. The study did not demonstrate significant changes in caregiver health, self-esteem, or family support.

Site of Death The study demonstrated a significant reduction in the rate of hospital deaths.

340

Evidence Table 1: Populations and Interventions

Uitdehaag et Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant Outcomes 26 al., 2014 EU Criteria: Cancer (upper None described Control Arm - Outpatient clinics, Usual care Comparing nurse-led follow up with Study Arms - 2 Patients were eligible for gastrointestinal) home conventional follow-up for patients with upper gastrointestinal cancer, inclusion if they had Level of Patient Functional Status: Intervention Arm - Home Intervention Arm: the study demonstrated significant unresectable or recurrent improvements in the intervention Randomization - Prognosis: Patients upper gastrointestinal cancer Not described Palliative case group in overall satisfaction and and a multidisciplinary panel burden of follow-up visits. The Not described Multidisciplinary Team and management Study Duration - 13 concluded that a curative study did not demonstrate Clinical Disciplines: months modality or disease-modifying Caregiver Symptoms: significant changes in HRQOL; Control Arm - No multidisciplinary cognitive functioning; depression; anti-tumor therapy (i.e., Palliative case Risk of Bias - 3 None described team; physicians. diarrhea; satisfaction with palliative chemotherapy, management availability of care; satisfaction with radiotherapy, or surgery) was Intervention Arm - No elements: knowledge of, confidence given by, not or no longer possible. multidisciplinary team; nurses. Nurse-led case information given by, advice given They were excluded if they management by, answers received from, and were admitted to a nursing support received from the home or hospice, could not be physician/nurse; satisfaction with involvement by the physician/nurse followed by a physician at an Intervention in care planning; or satisfaction outpatient clinic, or were Duration: with management of physical unable to understand the 13 months complaints. Dutch language and complete questionnaires. Caregiver QOL-Relevant Outcomes Number of Patient The study demonstrated significant Participants and Decedents: improvements in the intervention group in overall satisfaction and Screened - 204 burden of follow-up visits. The Enrolled - 138 study did not demonstrate significant changes in satisfaction Completed - 5 with availability of care; satisfaction with knowledge of, confidence Decedents - 88 given by, information given by, advice given by, answers received from, and support received from the physician/nurse; satisfaction with involvement by the physician/nurse in care planning; or satisfaction with management of physical complaints.

Health Care Use The study did not demonstrate significant changes in GP visits or hospital admissions.

Health Care Costs The study had an unclear effect on total follow-up costs per patient.

341

Evidence Table 2: Outcomes

Uitdehaag et Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient None Patient QOL-Relevant Outcomes 26 al., 2014 EU Depression - HADS depression hospital, outpatient visits Comparing nurse-led follow up with Study Arms - 2 subscale Health Care Costs - Total follow- conventional follow-up for patients Experience or Satisfaction: with upper gastrointestinal cancer, up costs per patient Level of Other Symptoms - EORTC QLQ-C30 Overall satisfaction (1-10 scale, higher the study demonstrated significant improvements in the intervention Randomization - diarrhea item is better): 1.5 months, control, mean Patients group in overall satisfaction and Experience or Satisfaction - Study- 7.5 (SD 2.02); intervention, mean 8.4 burden of follow-up visits. The Study Duration - 13 specific satisfaction questionnaire, (SD 0.95); four months, control, mean study did not demonstrate months including satisfaction with availability 7.1 (SD 1.18); intervention, mean 8.5 significant changes in HRQOL; cognitive functioning; depression; Risk of Bias - 3 of care; knowledge of, confidence (SD 1.03); intervention vs. control, 1.5 given by, information given by, months, p=.02; four months, p<.001. diarrhea; satisfaction with availability of care; satisfaction with advice given by, answers received Burden of follow-up visit (1-10 scale, knowledge of, confidence given by, from, and support received from higher is better): 1.5 months, control, information given by, advice given physician/nurse; involvement by mean 6.0 (SD 2.65); intervention, by, answers received from, and physician/nurse in care planning; mean 8.4 (SD 1.90); four months, support received from the physician/nurse; satisfaction with and management of physical control, mean 5.2 (SD 2.69); complaints items involvement by the physician/nurse intervention, mean 8.5 (SD 1.42); in care planning; or satisfaction HRQOL - EQ-5D intervention vs. control, 1.5 months, with management of physical Other QOL-Relevant Outcomes - p<.001; four months, p<.001. complaints. EORTC QLQ-C30 cognitive functioning scale Caregiver Outcomes: Caregiver QOL-Relevant Outcomes The study demonstrated significant Caregiver Outcomes: Experience or Satisfaction: improvements in the intervention group in overall satisfaction and Experience or Satisfaction - Study- Overall satisfaction (1-10 scale, higher specific satisfaction questionnaire burden of follow-up visits. The is better): 1.5 months, control, mean study did not demonstrate 6.8 (SD 2.32); intervention, mean 8.0 significant changes in satisfaction (SD 1.13); four months, control, mean with availability of care; satisfaction 6.9 (SD 2.38); intervention, mean 8.5 with knowledge of, confidence (SD 0.98); intervention vs. control, 1.5 given by, information given by, advice given by, answers received months, p=.02; four months, p=.01. from, and support received from Burden of follow-up visit (1-10 scale, the physician/nurse; satisfaction higher is better): 1.5 months, control, with involvement by the physician/nurse in care planning; or mean 4.9 (SD 2.44); intervention, satisfaction with management of mean 8.4 (SD 1.56); four months, physical complaints. control, mean 4.8 (SD 2.64);

intervention, mean 8.8 (SD 0.98); intervention vs. control, 1.5 months, Health Care Use p<.001; four months, p<.001. The study did not demonstrate significant changes in GP visits or hospital admissions.

Health Care Costs The study had an unclear effect on total follow-up costs per patient.

342

Evidence Table 1: Populations and Interventions

Yount et al., Study Location - Inclusion and Exclusion Condition(s): Patient Symptoms: Setting(s) of Care: Control Arm: Patient QOL-Relevant 41 2014 US Criteria: Cancer (lung) None described Control Arm - Home Telephone-based Outcomes Study Arms - 2 Patients were eligible for Intervention Arm - Outpatient symptom monitoring Comparing telephone-based Level of inclusion if they were at least clinics, home symptom monitoring and 18 years old, spoke English, Prognosis: Patient Functional Status: reporting of symptoms to Randomization - Intervention Arm: Patients had Stage III or IV NSCLC or Not described ECOG performance status (0- providers with telephone- small cell lung cancer, were 5 scale, lower is better): Multidisciplinary Team and Telephone-based based symptom monitoring Study Duration - 12 receiving active treatment with control, 0, 30.2%; 1, 44.2%; 2, Clinical Disciplines: symptom monitoring alone for patients with weeks traditional chemotherapy no advanced lung cancer, the 16.3%; 3, 9.3%; 4, 0.0%; Control Arm - No multidisciplinary and reporting of later than Day 1 of Cycle 2 or study demonstrated significant Risk of Bias - 4 intervention, 0, 31.7%; 1, team; physicians, physician symptoms to receiving oral therapy, had worsenings in the intervention 43.9%; 2, 16.3%; 3, 7.3%; 4, assistants, nurses. providers access to a telephone, and group in perceptions of having 0.8%; p=.79. had a life expectancy of at Intervention Arm - No concerns understood and

least six months. multidisciplinary team; physicians, Intervention receipt of adequate Caregiver Symptoms: physician assistants, nurses. Duration: information, time and support. None described 12 weeks The study did not demonstrate Number of Patient significant changes in lung Participants and Decedents: cancer symptom burden; Screened - 420 symptom distress; HRQOL; physical, functional, and Enrolled - 253 emotional well-being; Completed - 191 attitudinal barriers to symptom Decedents - 3 management; self-efficacy for patient/physician communication, health behaviors, and knowledge about accessing care; or satisfaction with provider explanations and interpersonal relationships with providers.

