Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives: a Policy Analysis

Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives: A Policy Analysis

December 1986

NTIS order #PB87-177465 Recommended Citation: U.S. Congress, Office of Technology Assessment, Nurse Practitioners, Physician Assis- tants, and Certified Nurse-Midwives: A Policy Analysis (Health Technology Case Study 37), OTA-HCS-37 (Washington, DC: U.S. Government Printing Office, December 1986).

Library of Congress Catalog Card Number 85-600596

For sale by the Superintendent of Documents U.S. Government Printing Office, Washington, DC 20402 Preface

Nurse Practitioners, Physician Assistants, and ● examples of types of technologies by physi- Certified Nurse-Midwives: A Policy Analysis is cal nature (drugs, devices, and procedures); Case Study 37 in OTA’s Health Technology Case ● examples of technologies in different stages Study Series. This case study has been prepared of development and diffusion (new, emerg- in response to a request by the Senate Commit- ing, and established); tee on Appropriations. ● examples from different areas of medicine (e.g., general medical practice, pediatrics, OTA case studies are designed to fulfill two radiology, and surgery); functions. The primary purpose is to provide ● examples addressing medical problems that OTA with specific information that can be used are important because of their high frequen- in forming general conclusions regarding broader cy or significant impacts (e. g., cost); policy issues. The first 19 cases in the Health Tech- ● examples of technologies with associated high nology Case Study Series, for example, were con- costs either because of high volume (for low- ducted in conjunction with OTA’s overall project cost technologies) or high individual costs; on The Implications of Cost-Effectiveness Anal- ● examples that could provide information ma- ysis of Medical Technology. By examining the 19 terial relating to the broader policy and meth- cases as a group and looking for common prob- odological issues being examined in the lems or strengths in the techniques of cost-effec- particular overall project; and tiveness or cost-benefit analysis, OTA was able ● examples with sufficient scientific literature. to better analyze the potential contribution that those techniques might make to the management Case studies are either prepared by OTA staff, of medical technology and health care costs and commissioned by OTA and performed under con- quality. tract by experts (generally in academia), or written by OTA staff on the basis of contractors’ The second function of the case studies is to papers. provide useful information on the specific tech- OTA subjects each case study to an extensive nologies covered. The design and the funding lev- review process. Initial drafts of cases are reviewed els of most of the case studies are such that they by OTA staff and by members of the advisory should be read primarily in the context of the as- panel to the associated project. For commissioned sociated overall OTA projects. Nevertheless, in cases, comments are provided to authors, along many instances, the case studies do represent ex- with OTA’s suggestions for revisions. Subsequent tensive reviews of the literature on the efficacy, drafts are sent by OTA to numerous experts for safety, and costs of the specific technologies and review and comment. Each case is seen by at least as such can stand on their own as a useful contri- 30 reviewers, and sometimes by 80 or more out- bution to the field. side reviewers. These individuals may be from relevant Government agencies, professional so- Case studies are prepared in some instances be- cieties, consumer and public interest groups, med- cause they have been specifically requested by ical practice, and academic medicine. Academi- congressional committees and in others because cians such as economists, sociologists, decision they have been selected through an extensive re- analysts, biologists, and so forth, as appropriate, view process involving OTA staff and consulta- also review the cases. tions with the congressional staffs, advisory panel to the associated overall project, the Health Pro- Although cases are not statements of official gram Advisory Committee, and other experts in OTA position, the review process is designed to various fields. Selection criteria were developed satisfy OTA’s concern with each case study’s to ensure that case studies provide the following: scientific quality and objectivity. During the various stages of the review and revision process, . examples of types of technologies by func- therefore, OTA encourages, and to the extent pos- tion (preventive, diagnostic, therapeutic, and sible requires, authors to present balanced infor- rehabilitative); mation and recognize divergent points of view.