Health Care Use The study demonstrated a significant increase in the intervention group in phone calls to nurses. The study did not demonstrate significant changes in hospital admissions, ED visits, unscheduled clinic visits, or number of phone calls to physicians.

343

Evidence Table 2: Outcomes

Yount et al., Study Location - Patient Outcomes: Patient Outcomes: Health Care Use - Inpatient Health Care Use: Patient QOL-Relevant 41 2014 US Other Symptoms - Lung cancer hospital, ED, outpatient visits, Phone calls to nurses over 12 Outcomes Study Arms - 2 symptom burden (FLSI), SDS phone calls to physicians, phone weeks: control, mean 1.14 (SD Comparing telephone-based Experience or Satisfaction: calls to nurses Level of Experience or Satisfaction - FACIT- 1.85); intervention, mean 1.79 symptom monitoring and Perceptions of having concerns (SD 2.52); p=.022. reporting of symptoms to Randomization - TS-PS comprehensive care and understood (scale not described, Patients interpersonal subscales, FACT-G providers with telephone- higher is better): 12 weeks, control, based symptom monitoring Study Duration - 12 physical well-being and functional mean 2.48 (SD 0.67); intervention, well-being subscales alone for patients with weeks mean 2.20 (SD 0.78); p=.012. advanced lung cancer, the Risk of Bias - 4 HRQOL - FACT-G Received adequate information, time study demonstrated significant Other QOL-Relevant Outcomes - and support (scale not described, worsenings in the intervention FACT-G emotional well-being higher is better): 12 weeks, control, group in perceptions of having subscale; attitudinal barriers to mean 2.52 (SD 0.61); intervention, concerns understood and symptom management (SMBQ); mean 2.27 (SD 0.86); p=.027. receipt of adequate information, time and support. study-specific self-efficacy for patient/physician communication, The study did not demonstrate health behaviors, and knowledge Caregiver Outcomes: significant changes in lung about accessing care measures N/A cancer symptom burden; symptom distress; HRQOL;

physical, functional, and Caregiver Outcomes: emotional well-being; None attitudinal barriers to symptom management; self-efficacy for patient/physician communication, health behaviors, and knowledge about accessing care; or satisfaction with provider explanations and interpersonal relationships with providers.

344

Chapter 4. Estimating the Value of Palliative Care for Older Adults: What Does the Evidence Support?

Abstract

Background: A 2014 Institute of Medicine report concluded that broad improvements in end‐of‐life care are within reach. Many currently available evidence‐based palliative health services interventions have not been implemented on a large scale.

Objective: Informed by the systematic review in Chapter 3, I project changes from 2016 to 2040 in health care costs, quality of life, and mortality resulting from the implementation of a PCT for cancer patients, PME for cancer patients, and CM for CHF patients.

Setting: Future Elderly Model, a microsimulation model that tracks cohorts of Health and Retirement

Study participants and projects their health status and quality of life and economic outcomes.

Outcome Measures: Societal and OOP health care costs; QALYs; population size, median age, and mortality; pain, depression, and ADL dependencies. For each measure, changes are given as average

(low, high) based on a range of RCT parameters.

Results: From 2016 to 2040 in both cancer and CHF patients in the absence of any intervention, societal and OOP health care costs rose, QALYs decreased, and mortality fell and population size of both patient populations increased. By design, the PCT intervention directly reduced depression in cancer patients by

75.3% (sensitivity analysis, 70.4%‐80.1%), the PME intervention reduced pain in cancer patients by

35.9% (13.6%, 45.5%), and the CM intervention reduced average ADL dependencies in CHF patients by

32.4% (18.3%, 35.1%). Each intervention had indirect effects on all other outcomes: from 2016 to 2040, as compared with the status quo, all three interventions reduced per capita societal andP OO health care costs; increased QALYs; reduced mortality and increased population size; and reduced average

345

pain, ADL dependencies, and the prevalence of depression. The PCT intervention saved $126 billion in societal health care costs ($118 billion, $134 billion) and $15.0 billion in OOP health care costs ($14.0 billion, $15.9 billion), and the CM scenario saved $35.9 billion in societal health care costs ($35.9 billion,

$52.3 billion) and $6.13 billion in OOP health care costs ($6.13 billion, $6.36 billion).

Conclusion: Cost and quality of life outcomes in end‐of‐life patients are largely projected to worsen over time in the absence of any intervention. Current RCT evidence indicates that palliative health services interventions could make positive and often large impacts on these outcomes over time.

346

Introduction

A 2014 IOM report on the state of end‐of‐life care acknowledged the shortcomings of end‐of‐life care as it is currently delivered but also concluded that broad improvements are within reach. It also defined a set of core components of high quality end‐of‐life care. Among these are referral to expert‐ level palliative care, management of pain and other symptoms, and round‐the‐clock access to coordinated care and services.

Here, I analyze the effect of implementing three palliative health services interventions identified Chapter 3 on future health care costs and quality of life. Each of the interventions considered here addresses a different core component from the IOM’s set. Together, they reflect evidence‐based approaches of varying comprehensiveness for addressing the needs of chronically ill older adults.

Policymakers, payers, and providers can use information on the future impact of these interventions to decide which ones best suit the needs of their patient populations as they develop, refine, and deliver end‐of‐life services.

The interventions are:

 An inpatient PCT for cancer patients, which addresses the core component of referral to expert‐

level palliative care.

 PME for cancer patients, which addresses the core component of pain management.

 CM for CHF patients, which addresses the core component of access to coordinated care and

services.

347

Methods

Future Elderly Model

I projected changes in all outcomes using the FEM, a microsimulation model that tracks cohorts of people and projects their health status and economic outcomes. The FEM has been used previously to examine the health and economic impacts of delayed aging,1 longer periods of drug exclusivity,2 improved disease prevention,3 and Medicare program changes,4 among others.

The FEM models nationally representative cohorts of people aged 51 and 52 using data from the

HRS, a biennial survey of community‐dwelling adults aged 51 or older in the contiguous United States.5 Given initial demographic characteristics and health states, the FEM predicts medical spending, QALYs health conditions, functional status, and symptomatology for the next two years.

Medical spending predictions use data from the Medical Expenditure Panel Survey6 for non‐

Medicare beneficiaries and from the Medicare Current Beneficiary Survey7 for Medicare beneficiaries.

Health states are derived from regression models that use HRS survey questions about demographics, health conditions, symptoms, and functional status, and are modeled using first‐order Markov processes that control for baseline unobserved factors.

Full details of the FEM design and implementation have been published elsewhere.1

Evidence‐Based Scenarios

In addition to the status quo scenarios, I developed three intervention scenarios derived from the systematic review in Chapter 3. In choosing interventions from the systematic review to model in the FEM, I used the following selection criteria:

 Interventions had to have an effect on patient quality of life outcomes available for modeling in

the FEM (namely, pain, depression, or functional status). Many interventions in the review were

348

effective for non‐clinical outcomes such as communication or satisfaction with care, but these

could not be implemented in the FEM.

 Interventions had to be studied in multiple studies or in a well‐known, high quality study.

 Studies had to report both baseline and post‐intervention outcome values to allow for

calculation of percent change in outcomes as a result of the intervention.

 Studies had to analyze all patients as a whole in reporting outcomes (i.e., not divided into

different groups according to race/ethnicity, gender, etc.) so as to be most relevant to the entire

FEM population of participants.

 Studies had to measure outcomes in a manner that could be translated to the FEM’s

measurement of outcomes. Pain had to be measured using average pain scales, depression had

to be measured in terms of prevalence, and functional status had to be measured using physical

functioning scales.

Status Quo Scenarios

These scenarios represent the status quo and serve as baselines for comparison with the intervention scenarios. I model two status quo scenarios: one in cancer patients, and one in CHF patients. Both scenarios’ mortality forecasts derive from all‐cause mortality in the intermediate projections of the Social Security Administration.8

Palliative Care Team Scenario

This scenario models the implementation of a PCT alongside standard oncologic care for patients diagnosed with cancer. In the systematic review, I identified one flagship RCT of early access to

PCT alongside standard oncologic care.9 This study measured the prevalence of depression using two different measures and found reductions of 66.6% and 79.2%.

As a proxy for care delivered by a PCT, I reduced the prevalence of depression in cancer patients in the FEM models. I implemented the average effect size (72.9%) as well as dthe high an low effect sizes

349

as a sensitivity analysis. This scenario assumes that the sole impact of a PCT is a reduction in the prevalence of depression.

Pain Management Education Scenario

This scenario models the implementation of PME for patients diagnosed with cancer. In the systematic review, I identified five RCTs of PME interventions designed for and delivered to cancer patients at home or in outpatient settings.10‐14 All of these studies significantly reduced pain (i.e., none of them made pain worse) and had a majority of significant pain results. Three of the five studies met all of my selection criteria and demonstrated reductions in average pain of 19.8%,10 34.4%,11 and 42.0%.12

As a proxy for PME, I reduced pain in cancer patients in the FEM models. I could not directly implement the average effect size (32.1%) or the high and low effect sizes because the three studies reported average pain reductions using a continuous 0‐10 visual analog scale, whereas the FEM uses a categorical 0‐3 pain scale with levels of no, mild, moderate, and severe pain. Instead, I approximated these reductions using the following approaches:

 Low effect size: I reduced severe pain to moderate pain and left the other pain levels

unchanged. This led to a reduction in average pain of approximately 13.6%.

 Average effect size: I reduced moderate pain to mild pain and left the other pain levels

unchanged. This led to a reduction in average pain of approximately 35.9%.

 High effect size: I reduced severe pain to moderate pain and moderate pain to mild pain, and

left the other pain levels unchanged. This led to a reduction in average pain of approximately

45.5%.

This scenario assumes that the sole impact of PME is a reduction in average pain.

Case Management Scenario

This scenario models the implementation of CM for patients diagnosed with CHF. In the systematic review, I identified four RCTs of CM interventions designed for and delivered to CHF

350

patients.15‐18 All of these studies significantly improved physical functioning (i.e., none of them made physical functioning worse) and had a majority of significant physical functioning results. Three of the four studies met all of my selection criteria and demonstrated improvements in average physical functioning of 20.7%,15 30.3%,16 and 32.7%.17

As a proxy for CM, I reduced average ADL dependencies in CHF patients in the FEM models. I could not directly implement the average effect size (27.9%) or the high and low effect sizes because the three studies reported average physical functioning improvements using a continuous 0‐40 scale, whereas the FEM uses a categorical 0‐6 scale which sums six different ADL dependencies. Instead, I approximated these reductions using the following approaches:

 Low effect size: on the categorical 0‐6 scale, I reduced values of 3 to 2, 4 to 3, 5 to 4, and 6 to 5.

This led to a reduction in average ADL dependencies of approximately 18.3%.

 Average effect size: on the categorical 0‐6 scale, I reduced values of 2 to 1, 3 to 2, 4 to 3, 5 to 4,

and 6 to 4. This led to a reduction in average pain of approximately 32.4%.

 High effect size: on the categorical 0‐6 scale, I reduced values of 2 to 1, 3 to 2, 4 to 3, 5 to 3, and

6 to 4. This led to a reduction in average pain of approximately 35.1%.

This scenario assumes that the sole impact of CM is a reduction in average ADL dependencies.

All Scenarios

All scenarios extend through 2040 and are modeled in all FEM participants diagnosed with either cancer or CHF, depending on the scenario, who have a mortality probability greater than the median mortality probability in their entry cohort.

Outcomes

In each scenario, I analyzed the following outcomes:

 Average pain, prevalence of depression, and average ADL dependencies

 Population size, median age, and mortality

351

 Societal health care costs (defined as the sum of Medicare and Medicaid costs and private

insurance payments) and OOP medical costs (in 2014 dollars)

 QALYs

Measures

Pain was captured with a 0‐3 categorical pain scale in which 0 represents no pain, 1 represents mild pain, 2 represents moderate pain, and 3 represents severe pain.19,20 Depression was captured with a score of four or more on an eight‐item subset of the 20‐item CES‐D.21 ADL dependencies were captured with a 0‐6 scale comprised of help with dressing, walking, bathing, eating, transferring, and toileting. Costs were captured using the Medical Expenditure Panel Survey6 for non‐Medicare beneficiaries and from the Medicare Current Beneficiary Survey7 for Medicare beneficiaries. QALYs were captured with the EQ‐5D.22 OOP costs were captured directly; societal costs were calculated by subtracting OOP costs from total costs, which are the sum of Medicare and Medicaid costs, private insurance payments, and OOP costs.

Analysis

Main figures plot changes in each outcome from 2016 to 2040 in status quo and intervention scenarios and are grouped according to patient population (cancer or CHF). Appendix figures are organized in the same fashion and plot the percent change in each outcome in intervention versus status quo scenarios in each time step. In both sets of graphs, average effect size scenarios are plotted as solid (main figures) or dashed lines (appendix figures). The area between high and low effect size scenarios is shaded to suggest a range of possible effects.

Appendix tables present FEM regression models for mortality and societal and OOP health care costs. I offer full models to highlight the impact of different covariates on these outcomes, and focus on mortality and costs because the mechanisms through which pain, depression, and ADL dependencies

352

affect these outcomes are less straightforward than the mechanisms by which they affect QALYs and one another.

All results pertain to patients ages 65 or older unless otherwise noted. Effect sizes are given as the average effect size over the course of the simulation unless designated for certain years. Bounds on estimates from high and low effect size scenarios are given in the style of 95% confidence intervals where relevant.

I conducted analyses with Stata 12.1 IC (StataCorp, College Station, Texas).