Ill.,, Health Technology Case Study Seriesa Case Study Case study title; author(s); Case Study Case study title; author(s); Series No. OTA Publication numberb Series No. OTA publication numberb 1 Formal Analysis, Policy Formulation, and End-Stage Renal 19 Assessment of Four Common X-Ray Procedures; Disease; Judith L. Wagner (OTA-BP-H-9(19))e Richard A. Rettig (OTA-BP-H-9(1)) C 20 Mandatory Passive Restraint Systems in Automobiles: Issues 2 The Feasibility of Economic Evaluation of Diagnostic Pro- and Evidence; cedures: The Case of CT Scanning; Kenneth E. Warner (OTA-BP-H-15(20))f Judith L. Wagner (OTA-BP-H-9(2)) 21 Selected Telecommunications Devices for Hearing-Impaired 3 Screening for Colon Cancer: A Technology Assessment; Persons; David M. Eddy (OTA-BP-H-9(3)) Virginia W. Stern and Martha Ross Redden 4 Cost Effectiveness of Automated Multichannel Chemistry (OTA-BP-H-16(21)) g Analyzers; 22 The Effectiveness and Costs of Alcoholism Treatment; Milton C. Weinstein and Laurie A. Pearlman Leonard Saxe, Denise Dougherty, Katharine Esty, (OTA-BP-H-9(4)) and Michelle Fine (OTA-HCS-22) 5 Periodontal Disease: Assessing the Effectiveness and Costs of 23 The Safety, Efficacy, and Cost Effectiveness of Therapeutic the Keyes Technique; Apheresis; Richard M. Scheffler and Sheldon Rovin John C. Langenbrunner (Office of Technology Assessment) (OTA-BP-H-9(5)) (OTA-HCS-23) 6 The Cost Effectiveness of Bone Marrow Transplant Therapy 24 Variation in Length of Hospital Stay: Their Relationship to and Its Policy Implications; Health Outcomes; Stuart O. Schweitzer and C. C. Scalzi (OTA-BP-H-9(6)) Mark R. Chassin (OTA-HCS-24) 7 Allocating Costs and Benefits in Disease Prevention Programs: 25 Technology and Learning Disabilities; An Application to Cervical Cancer Screening; Candis Cousins and Leonard Duhl (OTA-HCS-25) Bryan R. Luce (Office of Technology Assessment) 26 Assistive Devices for Severe Speech Impairments; (OTA-BP-H-9(7)) Judith Randal (Office of Technology Assessment) 8 The Cost Effectiveness of Upper Gastrointestinal Endoscopy; (OTA-HCS-26) Jonathan A. Showstack and Steven A. Schroeder 27 Nuclear Magnetic Resonance Imaging Technology: A Clinical, (OTA-BP-H-9(8)) Industrial, and Policy Analysis; 9 The Artificial Heart: Cost, Risks, and Benefits; Earl P. Steinberg and Alan Cohen (OTA-HCS-27) Deborah P. Lubeck and John P. Bunker 28 Intensive Care Units (ICUs): Clinical Outcomes, Costs, and (OTA-BP-H-9(9)) Decisionmaking; 10 The Costs and Effectiveness of Neonatal Intensive Care; Robert A. Berenson (OTA-HCS-28) Peter Budetti, Peggy McManus, Nancy Barrand, and 29 The Boston Elbow; Lu Ann Heinen (OTA-BP-H-9(1O)) Sandra J. Tanenbaum (OTA-HCS-29) 11 Benefit and Cost Analysis of Medical Interventions: The Case 30 The Market for Wheelchairs: Innovations and Federal Policy; of Cimetidine and Peptic Ulcer Disease; Donald S. Shepard and Sarita L. Karen (OTA-HCS-30) Harvey V. Fineberg and Laurie A. Pearlman 31 The Contact Lens Industry: Structure, Competition, and Public (OTA-BP-H-9(11)) Policy; 12 Assessing Selected Respiratory Therapy Modalities: Trends and Leonard G. Schifrin and William J. Rich (OTA-HCS-31) Relative Costs in the Washington, D.C. Area; 32 The Hemodialysis Equipment and Disposable Industry; Richard M. Scheffler and Morgan Delaney Anthony A. Romeo (OTA-HCS-32) (OTA-BP-H-9(12)) 33 Technologies for Managing Urinary Incontinence; 13 Cardiac Radionuclide Imaging and Cost Effectiveness; Joseph Ouslander, Robert Kane, Shira Vollmer, and Melvyn William B. Stason and Eric Fortess (OTA-BP-H-9(13)) Menezes (OTA-HCS-33) 14 Cost Benefit/Cost Effectiveness of Medical Technologies: A 34 The Cost Effectiveness of Digital Subtraction Angiography in Case Study of Orthopedic Joint Implants; the Diagnosis of Cerebrovascular Disease; Judith D. Bentkover and Philip G. Drew (OTA-BP-H-9(14)) Matthew Menken, Gordon H. DeFriese, Thomas R. Oliver, 15 Elective Hysterectomy: Costs, Risks, and Benefits; and Irwin Litt (OTA-HCS-34) Carol Korenbrot, Ann B. Flood, Michael Higgins, 35 The Effectiveness and Costs of Continuous Ambulatory Noralou Roos, and John P. Bunker (OTA-BP-H-9(15)) Peritoneal Dialysis (CAPD) 16 The Costs and Effectiveness of Nurse Practitioners; William B. Stason and Benjamin A. Barnes (OTA-HCS-35) Lauren LeRoy and Sharon Solkowitz (OTA-BP-H-9(16)) 36 Effects of Federal Policies on Extracorporeal Shock Wave 17 Surgery for Breast Cancer; Lithotripsy Karen Schachter Weingrod and Duncan Neuhauser Elaine J. Power (Office of Technology Assessment) (O-I-A-BP-H-9(17)) (OTA-HCS-36) 18 The Efficacy and Cost Effectiveness of Psychotherapy; 37 Nurse Practitioners, Physician Assistants, and Certified Nurse- Leonard Saxe (Office of Technology Assessment) Midwives: A Policy Analysis; (OTA-BP-H-9(18)) d (O-I-A-I-EC-37) aAvailab]e for sale by the Superintendent of Documents, U.S. Government dBackground paper #3 to The Implications of Cost-Effectiveness Analysis of Printing Office, Washington, DC, 20402, and by the National Technical Medical Technology. Information Service, 5285 Port Royal Rd., Springfield, VA, 22161. Call egackground paper #S to The Implications of Cost-Effectiveness Analysis of OTA’S Publishing Office (224-8996) for availability and ordering infor- Medical Technology. mation. fgackground paper #l to OTA’S May 1982 report Technology and ~andi- borigina] publication numbers appear in parentheses. capped People. cThe first 17 Ca= jn the Series were 17 separately issued cases in Background ggackground paper #2 to Technology and Handicapped People. Paper #2: Case Studies of Medical Technologies, prepared in conjunction with OTA’S August 1980 report The Implications of Cost-Effectiveness Anal- ysis of Medical Technology.

iv OTA Project Staff for Case Study #37 Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives: A Policy Analysis

Roger C. Herdman, Assistant Director, OTA Health and Life Sciences Division

Clyde J. Behney, Health Program Manager

Gloria Ruby, Project Director

Steven Sisskind, Research Assistant” Virginia Cwalina, Administrative Assistant Diann G. Hohenthaner, F’. C. Specialist Carol A. Guntow, Secretary/Word Processor Specialist

Principal Contractor Edward G. Brooks, University of North Carolina, Chapel Hill, NC