353

Results

Pain, Depression, and ADL Dependencies

In the status quo scenarios, most of the outcomes in this section worsened in both cancer and

CHF patients. Mean pain scores and ADL dependencies increased from 2016 to 2040 (cancer patients: pain, 0.68 to 0.77; ADL dependencies, 0.69 to 0.94; CHF patients: pain, 0.85 to 0.92; ADL dependencies,

1.13 ).to 1.36 The prevalence of depression increased slightly in cancer patients (15.0% to 15.4%) and decreased in CHF patients (19.8% to 18.7%).

By design, each intervention scenario directly affected one of the outcomes in this section: the

PCT scenario reduced the prevalence of depression, the PME scenario reduced average pain, and the

CM scenario reduced average ADL dependencies. Accordingly, there were large reductions observed in those outcomes in those scenarios as compared to the status quo scenarios (Table 1).

Each intervention scenario also had indirect effects on the two outcomes that it did not directly impact. Average, high, and low effect size scenarios in all interventions yielded reductions in all outcomes, except for the low effect size scenario of CM in CHF patients, which yielded a small increase in the prevalence of depression in CHF patients.

Population Size, Median Age, and Mortality

The population sizes of both cancer and CHF patients increased in the status quo scenarios from

2016 to 2040: cancer patients increased from 10.8 million to 19.0 million (a 75.4% increase), and CHF patients increased from 5.7 million to 12.8 million (a 124% increase). The median ages of both populations rose as well: the median age of cancer patients increased from 75.6 to 78.1, and the median age of CHF patients increased from 78.2 to 80.0. These changes were driven by reductions in mortality which grew comparatively larger in older age ranges. In cancer patients ages 65‐74, mortality decreased from 6.68% to 6.21%; ages 75‐84, 13.6% to 12.9%; and ages 85+, 30.1% to 27.5%. In CHF patients ages

354

65‐74, mortality decreased from 9.91% to 8.87%; ages 75‐84, 16.9% to 15.9%; and ages 85+, 34.1% to

30.9%.

Mortality was lower in all three intervention scenarios as compared with status quo scenarios

(PCT in cancer patients, 0.96% lower [0.89%, 1.0%]; PME in cancer patients, 1.5% lower [0%, 1.5%]; CM in CHF patients, 2.4% lower [1.2%, 2.4%]). These reductions in mortality drove modest increases in population size in each of the three intervention scenarios over the status quo scenarios.

Appendix Table 1 presents the FEM regression model for mortality. The ORs of the covariates in this model suggest the mechanisms by which each intervention scenario reduces mortality as compared with status quo scenarios. Pain, depression, and ADL dependencies feed directly into the mortality model, as does smoking status. According to the model, reductions in depression, ADL dependencies, and current smoking reduce mortality; reductions in pain from moderate to mild reduce mortality, while reductions in pain from severe to moderate slightly increase mortality. Reductions in pain and depression also reduce the likelihood of current smoking (data not shown).

All three interventions reduced average pain, the prevalence of depression, and average ADL dependencies, and thus all three intervention scenarios reduced mortality. Of these outcomes, reductions in ADL dependencies have the largest effect on mortality, which is consistent with the mortality reduction observed in the CM scenario (which led to large, direct reductions in ADL dependencies). Low and high effect size scenarios in the PME intervention reduced severe pain to moderate pain, which nullified and tempered their reductions in mortality, respectively. In addition, reductions in pain and depression reduced the likelihood of smoking, which contributed to the reductions in mortality in the PCT and PME intervention scenarios.

Health Care Costs

In the status quo scenarios in both cancer and CHF patients, median per capita societal and OOP health care costs increased monotonically between 2016 and 2040 (Figure 1 and Appendix Figure 1).

355

They followed the same patterns in all three intervention scenarios, but in each of these scenarios, they were lower in each time step in all effect size scenarios. The PCT and CM scenarios were particularly effective at lowering both measures of costs. The PCT intervention would have to cost on average

$1,202 per patient ($1,127, $1,274) to nullify all of its societal and OOP health care costs savings; the

PME intervention, $201 ($91, $268); and the CM intervention, $993 ($853, $993).

These per capita reductions drove large cumulative reductions in both measures of costs in the

PCT and CM scenarios. As compared with the status quo scenarios, between 2016 and 2040, the PCT scenario saved $126 billion in societal health care costs ($118 billion, $134 billion) and $15.0 billion in

OOP health care costs ($14.0 billion, $15.9 billion), and the CM scenario saved $35.9 billion in societal health care costs ($35.9 billion, $52.3 billion) and $6.13 billion in OOP health care costs ($6.13 billion,

$6.36 billion). By contrast, the pain management intervention scenario added $26.2 billion in societal health care costs (saved $19.9 billion, added $41.6 billion) and $9.74 billion in OOP health care costs

(saved $4.22 billion, added $6.36 billion).

The differences in the direction of cumulative cost changes (addition vs. savings) between the

PME scenario and the other two scenarios were reflective of population size changes and not of the cost savings of the interventions themselves. All three intervention scenarios reduced both measures of costs on a per capita basis. However, these scenarios also increased population size as compared with the status quo scenarios, which increased the number of FEM participants contributing to cumulative costs.

In the PCT and CM scenarios, the rates of per capita cost savings outpaced the increase in population size resulting from the interventions; in the PME scenario, the opposite was true.

Appendix Table 2 presents the FEM regression models for total health care costs and OOP health care costs. The coefficients of the covariates in these models suggest the mechanisms by which each intervention scenario reduces both measures of costs as compared with status quo scenarios.

Depression and ADL dependencies feed directly into the cost models, as do age brackets, mortality,

356

living in a nursing home, and the interaction of claiming SSDI with living in a nursing home. According to both models, reductions in depression, ADL dependencies, and living in a nursing home reduce costs, while reductions in living in a nursing home while claiming SSDI increases costs. Increasing median age increases costs, while reducing mortality reduces costs. In addition, reducing pain, depression, and ADL dependencies reduce the likelihood of claiming SSDI, and reducing ADL dependencies reduces the likelihood of living in a nursing home (data not shown).

Together, these model effects combined to reduce per capita societal and OOP health care costs in all three intervention scenarios as compared with status quo scenarios. All three interventions reduced ADL dependencies, which directly reduced both cost measures and also reduced nursing home use, which also reduced costs (living in a nursing home has the largest effect of any covariate in each model). All three interventions also reduced pain and depression, which reduced the likelihood of claiming SSDI. This tempered the reduction in nursing home use. Reductions in depression also directly lowered costs.

All three interventions increased median age and reduced mortality. The former tempered cost reductions while the latter contributed to them.

QALYs

In the status quo scenarios in both cancer and CHF patients, median per capita QALYs decreased monotonically between 2016 and 2040 (Figure 2 and Appendix Figure 2). They followed the same patterns in all three intervention scenarios, but in each of these scenarios, they were higher in each time step in all effect size scenarios.

357

Discussion

I modeled the impact of implementing three palliative health services interventions (PCT for cancer patients, PME for cancer patients, and CM for CHF patients) on health care costs, quality of life, and population size and median age from 2016 to 2040. Each intervention directly reduced pain, depression, or ADL dependencies, leading to large reductions in the outcomes directly affected. I found that all three interventions reduced the two symptom and/or functional status outcomes not directly affected; reduced mortality, increased population size, and increased the median age of the population; lowered per capita societal and OOP health care costs; and increased QALYs.e Th PCT and CM interventions led to large cumulative reductions in both cost measures. The PME intervention increased cumulative costs, but this reflected its effect on population size and not on the cost savings from the intervention itself.

These results suggest that the future of palliative care could be a promising one: the vast majority of all the outcomes I analyzed in all intervention scenarios and sensitivity analyses showed improvements. From a health systems perspective, the interventions I modeled would not necessarily be difficult to implement: PCT interventions have been studied for years,23 and PME interventions have an equally long history,24 are often short‐term, and require only training of existing personnel to implement. This suggests to policymakers that offering alternative formulations of palliative care may require less effort than expected but still yield large positive individual and societal benefits.

That said, these are the results of a simulation and can only be as useful as the simulation is valid. With regard to model inputs, I used an evidence‐based approach to parameter estimation in the form of a systematic review. However, there is a relative paucity of RCTs of palliative health services interventions as compared with RCTs of other kinds of palliative and curative end‐of‐life interventions

(e.g., chemotherapy), and we did not consider non‐RCT study designs in our systematic review. I could only find three RCTs each for the PME and CM scenario estimates, and only one for the PCT scenario

358

estimates, which could limit the validity of my model inputs. In addition, our systematic review found a variety of areas in which palliative health services interventions are effective (e.g., communication outcomes), but I could not model most interventions because many outcomes relevant to palliative care are not implemented in the FEM. As such, I may not have been able to focus on the most effective or most policy‐relevant interventions.

The covariates in FEM regression models provide the mechanisms through which each intervention affects the outcomes I analyzed. The validity of my projections is thus a function of these covariates. Some of the mortality and cost model covariates I analyzed suggest mechanisms that are plausible and fully formed: in the models, for example, reducing ADL dependencies reduces the likelihoods of living in a nursing home and of claiming SSDI, which combine to reduce costs overall; and reducing pain and depression reduces the likelihood of smoking, which in turn reduces mortality. Other pathways, however, are undeveloped: in the models, pain, depression, and ADL dependencies directly affect mortality, and the latter two directly affect costs. In reality, all of these effects would likely be mediated through reductions in the use of health services, but the models do not capture many relevant services (e.g., hospital use, outpatient visits, etc.). My results thus rely in part on broad and potentially unsubstantiated assumptions about the impact of symptoms and functional status on health care utilization.

Even with these caveats, my projections can still provide useful information. Societal and personal health care costs at the end of life are substantial. Health care spending is projected to rise significantly in the coming decades: federal spending on the major health programs, which include

Medicare and Medicaid, represented 4.8% of GDP in 2014 and is projected to reach 8% of GDP by

2039.25 Medicare alone accounts for more than half of all federal health spending, and 30 percent of lifetime Medicare expenditures are spent in the last year of life.6 Of all patients in the last five years of life, 25% and 43% of have OOP spending that exceeds household and non‐housing assets, respectively,26

359

and over 60% of personal bankruptcies are caused by medical bills.27 In light of rising health care costs and the high and cost of end‐of‐life care, any cost reductions resulting from palliative approaches would be beneficial. The PCT and CM interventions I examined collectively saved over $180 billion in societal and personal health care costs through 2040, which would be a much‐needed boon both to patients and to a health care system struggling to remain fiscally solvent.28 By way of comparison, $180 billion could fund the construction of 25,000 new elementary schools or pay the salaries of over 130,000 additional elementary school teachers through 2040.29,30

All three interventions I analyzed reduced mortality and improved QALYs, indicating that patients who receive these interventions could live longer in better health. The original study of the PCT intervention demonstrated a direct increase in survival as a result of the intervention,9 but it is one of the few studies of palliative care to do so. My results demonstrate that there could be survival and quality of life benefits to palliative interventions that may not be observable within the time and scope of a single intervention study. The impact of these quality of life benefits is bolstered by evidence suggesting that QALYs gained at the end of life have a higher social value than would be expected due to the proximity to death.31

Pain, depression, and ADL dependencies are common in end‐of‐life patients: 64% of patients with advanced cancer have pain,32 15% of patients with advanced illness have depression33, and 56% of patients in the last month of life have at least one ADL dependency.34 Pain and depression significantly reduce quality of life in end‐of‐life patients,35,36 and patients and families report that they fear and wish to avoid pain near the end of life.37 ADL dependencies can have a significant impact on patient function and thus quality of life: patients who move from zero to one ADL dependency often require daily help, while those who move from two to three often require full‐time care.38 Given the high prevalence of and significant burden caused by these outcomes, the reductions I observed in these outcomes across all

360

three interventions throughout the simulations could have an important impact on the quality of life of future end‐of‐life populations.

I faced several limitations. First, I assumed that each of the three interventions directly affected only one outcome; in reality, these interventions, particularly PCT and CM, would be expected to affect a range of patient outcomes. Accounting for all effects of these interventions could actually improve my results, however, as palliative interventions rarely worsen any component of quality of life. Second, I were unable to implement the exact average, high, and low effect size scenarios for both the PME and

CM interventions. That said, the effect sizes I implemented were close enough to the observed effect sizes in the original studies to provide worthwhile results in the context of a simulation.

Third, I observed reductions in per capita health care costs and mortality in all three intervention scenarios, but the mechanisms by which the FEM models support these reductions are not fully developed. Further analysis is needed to establish the specific pathways through which reducing pain, depression, and ADL dependencies can affect health care costs and mortality. Fourth, I did not include the costs of the interventions themselves in my simulations, as only one of the seven original studies provided estimates of intervention costs. Including these costs would not change any of the population or quality of life outcomes but would reduce or possibly reverse the cost savings of the interventions. Lastly, I assumed all eligible patients received each intervention, which is not likely to occur even with widespread implementation of the interventions. My results thus represent hypothetical best‐case scenarios of intervention uptake and can serve as a benchmark in understanding the potential benefits of palliative services interventions.

In summary, implementing three currently available palliative health services interventions could reduce health care costs, extend survival, and improve quality of life among end‐of‐life cancer and

CHF patients through 2040. Future work should focus on (1) more nuanced understandings of the direct and indirect impacts of these interventions on a broader array of outcomes and the mechanisms

361

through which these interventions affect their outcomes, and (2) simulations of other effective palliative health services interventions on these and other outcomes. That said, my results indicate that palliative health services can offer positive and often large benefits to patients and society, and policymakers and payers should actively consider implementing them today.