Contractors Louis P. Garrison, The Project Hope Health Sciences Education Center, Millwood, VA Anne Meadows, Washington, DC (Editing)

*From July to October 1986. Advisory Panel— Nurse Practitioners, Physician Assistants, and Certified Nurse= Midwives: A Policy Analysis

Rosemary Stevens, Chair Department of History and the Sociology of Science University of Pennsylvania, Philadelphia, PA Walter H. Caulfield Patricia A. Prescott Kaiser Permanence School of Nursing Oakland, CA University of Maryland Baltimore, MD Philip D. Cleveland Family Medicine Spokane Judith Rooks Spokane, WA Consultant Portland, OR Lynn Etheredge Health Policy Consultant George M. Ryan Washington, DC Department of Obstetrics and Gynecology College of Medicine Willis Goldbeck University of Tennessee Washington Business Group on Health Memphis, TN Washington, DC Richard M. Scheffler Sandra Greene Health Policy and Administration Program Health Economics Research School of Public Health Blue Cross/Blue Shield of North Carolina University of California Durham, NC Berkeley, CA Hurdis Griffith Henry M. Seidel Robert Wood Johnson Fellow School of Medicine Institute of Medicine The Johns Hopkins University National Academy of Sciences Baltimore, MD Washington, DC Gerry Shea Charles G. Huntington Health Care Division Hermon Medical Group Service Employees International Union Hermon, NY Washington, DC Lauren LeRoy Barbara Warden Physician Payment Review Commission National Consumers’ League Washington, ‘DC Washington, DC Kathy Lohr Ivan Williams The Rand Corp. Kellogg Center Washington, DC Montreal General Hospital Ruth Lubic Montreal, Quebec Maternity Center Association New York, NY Michael R. Pollard Office of Policy Analysis Pharmaceutical Manufacturers’ Association Washington, DC

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members. The panel does not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full responsibility for the report and the accuracy of its contents, vi Contents

Page CHAPTER 1: SUMMARY AND POLICY CONCLUSIONS...... 3 Introduction ...... 3 Background and Scope of the Case Study ...... 3 Organization of the Case Study ...... 4 Summary ...... 5 Contributions of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives ...... 5 Effects of Changing Payment Methods ...... 6 Policy Conclusions ...... 10 Addendum: Definitions and Descriptions ...... 12 CHAPTER 2: QUALITY OF CARE...... 17 Indicators of Quality ...... 17 Comparisons With Physicians ...... 17 Patients’ Satisfaction ...... 18 Physicians’ Acceptance ...... 18 Methodological Problems of Studies ...... 18 Quality of Nurse Practitioners’ Care ...... 19 Comparisons With Physicians ...... 19 Patients’ Satisfaction ...... 19 Physicians’ Acceptance ...... 21 Quality of Physician Assistants’ Care ...... 22 Comparisons With Physicians ...... 22 Patients’ Satisfaction ...... 22 Physicians’ Acceptance ...... 23 Quality of Certified Nurse-Midwives’ Care ...... 23 Comparisons With Physicians ...... 23 Patients’ Satisfaction ...... 24 Physicians’ Acceptance ...... 24 Summary ...... 25 CHAPTER 3: ACCESS T0 CARE ...... 29 Nurse Practitioners’ Contribution to Access to Care ...... 30 Physician Assistants’ Contribution to Access to Care ...... 32 Certified Nurse-Midwives’ Contribution to Access to Care ...... 33 Summary ...... 34 CHAPTER 4: PRODUCTIVITY, COSTS, AND EMPLOYMENT ...... 39 Scope of Professional Practice ...... 39 Services Provided by Nurse Practitioners and Physician Assistants ...... 39 Services Provided by Certified Nurse-Midwives ...... 40 Productivity ...... 40 Nurse Practitioners’ and Physician Assistants’ Productivity ...... 41 Certified Nurse-Midwives’ Productivity ...... 43 Costs and Employment...... 44 Costs and Benefits of Training Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives...... 44 Costs and Benefits of Private Employment of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives ...... 46

vii Contents—continued Page

Current Employment: Settings and Trends...... 47 Nurse Practitioners’ and Physician Assistants’ Employment ...... 47 Certified Nurse-Midwives’ Employment ...... 48 Summary ...... 49 CHAPTER 5: PAYMENT ISSUES ...... 53 Effects of Modifying Payment for Services of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives ...... 54 Effects on Independent Practices of Nurse Practitioners and Certified Nurse-Midwives...... 54 Effects on Physicians’ Practices ...... 57 Effects on Health Maintenance Organizations ...... 58 Effects on Hospitals ...... 58 Effects on Nursing Homes ...... 59 The Changing Context of Health Care ...... 60 Financing ...... 60 Supply of Physicians...... 60 Delivery Sites and Organizations ...... 61 Effects of Changes in the Health-Care Environment on Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives ...... 62 Summary ...... 64 APPENDIX A.–METHODS AND ACKNOWLEDGMENTS...... 69 APPENDIX B.–PAYMENT FOR THE SERVICES OF NURSE PRACTITIONERS, PHYSICIAN ASSISTANTS, AND CERTIFIED NURSE-MIDWIVES ...... 71 REFERENCES ...... 79

Tables Table No. Page 1-1. Coverage and Direct Payment for Services of Nurse Practitioners, Physician “ Assistants, and Certified Nurse-Midwives ...... 7 1-2. Comparison of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives ...... 13 2-1. Equivalence in Quality of Care Provided by Nurse Practitioners (NPs) and Physicians (MDs) ...... 20 2-2. Difference in Quality of Care Provided by Nurse Practitioners (NPs) and Physicians (MDs) ...... 21 2-3. Percentage of U.S. Resident Certified Nurse-Midwives by Type of Organization, 1976-77 and 1982 ...... , ...... 25 5-1. Selected Alternatives to Traditional Health-Care Delivery ...... 62 B-1. Coverage and Direct Payment for Services of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives ...... 72