362

References

1. Goldman DP, Cutler D, Rowe JW, et al. Substantial health and economic returns from delayed aging may warrant a new focus for medical research. Health Aff (Millwood). 2013;32(10):1698‐ 1705. 2. Goldman DP, Lakdawalla DN, Malkin JD, Romley J, Philipson T. The benefits from giving makers of conventional 'small molecule' drugs longer exclusivity over clinical trial data. Health Aff (Millwood). 2011;30(1):84‐90. 3. Lakdawalla D, Eber MR, Forma FM, et al. Measuring the value of better diabetes management. Am J Manag Care. 2013;19 Spec No. 2:E11. 4. Eibner C, Goldman DP, Sullivan J, Garber AM. Three large‐scale changes to the Medicare program could curb its costs but also reduce enrollment. Health Aff (Millwood). 2013;32(5):891‐ 899. 5. Juster FT, Suzman R. An Overview of the Health and Retirement Study. J Hum Resour. 1995;30(Suppl):S7‐S56. 6. Hogan C, Lunney J, Gabel J, Lynn J. Medicare beneficiaries' costs of care in the last year. of life Health Aff. 2001;20(4):188‐195. 7. Lubitz JD, Riley GF. Trends in Medicare payments in the last year of life. N Engl J Med. 1993;328(15):1092‐1096. 8. The 2011 Annual Report of the Board of Trustees of the Federal Old‐Age and Survivors Insurance and Federal Disability Insurance Trust Funds. 2011; http://www.socialsecurity.gov/oact/TR/2011/tr2011.pdf. 9. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non‐ small‐cell lung cancer. N Engl J Med. 2010;363(8):733‐742. 10. Syrjala KL, Abrams JR, Polissar NL, et al. Patient training in cancer pain management using integrated print and video materials: a multisite randomized controlled trial. Pain. 2008;135(1‐ 2):175‐186. 11. Miaskowski C, Dodd M, West C, et al. Randomized clinical trial of the effectiveness of a self‐care intervention to improve cancer pain management. J Clin Oncol. 2004;22(9):1713‐1720. 12. Vallieres I, Aubin M, Blondeau L, Simard S, Giguere A. Effectiveness of a clinical intervention in improving pain control in outpatients with cancer treated by radiation therapy. Int J Radiat Oncol Biol Phys. 2006;66(1):234‐237. 13. Kalauokalani D, Franks P, Oliver JW, Meyers FJ, Kravitz RL. Can patient coaching reduce racial/ethnic disparities in cancer pain control? Secondary analysis of a randomized controlled trial. Pain Med. 2007;8(1):17‐24. 14. Miaskowski C, Dodd M, West C, et al. The use of a responder analysis to identify differences in patient outcomes following a self‐care intervention to improve cancer pain management. Pain. 2007;129(1‐2):55‐63. 15. Doughty RN, Wright SP, Pearl A, et al. Randomized, controlled trial of integrated heart failure management: The Auckland Heart Failure Management Study. Eur Heart J. 2002;23(2):149‐146. 16. Kasper EK, Gerstenblith G, Hefter G, et al. A randomized trial of the efficacy of multidisciplinary care in heart failure outpatients at high risk of hospital readmission. J Am Coll Cardiol. 2002;39(3):471‐480. 17. Harrison MB, Browne GB, Roberts J, Tugwell P, Gafni A, Graham ID. Quality of life of individuals with heart failure: a randomized trial of the effectiveness of two models of hospital‐to‐home transition. Med Care. 2002;40(4):271‐282.

363

18. Capomolla S, Febo O, Ceresa M, et al. Cost/utility ratio in chronic heart failure: comparison between heart failure management program delivered by day‐hospital and usual care. J Am Coll Cardiol. 2002;40(7):1259‐1266. 19. Williamson A, Hoggart B. Pain: a review of three commonly used pain rating scales. J Clin Nurs. 2005;14(7):798‐804. 20. Serlin RC, Mendoza TR, Nakamura Y, Edwards KR, Cleeland CS. When is cancer pain mild, moderate or severe? Grading pain severity by its interference with function. Pain. 1995;61(2):277‐284. 21. Steffick DE. Documentation of Affective Functioning Measures in the Health and Retirement Study. Ann Arbor, MI: University of Michigan;2000. 22. Rabin R, de Charro F. EQ‐5D: a measure of health status from the EuroQol Group. Ann Med. 2001;33(5):337‐343. 23. Jordhoy MS, Fayers P, Loge JH, Ahlner‐Elmqvist M, Kaasa S. Quality of life in palliative cancer care: results from a cluster randomized trial. J Clin Oncol. 2001;19(18):3884‐3894. 24. Larson PJ, Dodd MJ, Aksamit I. A symptom‐management program for patients undergoing cancer treatment: the Pro‐Self Program. J Cancer Educ. 1998;13(4):248‐252. 25. The 2014 Long‐Term Budget Outlook. Washington, DC: Congressional Budget Office;2014. 26. Kelley AS, McGarry K, Fahle S, Marshall SM, Du Q, Skinner JS. Out‐of‐pocket spending in the last five years of life. J Gen Intern Med. Feb 2013;28(2):304‐309. 27. Himmelstein DU, Thorne D, Warren E, Woolhandler S. Medical bankruptcy in the United States, 2007: results of a national study. Am J Med. 2009;122(8):741‐746. 28. Potetz L. Medicare Spending and Financing: A Primer. Menlo Park, CA: The Henry J. Kaiser Family Foundation;2011. 29. Langley PC, Liedgens H. The impact of pain severity and frequency on HRQoL in the big 5 european union countries2010. 30. Lorenz KA, Shapiro MF, Asch SM, Bozzette SA, Hays RD. Associations of symptoms and health‐ related quality of life: findings from a national study of persons with HIV infection. Ann Intern Med. 2001;134(9 Pt 2):854‐860. 31. Pinto‐Prades JL, Sanchez‐Martinez FI, Corbacho B, Baker R. Valuing QALYs at the end of life. Soc Sci Med. 2014;113:5‐14. 32. van den Beuken‐van Everdingen MH, de Rijke JM, Kessels AG, Schouten HC, van Kleef M, Patijn J. Prevalence of pain in patients with cancer: a systematic review of the past 40 years. Ann Oncol. 2007;18(9):1437‐1449. 33. Hotopf M, Chidgey J, Addington‐Hall J, Ly KL. Depression in advanced disease: a systematic review Part 1. Prevalence and case finding. Palliat Med. 2002;16(2):81‐97. 34. Smith AK, Walter LC, Miao Y, Boscardin WJ, Covinsky KE. Disability during the last two years of life. JAMA Intern Med. 2013;173(16):1506‐1513. 35. Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research. Washington, D.C.: Institute of Medicine;2011. 36. Laird B, Mitchell J. The assessment and management of depression in the terminally ill. Eur J Palliat Care. 2005;12:101‐104. 37. Steinhauser KE, Christakis NA, Clipp EC, McNeilly M, McIntyre L, Tulsky JA. Factors considered important at the end of life by patients, family, physicians, and other care providers. Jama. 2000;284(19):2476‐2482. 38. Gaugler JE, Duval S, Anderson KA, Kane RL. Predicting nursing home admission in the U.S: a meta‐analysis. BMC Geriatr. 2007;7:13.

364

Tables and Figures

Table 8. Intervention Effects on Depression, Pain, and ADL Dependencies

Intervention and Patient Populationa Outcome PCT, cancer patients PME, cancer patients CM, CHF patients Prevalence of depression ‐75.3% (‐80.1%, ‐70.4%)b ‐5.11% (‐6.43%, ‐1.86%) ‐1.32% (‐1.32%, 0.25%) Average pain ‐3.45% (‐3.68%, ‐3.22%) ‐35.9% (‐45.5%, ‐13.6%)b ‐1.38% (‐1.38%, ‐0.31%) Average ADL dependencies ‐1.27% (‐1.36%, ‐1.17%) ‐5.53% (‐7.49%, ‐2.73%) ‐32.4% (‐35.1%, ‐18.3%)b a Percent changes are given as: mean change over status quo scenario in average effect size scenario (high effect size scenario, low effect size scenario) b By design, intervention had a direct effect on outcome in given patient population

365

Figure 1a. Projections of Per Capita Societal Costs in Status Quo and Intervention Scenarios, 2016‐2040