Figure Figure No. Page 3-1. Distribution of Physician Assistants by Size of Community...... 32

Vlll. . . Chapter 1 Summary and Policy Conclusions Chapter 1 Summary and Policy Conclusions

INTRODUCTION

The use of nurse practitioners (NPs) and phy- physicians, restrictive State laws and regulations, sician assistants (PAs) to provide primary health and the inaccessibility and cost of malpractice in- care traditionally provided only by physicians de- surance. Although these problems are significant veloped during the 1960s in response to a per- (see box 1-A), they are beyond the scope of this ceived shortage and maldistribution of physicians. study, which focuses on another major barrier— Societal support for this innovation in the deliv- limited third-party payment for the services of ery of health-care was based on the potential for NPs, PAs, and CNMs. NPs and PAs to improve access and to lower costs while maintaining the quality of care. At about Background and Scope of the same time the number of certified nurse-mid- the Case Study wives (CNMs),1 who had been providing health care for some 30 years, began to increase substan- This case study was prepared in response to a tially. request by the Senate Committee on Appropria- In the past two decades, the ranks of NPs, PAs, tions to update a previous OTA case study, “The and CNMs and their responsibilities for provid- Cost and Effectiveness of Nurse Practitioners. ” ing care to patients have increased, despite the The committee also requested that OTA address resistance these practitioners have encountered in the extent to which various Federal health-care their attempts to assume more prominent or more programs and private third-party payers pay for independent roles in delivering health care. Today, the services of NPs and CNMs. Of particular in- approximately 15,400 NPs, 16,000 PAs, and 2,000 terest to the committee were the issues of cover- CNMs are practicing in the United States. age (i.e., authorization for payment) and direct payment (i.e., payment to NPs and CNMs) for Changes in the health-care environment have their services.2 The committee also requested that altered the forces that spurred the development OTA review the evidence on the quality and costs and growth of these groups of providers. The of the care NPs and CNMs provide. The analy- health-care sector has become increasingly com- sis also addresses PAs because their historical petitive as the supply of physicians has grown and background and current roles are similar to that as the proportion of physicians practicing in the of NPs, and because information on NPs often primary-care specialties has decreased. New forms overlaps with information on PAs. of organization for the delivery of medical care have emerged. Concern over the rapidly rising In considering NPs and PAs, the study focuses costs of health care has grown, and new meth- on the large majority who provide primary care, ods of paying for hospitals’ inpatient services have although some attention is given to the roles of been implemented. All of these changes have im- NPs and PAs in nonprimary-care settings. No dis- plications for the roles NPs, PAs, and CNMs will tinction is made between primary-care PAs and play in the future, and for the quality, accessibil- PAs trained in Medex programs specifically to ity, and costs of health care. provide primary care to underserved populations. As the health-care delivery system evolves, NPs, PAs, and CNMs are exploring ways to over- come several obstacles, such as unsupportive ‘The Medicare program and other third-part y payers distinguish ‘This case study uses the word certified to distinguish formally between coverage and payment. Coverage refers to benefits avail- trained and certified nurse-midwives from lay midwives, who may able to eligible beneficiaries or subscribers; payment refers to the or may not be nurses and who have informal training in midwifery. amounts and methods of payment for covered services.

3 4

The central questions the study attempts to an- . How would changing the payment method swer are: affect health-care costs for patients, third- party payers, and society? ● What contributions do NPs, PAs, and CNMs make in meeting the Nation’s health-care needs? Organization of the Case Study ● How would changing the method of payment for the services of NPs, PAs, and CNMs af- The case study is organized into five chapters fect the roles these practitioners would play and two appendixes. Chapter 1 presents a sum- in the evolving health-care delivery system? mary of the case study and in an addendum de- 5 fines and describes NPs, PAs, and CNMs. Chap- care sector could have on NPs, PAs, and CNMs; ters 2 through 4 discuss the contributions of NPs, and assesses how payment modifications in the PAs, and CNMs to health care. Chapter 2 ad- context of a rapidly changing health-care system dresses the quality of care, reviewing studies that might influence the roles of these practitioners and compare the care provided by NPs, PAs, and the costs of health care. CNMs with that provided by physicians and studies that gauge patients’ satisfaction with and physi- Appendix A describes the method of the study cians’ acceptance of the care provided by NPs, and acknowledges the assistance of the individ- PAs, and CNMs. Chapter 3 considers access to uals and organizations that reviewed this case health care; and chapter 4 focuses on productivity, study and provided valuable advice on its con- costs, and employment. Chapter 5 analyzes what tent. Appendix B presents a detailed description implications various payment modifications would of payment for the services of NPs, PAs, and have for the employment and practice of NPs, CNMs by third-party payers in the public and private sectors. PAs, and CNMs and for health-care costs; exam- ines the effects new developments in the health-

SUMMARY

Understanding how the use of NPs, PAs, and CNMs affects the quality of care, the access to care, the productivity of providers, and the costs of care is crucial for analyzing the effects of alternative policies regarding payment for the services of these providers. Drawing general conclusions is possible, despite the methodological limitations of many studies.

Contributions of NPs, PAs, and CNMs Direct measurement of the quality of the care provided by NPs, PAs, and CNMs is not possible at this time. Instead, the quality must be gauged by comparing their care with the care provided by physicians; by examining the extent to which patients are satisfied with the care provided by NPs, PAs, and CNMs; and by assessing physicians’ acceptance of such care. Many studies that analyze these relationships are methodologically flawed and almost none examine the quality of services provided without physician involvement. The weight of the evidence indicates that, within their areas of competence, NPs, PAs, and CNMs provide care whose quality is equivalent to that of care provided by physicians.3 More- over, NPs and CNMs are more adept than phy- Photo credit American College of Nurse-Midwives

3This study examined the quality of the care provided by NPs CNM’s improve quality of care and access to care by and PAs in primaW-care ambulatory settings and the quality of care providing person-oriented services such as health provided by CNMs in ambulatory and inpatient settings. education and counseling. 6 sicians at providing services that depend on com- access to primary care in other settings, such as, munication with patients and preventive actions. in the home and in correctional institutions, where The evidence indicates that PAs also perform bet- needed medical care is not always available. ter than many physicians in supportive-care and In principle, the scope of NPs’ and PAs’ prac- health-promotion activities. Patients are generally tice encompasses most of the primary-care serv- satisfied with the quality of care provided by NPs, ices provided by their physician counterparts. PAs, and CNMs, particularly with the interper- Productivity studies indicate that NPs and PAs sonal aspects of care. Although most physicians working under physicians’ supervision can in- who employ these practitioners are satisfied with crease total practice output by some 20 to 50 per- their performance, physicians’ willingness to del- cent. Increases in productivity resulting from the egate medical tasks is limited. Many physicians use of NPs and PAs vary widely depending on are more comfortable delegating the routine tasks the practice settings, on the responsibilities dele- related to primary care, such as taking histories, gated to these practitioners, on the severity and than the more technical procedures, such as phys- stability of the patients’ illnesses, and on how the ical examinations. Employment statistics also re- physicians choose to use the free time that results flect physicians’ acceptance of these practitioners. from delegating tasks. Although much less infor- Historically, NPs, PAs, and CNMs have been mation on productivity is available for CNMs credited with improving the geographic distribu- than for NPs and PAs, the degree to which CNMs tion of care, because many of them have been can substitute for physicians appears to be con- willing to locate in underserved rural and inner- siderable. city areas. As a result of increases in the supply Indirect evidence indicates these providers could of physicians, some physicians are beginning to decrease costs to employers and society. Employ- practice in smaller communities. Although some ment levels for NPs, PAs, and CNMs suggest that experts believe that the maldistribution of physi- health-care providers consider these practitioners cian manpower will improve over time, access to to be cost-effective substitutes for physicians in primary care is still limited and may persist as a delivering many services. From a societal stand- problem in certain geographic areas. How chang- point, training NPs, PAs, and CNMs costs much ing patterns in the distribution of primary-care less than training physicians. Given that the qual- physicians will affect the employment and the ity of care provided by NPs, PAs, and CNMs practice patterns of NPs, PAs, and CNMs is un- within their areas of competence is equivalent to certain, but these practitioners will continue to the quality of comparable services provided by play valuable roles in underserved areas. physicians; using NPs, PAs, and CNMs rather In addition to improving access to care in ru- than physicians to provide certain services would ral areas, NPs, PAs, and CNMs increase access appear to be cost-effective from a societal per- to primary care in a wide variety of nongeographic spective. settings and for populations not adequately served by physicians. Studies have shown, for example, Effects of Changing Payment Methods that NPs increase access to primary care for underserved children in school settings, and elderly Although the evidence indicates that NPs, PAs, patients in nursing homes. CNMs provide effec- and CNMs have made positive contributions to tive and low-cost maternity care to underserved, the delivery of health care, these practitioners socioeconomically high-risk pregnant women and have not been used to their fullest potential. Ma- adolescents. NPs, PAs, and CNMs have also im- jor obstacles to the greater employment and approved access by adding to the scope of primary- propriate use of NPs, PAs, and CNMs are that care services available to patients. NPs and PAs most third-party payers do not cover (authorize are competent in guiding individuals through to- for payment) the provision by NPs, PAs, and day’s complex health-care system and in caring CNMs of many services that are typically and for chronically ill adults and children. Preliminary characteristically provided by physicians, and, in reports indicate that NPs and PAs also increase those instances where third-party payers do cover 7 the services of NPs, PAs, and CNMs, the pay- sicians and in physicians’ practice arrangements. ments are most often indirect (i.e., to the employ- Innovations in methods of paying other providers ing physicians or institutions) rather than direct are multiplying. For example, some third-party (i.e., to the NPs or CNMs). PAs have not sought payers are paying prospectively for hospitals’ indirect payment. patient services (e.g., Medicare is paying on the 4 Most NPs, PAs, and CNMs are employed in or- basis of diagnosis related groups ), and cavitation’ is a growing mode of payment. These changes, ganized settings where employment is usually not along with the fact that an increasing proportion contingent upon coverage, However, the reluc- 65 tance of some physicians in private practice to hire of the population is aged or older, and thus in need of significant amounts of health-care serv- these practitioners stems partly from uncertainties ices, have major implications for the employment about payment for their services. NPs and CNMs and use of NPs, PAs, and CNMs and for health- in independent practices must depend on patients’ care costs. The uncertainty surrounding the mar- out-of-pocket payments. Some third-party payers kets for the services of NPs, PAs, and CNMs in in the public and private sectors cover the services of NPs, PAs, and CNMs (see table 1-1). Coverage a health-care system in a state of flux makes it difficult to predict the effect of payment changes. and direct payment has been mandated most often for CNMs, and to some extent they have been able to operate with suitable physician collaboration. 4Diagnosis related groups are groupings of diagnostic categories drawn from the International Classification of Diseases and modi- The effects of extending coverage for the serv- fied on the basis of surgical procedures, patients’ age, significant comorbidities or complications, and other relevant criteria. DRGs ices of NPs, PAs, and CNMs and paying directly are the case-mix measure mandated for Medicare’s prospective hos- for the services of NPs and CNMs would un- pital payment system by the Social Security Amendments of 1983 doubtedly be influenced by the markets for their (Public Law 98-21). ‘Cavitation payment is prospective payment of a per-capita amount services. The health-care system is currently un- for all services received by an enrollee or beneficiary during a given dergoing substantial changes in the supply of phy- period.