366

Figure 1b. Projections of Per Capita OOP Costs in Status Quo and Intervention Scenarios, 2016‐2040

367

Figure 2. Projections of Per Capita QALYs in Status Quo and Intervention Scenarios, 2016‐2040

368

Appendix Figure 1a. Projections of Percent Change in Per Capita Societal Costs in Intervention vs. Status Quo Scenarios, 2016‐2040

369

Appendix Figure 1b. Projections of Percent Change in Per Capita OOP Costs in Intervention vs. Status Quo Scenarios, 2016‐2040

370

Appendix Figure 2. Projections of Percent Change in Per Capita QALYs in Intervention vs. Status Quo Scenarios, 2016‐2040

371

Appendix Table 10. FEM Regression Model for Mortality

Covariate ORa Age Min(63, two‐year lag of age) 1.04** Min(10, Max(0, two‐year lag age ‐ 63)) 1.02** Max(0, two‐year lag age ‐ 73) 1.05** Male 1.26** Race/ethnicity Non‐Hispanic black 0.99 Non‐Hispanic black and male 1.09 Hispanic 0.89 Hispanic and male 1.16 Education Less than high school 1.05 Less than high school and male 0.93 Some college and above 0.97 Marital Status Two‐year lag of widowed 1.19** Widowed at age 50 1.01 Single at age 50 1.20** Conditions Cancer Two‐year lag of cancer 1.51** Cancer status at age 50 (imputed) 0.85* Heart disease Two‐year lag of heart disease 1.22** Two‐year lag of ever had CHF 1.28 Heart disease status at age 50 (imputed) 1.07 Hypertension Two‐year lag of hypertension 1.13** Hypertension status at age 50 (imputed) 1.10 Lung disease Two‐year lag of lung disease 1.39** Lung disease status at age 50 (imputed) 1.10 Diabetes Two‐year lag of diabetes 1.21** Diabetes status at age 50 (imputed) 1.09 Stroke Two‐year lag of stroke 1.09** Stroke status at age 50 (imputed) 0.82

372

Smoking Two‐year lag of current smoking 1.15** Ever smoked 1.04 Smoking status at age 50 (imputed) 1.23** Symptoms Two‐year lag of depressive symptoms 1.08 Two‐year lag of mild pain most of the time 0.87 Two‐year lag of moderate pain most of the time 0.94 Two‐year lag of severe pain most of the time 0.91 Functional Status Two‐year lag of has exactly 1 ADL dependency 1.23** Two‐year lag of has exactly 2 ADL dependencies 1.40** Two‐year lag of has 3 or more ADL dependencies 1.83** Two‐year lag of has exactly 1 IADL dependency 1.20** Two‐year lag of has 2 or more IADL dependencies 1.82** a * = p < 0.05; ** = p < 0.01

373

Appendix Table 11. FEM Regression Models for Costs

Coefficienta Covariate Total Costs OOP Costs Age 65 to 69 $1,113 $297 70 to 74 $1,846 $262 75 to 79 $4,257** $885** 80 to 84 $4,228** $805** > 84 $4,942** $1,340** Male ‐$418 ‐$365** Race/ethnicity Black $1,206 ‐$676** Black and male $833 $279 Hispanic ‐$168 ‐$358* Hispanic and male $530 $25 Education Less than high school ‐$855* ‐$805** Less than high school and male $80 $526** Some college and above $392 $527** Marital status Widowed ‐$261 ‐$125 Single ‐$316 ‐$272** Conditions Cancer Diagnosis stage $15,802** $1,868** Maintenance stage $2,218** $150 Terminal stage $3,822 $249 Heart disease Diagnosis stage $8,964** $1,433** Maintenance stage $3,026** $318** Terminal stage $2,492 ‐$136 Hypertension Diagnosis stage $4,934** $1,311** Maintenance stage $1,064** $252* Terminal stage $2,007 ‐$483 Lung disease Diagnosis stage $5,554** $471 Maintenance stage $2,595** $154 Terminal stage ‐$2,824 $344 Diabetes

374

Diagnosis stage $3,957** $622* Maintenance stage $2,279** $124 Terminal stage $5,168 $203 Stroke Diagnosis stage $10,588** $1,597* Maintenance stage $929 ‐$119 Terminal stage ‐$921 ‐$679 Heart disease and diabetes $290 ‐$36 Heart disease and hip fracture $304 ‐$303 Stroke and hip fracture $1,529 $671 Diabetes and hip fracture $1,203 $175 Disability Living in nursing home while receiving SSDI ‐$8,556* ‐$7,619** Died while receiving SSDI $21,539** $384 Eligible for Medicare due to disability $3,181** $426 Living in nursing home $42,742** $15,681** 3+ ADL dependencies and not living in nursing home $7,323** $947** Depressive symptoms $3,710** $495** Death Died ‐$365 $813 Died, ages 65‐69 $21,777** $344 Died, ages 70‐74 $27,595** $603 Died, ages 75‐79 $13,866 ‐$65 Died, ages 80‐84 $16,657* $979 Died, ages 85+ $9,186 ‐$169 Died while living in nursing home ‐$27,170** ‐$7,480** a * = p < 0.05; ** = p < 0.01

375

Chapter 5. Conclusions

Summary

In this dissertation, I examined the past, present, and future of end‐of‐life care in order to shed light on the successes and shortcomings in the field. I analyzed a variety of patient and family quality of life and economic outcomes.

In Chapter 2, I used a nationally representative survey of community‐dwelling older adults to showed that the prevalence of many important and troubling symptoms, including pain, depression, and periodic confusion, became more common in the last year of life from 1998 to 2010, and that no symptom became less common. Palliative and hospice care expanded considerably during that time, but these results call into question the specifics of their organization and delivery.

In Chapter 3, I systematically reviewed the RCT evidence for palliative health services interventions to identify the populations appropriate for palliative care and the interventions that improve patient and family quality life, health care use, and health care costs. I showed that palliative interventions are effective overall for cancer, CHF, COPD, and dementia patients and their families, but that certain kinds of interventions are more or less effective for certain outcomes. There remain gaps in research that addresses various quality of life outcomes among specific patient populations, hand healt care use and cost outcomes were poorly studied overall. These results highlight the value of multifaceted approaches to palliative care targeted for specific patient and family populations.

In Chapter 4, I used a microsimulation to project the impact through 2040 of implementing three of the most effective interventions identified in Chapter 3: a PCT for cancer patients, PME for cancer patients, and CM for CHF patients. I showed that all three interventions reduced per capita societal and OOP costs; increased QALYs; reduced mortality and increased population size and median age; and reduced pain, depression, and ADL dependencies. The PCT and CM interventions also led to

376

large reductions in cumulative population costs. These results demonstrate the long‐term value to individuals and to society of implementing evidence‐based approaches to palliative care.

Implications

Together, these results reinforce the value of palliative care but also highlight the critical importance of implementing it in the right way. While palliative care was expanding in the 1990s and

2000s, it focused largely on inpatient settings, team‐based approaches, and physician contributions. This kind of growth was natural in a field in which physicians sit atop the medical hierarchy and the sickest patients are hospitalized. To be sure, the rapid expansion of palliative care during this time was unequivocally positive. The benefits of palliative care are not in question, and palliative approaches stand in opposition to the overly intensive, curative approaches that often lead to unnecessary suffering and high costs.