Table 1-1 .—Coverage and Direct Payment for Servicesa of Nurse Practitioners, Physician Assistants, and Certified Nurse. Midwives

Nurse practitioners Physician assistants Certified nurse-midwives Direct Direct Direct Third-party payer Coverage payment Coverage payment Coverage payment Medicare: Part A ...... No No No No No No Part B ...... No No Nob No No No C HMO S ...... , Yes NA Yes NA Yes NA State Medicaid programsd . . . . Some A few Some None Almost all Almost all programs programs programs programs programs Medicare and Medicaid: Rural Health Clinics. . . Yes No Yes No Yes No CHAMPUS e ...... Yes Yes No No Yes Yes FEHBP f ...... , 7 plans 7 plans 6 plans 6 plans 20 plans 20 plans Private insurance ...... In some In some No No In some In some States States States States NA = not available. aseryices that are typically and characteristically provided by physicians. bDurlng the publication of this case study, the omnibus Reconciliation Act of 1~ (public Law 99~9) was enacted. The act modifies part B of Medicare and authorizes payment for (covers) services of physician assistants working under the supervision of physicians in hospitals, skilled nursing facilities, intermediate-care facilities, and as an assistant at surgery The payment is indirect and at levels lower than physicians would receive for providing comparable services. cHealth maintenance organizations dstate Medicaid programs have the option of including Np and pA Services in their Siate Medicaid plans, Congress mandated coverage of CNMS’ SeWiCeS ifl 19S0, As of January 1985, all States in which CNMS practiced either were complying with the law (Public Law 96-499) or were considering changes in their Medical plans to comply with the law ecivilian Health and Medical Program of the Uniformed SerViCeS. fFederal Employees Health Benefit Program. FEHBP has 21 fee-for-service plans, some of which authorize PaYment to Nps, pAs, and CNMS gwhether State laws and regulatlofls require or pemlit lrlsurailce Coverage and direct payment for the services of NPs, PAs, and CNMS SOURCE Office of Technology Assessment, 1986. 8

The effect of modifying the payment system to ously affect costs. However, NPs and CNMs in cover and allow direct payment for the service administratively independent practices could po- of NPs, PAs, and CNMs depends on their em- tentially lower costs to third-party payers, pa- ployment setting. Such changes could spur the tients, and society. If the provision of services by growth of NPs’ and CNMs’ independent fee-for- NPs, CNMs, and physicians did not increase, ’ and service practices and joint practices with physi- if NPs’ and CNMs’ payment level were lower than cians, to the extent permitted under State laws and those of physicians for comparable services, lower regulations. Because CNMs are currently less lim- costs for third-party payers would be likely. If the ited than NPs by payment limitations of third- fees to patients reflected the lower payment levels, party payers, NPs would benefit most from cov- costs to patients’ and society could be lower. For erage and direct payment. primary care services, such as office visits, savings to patients would be small, because the fee Even with coverage and direct payment, the for the service is small, and because insurance usu- number of NPs and CNMs engaging in independent practice should be expected to remain ver ally covers most of the providers’ fees. Savings y for maternity care could be important, because small. In addition to the restriction imposed by the care itself is costly and insurance coverage is State laws and regulations, there are many diffi- culties in undertaking such a practice, including incomplete. Patients, third-party payers, and so- high startup costs, obtaining malpractice insur- ciety could have lower costs if the total costs of ance, and high premium malpractice insurance care provided by these practitioners was lower rates. NPs in independent practices also depend than the total costs of care provided by physicians on physician referrals to establish a clientele. Con- for similar medical conditions. cerns expressed by physicians and the current NPs and CNMs in independent practices would competitive market suggest that such referrals benefit by being able to offer lower prices as a might not be forthcoming, Independent practices competitive strategy. Individual practice associa- of CNMs are limited by physician concern with tion (IPA)-model health maintenance organiza- competition and difficulty in obtaining physician tions (HMOs), which contract with individual collaboration and hospital privileges. Although physicians for services, might turn to NPs as con- many patients might continue to prefer a physi- tractors for primary-care services and CNMs as cian, direct payment would give patients the choice contractors for maternity services. Preferred pro- of a wider range of providers. vider organizations (PPOs), which contract with One possible drawback of coverage and direct providers to supply services at discounted fees, payment is that additional covered providers might also consider NPs and CNMs as contrac- might increase the volume of services provided tors. These developments, however, would be and increase costs to patients and third-party limited by the increasing availability of primary- payers. Although the sparsity of conclusive data care physicians (including obstetricians) and other makes it difficult to allay this concern, the increas- barriers (see box 1-A). Moreover, physicians ap- ing emphasis most third-party payers place on pear to be engaging in price competition as a re- monitoring the use of services might help control sult of the changing health-care market. any increase in the volume of services provided. How coverage for NPs, PAs, and CNMs would Because of their potentially small number, NPs affect their employment and appropriate use by and CNMs in independent practice might not seri- fee-for-service physicians’ practices is uncertain, because many variables affect physicians’ deci- ‘Such practices would be administratively independent. Adminis- sions to employ these practitioners and to dele- tratively independent practices are not clinically independent from physicians when NPs and CNMs are performing delegated medical 7 tasks. In addition to the nursing profession’s agreement to clinical N0 direct evidence is available as to how coverage and direct collaboration with physicians, State laws and regulations that pro- payment would affect the volume of services provided by NPs and scribe the scope of practice of NPs and CNMs and specify require- CNMs. Although research on physicians’ influence on the volume ments for physician supervision serve as a more formal control on of services has been conducted for many years, none of the studies clinical independence. NPs and CNMs may legally be clinically in- unequivocally proves the magnitude or even the existence of phy- dependent from physicians when performing nursing tasks. sicians’ ability to control the volume of services (246). 9 gate tasks commensurate with the training of these physicians’ disinterest in visiting elderly residents providers. If NPs’, PAs’, and CNMs’ services were of nursing homes (166) been established, but there authorized for payment, some physicians might are very few physicians trained in geriatrics (126). be encouraged to employ and integrate these Furthermore the elderly institutionalized popula- providers into their practices, knowing that prac- tion is growing. Although more and better phy- tices that employ NPs and PAs are better able to sician care for these patients may be available in offer competitive prices and broader ranges of the future, their ability to furnish all the health services than are other practices (17). Some phy- needs of this group is questionable. The geriatric sicians might find it advantageous to hire new component of many of the training programs of physicians, rather than NPs, PAs, or CNMs, be- NPs and PAs has been increased and the 1- to 2- cause the rate at which physicians’ income is grow- year length of NP and PA training programs makes ing is decreasing, and new physicians are express- NPs and PAs readily available for providing care. ing interest in salaried positions and are willing NPs and PAs have the demonstrated ability to to work for less money than established physi- provide care for a population with chronic prob- cians earn. Employing physicians, rather than lems and functional disabilities. Coverage would NPs, PAs, or CNMs, might make some practices permit NPs and PAs8 to legally provide the pri- more competitive, because of the status patients mary care services for which they are trained and often confer on physicians. Physicians with declin- licensed—services that many nursing homes find ing patient bases might not be able to justify tak- difficult to supply. ing on additional providers and expenses and If coverage were extended, NPs and PAs would might compete by increasing the time spent with most likely provide nursing home visits as em- individual patients. ployees of physicians’ practices or as team mem- The advantages of extending coverage for NPs’, bers in group practices to provide nursing-home PAs’, and CNMs’ services in fee-for-service set- visits. If NPs were paid directly, they could func- tings is apparent in certain settings, for certain tion as independent practitioners, supplying pri- populations and where there are demonstrated mary-care services to nursing homes. Except when shortages of trained personnel. For example, rapid more intensive care can be substantiated, the growth in the elderly population and in the use Medicare program currently limits the frequency of nursing-home care has raised concerns about of physicians’ visits to nursing homes, so third- the quality and costs of such care. Not only has party payer costs in this setting might not be affected as long as payment levels were the same for NPs and PAs as for physicians. Total costs to third-party payers would probably decrease because visits to nursing homes by teams of physicians and NPs or PAs would decrease the use of hospital facilities (128,155,257 ).’

8During the publication of this case study, the Omnibus Recon- ciliation Act of 1986 (Public Law 99-509) was enacted. The act changes the Medicare law and authorizes the coverage of the serv- # ices furnished by PAs under the supervision of physicians in skilled nursing facilities and intermediate care facilities in States where PAs are legally authorized to perform the services. This provision takes effect Jan. 1, 1987. Payments, which go to the employer are 85 percent of the prevailing charges of physician services for comparable services provided by nonspecialist physicians. 9As app. B describes, a number of other Medicare and Medicaid regulations specific to nursing homes limit the roles of NPs and PAs and specify services that must be performed by physicians in order Extending coverage for NPs to provide primary care for the nursing homes’ services to be covered. In addition to per- services to elderly nursing home residents would mitting coverage under Medicare and Medicaid, amendments to these alleviate a demonstrated shortage of trained regulations would be required in order to encourage the employ- personnel for that population. ment and appropriate use of NPs’ and PAs’ services in this setting. 10

Coverage for the services of NPs and PAs could ready covered by public and private third-party also be advantageous for home-bound elderly pa- payers. Thus, coverage and direct payment for tients and for allowing pediatric NPs to care for the services of these practitioners would not chronically ill children at home. Medical teams directly affect their employment by HMOs. of pediatricians and PNPs—with the PNPs pro- Such employment might diminish, however, if viding routine care, teaching children at home, competition leads physicians to accept salaries and monitoring the program—have been shown that are sufficiently low to entice HMOs to em- to be effective in minimizing the social and psy- ploy physicians instead of NPs, PAs, or CNMs. (234). chological consequences of chronic illness Another factor that might negatively affect CNMs could be covered for the maternity care HMOs’ employment of these practitioners is the of pregnant disabled women, in cases where the increase in the number of IPA-model HMOs. Be- disabling condition did not complicate the preg- cause they are primarily organized around phy- nancy and birth process. Such women might ben- sicians who usually practice in private offices, efit from the individualized care that CNMs typi- IPA-model HMOs are less likely than are large cally provide. group- or staff-model HMOs to employ these pro- Coverage would be advantageous in rural areas viders. Although the number of IPA-model HMOs where the lack of medical personnel is a persist- has increased, the group- and staff-model HMOs ing problem. Although the Rural Health Clinics have the greatest number of enrollees. Services Act of 1977 extended coverage to NPs, The data suggest that NPs, PAs, and CNMs of- PAs, and CNMs working in rural clinics, not all fer financial savings to capitated HMOs. An in- residents of such areas have access to clinics. Cov- creasingly competitive environment might en- erage for NPs, PAs, and CNMs might encourage courage providers to pass on to consumers the their use by physicians in fee-for-service practices savings generated by the employment and appro- in rural areas who, because of fewer numbers, priate use of NPs, PAs, and CNMs, which would must see considerably more patients and work benefit society. longer hours than their urban counterparts. Fur- thermore, direct payment might encourage qual- Providing coverage or direct payment for the ified NPs and CNMs to move into unserved and services of NPs, PAs, and CNMs would not nec- underserved areas to expand access to heath care. essarily affect their employment by hospitals for inpatient care. NPs, PAs, and CNMs who work Competition among health-care organizations in hospitals are usually hospital employees, and and the growth of HMOs—which have employed the hospitals pay their salaries. Furthermore, there and used NPs, PAs, and CNMs extensively in the is no statutory permission or lack of permission past—augurs larger roles for these providers in under Medicare or Medicaid for payment of NPs’, the health-care system as employees of HMOs. PAs’, or CNM’s services as inpatient hospital serv- Cavitation, the method used to pay most HMOs, ices when these providers are employed by hos- does not require providers to bill for specific serv- pitals. Most other third-party payers are also si- ices, and the services provided by NPs, PAs, and lent on this issue. With coverage, these services CNMs in such settings are, for the most part, al- could be billed for as professional services.