That said, this dissertation demonstrates that there is room for improvement. Today, many palliative care programs still emphasize the same elements as they did 20 years ago. These benefit patients and families in a variety of ways, but the evidence indicates that they most likely do not meet all of the complex needs of certain patient and family populations. Put simply, these populations may be slipping through the cracks, and a harder push in the same direction likely will not help them. What is needed, then, is new pushes in new directions. These new directions can be thought of in terms of what policymakers, payers, providers, and researchers can do today and what they should focus on for the future.

Improving Palliative Care Programs

Today, palliative care programs should diversify their offerings. Policymakers and payers must make efforts to incorporate additional evidence‐based elements into these programs, and they should focus on areas of low‐hanging fruit that have largely not been implemented on a health system scale.

Chapters 3 and 4 highlighted two services in particular (PME and CM) that exemplify these kinds of

377

elements: they are relatively well‐studied, effective, and require only retraining of existing personnel and/or hiring of a small number of new personnel to implement. Chapter 3 also highlighted a number of other currently available approaches that could yield significant benefits for end‐of‐life populations: education and psychosocial support, interventions delivered at home, and nurse‐ and social worker‐ based intervention, among others.

Palliative care programs that emphasize these elements may find that they solve two problems at once: improving the quality of life of their patient and family populations, and alleviating workforce shortages. Workforce shortage calculations suggest that between 6,000 and 18,000 additional palliative care physicians are needed to properly staff current palliative care programs.1 But not every end‐of‐life concern needs to be (or should be) addressed by physicians, who may be less effective than other palliative care personnel in addressing certain issues in certain patient and family populations. Using the available evidence, programs can redirect resources away from physicians where appropriate and toward nurses and social workers (who cost roughly 3‐4 times less than palliative care physicians2,3) in order to improve the value of the care they offer.

Research Priorities

Looking ahead, research and funding priorities over the next several decades will play an important role in the future of palliative care. More evidence is needed in a variety of areas: focusing on symptoms in noncancer populations; focusing in general on conditions other than cancer, CHF, COPD, and dementia (such asRD ES and ESLD); incorporating technology into palliative care; using data‐driven approaches to target patients for palliative care; reducing the use of non‐hospital intensive health services (such as emergency departments) in end‐of‐life populations; and identifying the components of palliative care that reduce health care spending and those that do not. Some of these areas have been largely neglected by the palliative care research community, while others need further research to fully answer remaining questions.

378

In addition, evidence is needed that bridges the gap between research and practice. Aside from the aforementioned research areas in need of attention, consistency in research methodology is a top priority. For example, functional status has the potential to be a high quality indicator of the need for palliative care: it is well‐known, well‐studied, and easy to assess, and it can be measured by validated scales. As such, it could be used in algorithms to identify appropriate palliative care populations.

However, it is often not measured in research studies, and when it is, it is not measured or reported consistently: there is a lack of regularity in which scales are used to measure it and how data collected from these scales is presented. This heterogeneity precludes synthesis of the evidence, and without that synthesis, no recommendations can be made for how to use functional status to improve the delivery of palliative care. This problem is not limited to functional status; it is pervasive in palliative care research, particularly with regard to the measurement and reporting of quality of life outcomes.

There is also a need to bridge the gaps among research modalities. This dissertation contains an example of the pitfalls of research silos: as shown in Chapter 3, palliative health services interventions are highly effective at improving communication outcomes, but the future impact of these interventions could not be modeled in Chapter 4 because communication outcomes are not collected by HRS and therefore not implemented in the FEM. This kind of disconnect is common in palliative care research and limits its impact: the field as a whole can only be as effective as the ability of its different branches to communicate with and complement one another. Researchers and research institutes should actively seek new collaborations that foster integrated approaches to data collection and analysis.

Funding Priorities

Funding priorities drive research priorities. Funders need to recognize these critical shortcomings in palliative care research and act to remedy them. They should do this by (1) offering grants specifically designed to address research gaps; (2) stipulating that certain outcomes be measured and reported in a manner that will be most useful to other researchers, policymakers, and practitioners;

379

and (3) incentivizing collaborations that bridge gaps between disparate research methodologies and data sources.

Policymakers must also increase funding for palliative care research in general to levels commensurate with the importance of high quality end‐of‐life care. In 2010, the NIH spent $31.2 billion on biomedical science but only $57 million on palliative and end‐of‐life research, which represented

0.2% of its research budget.4 The impact of these funding priorities is dramatic. Chapter 3 identified 124 total RCTs of palliative health services interventions published between 2001 and 2015; over the same time frame, the Journal of Clinical Oncology alone published over 10 times that number of RCTs of chemotherapy agents and regimens.5 Curative approaches play an important role in the management of any disease, and they should undoubtedly receive significant research funding. They should not, however, be funded to the detriment of palliative approaches, which address fundamental aspects of the end‐of‐life experience that curative approaches cannot.

Closing Thoughts

Palliative care has made great strides in the last two decades: the number of hospitals with traditional inpatient palliative care programs is reaching saturation,6 and organizations such as CAPC have established a strong advocacy presence in promoting traditional models of palliative care.7

Inpatient palliative programs have improved the lives and deaths of countless numbers of patients and their families.

The work of expanding and promoting palliative care programs is far from over. But it is time for policymakers to take a critical look at the successes and shortcomings of these programs, to augment them with evidence‐based approaches and alternative models, and to adequately fund the generation of new knowledge so that the end‐of‐life experience can continuously improve in the future. By making these constructive changes, policymakers can fulfill their responsibility to the public to ensure that the

380

end‐of‐life experience is as humane as possible for patients and families. This dissertation provides them with the beginnings of a roadmap for change.

381

References

1. Lupu D. Estimate of current hospice and palliative medicine physician workforce shortage. J Pain Symptom Manage. 2010;40(6):899‐911. 2. Average Salary for Skill: Palliative Care. http://www.payscale.com/research/US/Skill=Palliative_Care/Salary. Accessed May 2, 2015. 3. Social Workers in Hospice and Palliative Care: Occupational Profile. Washington, DC: National Association of Social Workers;2010. 4. Building Momentum: The Science of End‐of‐Life and Palliative Care. Bethesda, MD: National Institute of Nursing Research. 5. PubMed Search Results. http://www.ncbi.nlm.nih.gov/pubmed/?term=%22J+Clin+Oncol%22[Journal]+AND+%22Random ized+Controlled+Trial%22[Publication+Type]+AND+%28%22Antineoplastic+Agents%22[Mesh]+O R+%22Antineoplastic+Combined+Chemotherapy+Protocols%22[Mesh]%29+AND+2001%3A2015 [DP]. Accessed May 2, 2015. 6. Growth of Palliative Care in U.S. Hospitals: 2013 Snapshot. New York, NY: Center to Advance Palliative Care;2013. 7. Center to Advance Palliative Care. http://www.capc.org/. Accessed May 2, 2015.

382

This product is part of the Pardee RAND Graduate School (PRGS) dissertation series. PRGS dissertations are produced by graduate fellows of the Pardee RAND Graduate School, the world’s leading producer of Ph.D.s in policy analysis. The dissertation has been supervised; reviewed; and approved by the faculty committee composed of Karl Lorenz (Chair), Daniella Meeker, and Joan Teno.

PARDEE RAND GRADUATE SCHOOL www.rand.org

RGSD-355