POLICY CONCLUSIONS

NPs, PAs, and CNMs have made important ● increasing the productivity of medical prac- contributions to meeting the Nation’s health-care tices and institutions. needs by: These practitioners have been accepted in a wide . improving the quality and accessibility of range of settings under many different payment health-care services; and schemes, have the potential to reduce health-care costs, and clearly play legitimate roles in the ● those in certain locales (geographically under- health-care system. served rural and inner-city areas); ● Although NPs, PAs, and CNMs are not em- those in certain settings (e.g., homes and ployed and used to their fullest potential, many nursing homes); and ● third-party payers in the public and private sec- specific populations (e. g., some disabled preg- tors are gradually lowering the barriers presented nant women and some chronically ill patients, by current payment methods and coverage re- both adults and children). strictions. Covering the services of NPs, PAs, and CNMs Although Federal third-party payers vary con- might encourage physician fee-for-service prac- siderably in the extent of their coverage of and tices to employ these providers and use them in payment for the services of these providers, in settings and for populations that are not receiv- general, coverage and direct payment is limited ing sufficient and adequate care. Because payment (see app. B). Federal third-party payers could be would be to employing physicians, physicians more in step with new and evolving payment would have the final authority for the employ- practices by liberalizing coverage and payment ment and the exact nature of NPs’, PAs’, and restrictions for the services of NPs, PAs, and CNMs’ responsibilities. Physicians would have to CNMs. A major policy question is the manner of recognize the advantages of using NPs, PAs, and CNMs in their practices for providing care to un- liberalizing coverage and policy restrictions. Cov- served and underserved individuals. erage could be extended for NPs’, PAs’, and CNMs’ services in all settings or only in certain Direct payment as well as coverage for serv- settings. Direct payment for the services of NPs ices of NPs and CNMs might enable them to de- and CNMs would further remove barriers to prac- velop independent practices in competition with tice. (PAs have not sought direct payment. ) physician practices. Legal and financial restric- How extending coverage for the services of tions could be expected to keep the numbers of NPs, PAs, and CNMs in all settings would affect NPs and CNMs in independent practice very small. their employment and use varies on the setting: Competition from an increasing supply of physicians might offset the gains direct payment would • little change would occur in HMOs and in- bring to the independent practice of NPs and patient hospital settings; and CNMs. • the effect in physician fee-for-service practice settings is unclear. How adding these practitioners, particularly as independent practitioners, to the health-care sys- Coverage for the services of NPs, PAs, and tem, would affect costs cannot be resolved at this CNMs by additional payers would have little ef- time. The suspicion exists that total costs would fect on the employment and use of these providers increase, but data are not available to answer the by HMOs or by hospitals for inpatient care. While question. If costs increased due to an increase in important changes in employment opportunities the provision of services, volume controls could could occur in physician fee-for-service practices, be instituted. the direction of change is not clear because of the large number of variables that affect physicians’ If the overall volume of services did not in- decisions. Since the effect on costs is directly re- crease, and if the NPs’ and CNMs’ payment levels lated to the extent of employment, this question were lower than physicians’ levels for compara- also remains unanswered. ble services, third-party payers’ costs might decrease. Patients might realize savings from de- Extending coverage for NPs’, PAs’, and CNMs’ creases in the fees for some services. The extent services in all settings or limiting coverage for their of any savings would depend on what payment services to certain settings where health-care serv- levels were established. In any event, patients ices are currently inaccessible or inadequate would could choose from a wider range of providers and benefit certain individuals, such as: might have greater access to primary-care services. 12

Direct payment for the services of NPs and PAs It seems clear that coverage for the services of could be limited to certain settings where there NPs, PAs, and CNMs in at least some settings are demonstrated shortages of primary-or mater- could improve health care for segments of the nity care services. For example, direct payment population that are not being served adequately. might be provided to NPs and CNMs who in- How coverage would affect costs is unclear, but crease geographic access to care. NPs and CNMs the long-term result could be notable savings. The in independent practice may prove a viable solu- effect of direct payment on costs is even less cer- tion for meeting the health-care needs of sparsely tain, but it might enable NPs and CNMs to prac- populated areas that cannot support a physicians’ tice in unserved and underserved areas to expand practice. However, limiting direct payment to cer- access to health care. tain areas and populations may not be an efficient cost containment measure because of the potentially small number of independent practices.

ADDENDUM: DEFINITIONS AND DESCRIPTIONS

Descriptions of the general roles of NPs, PAs, within a health care system which provides for and CNMs indicate the similarities and differences medical consultation, collaborative manage- of these three types of health practitioners. (See ment, and referral. table 1-2 for a comparison of their general char- —American College of Nurse-Midwives, 1984 acteristics. ) PAs differ from NPs and CNMs in their working relations with physicians. PAs always work Today’s nurse, operating in an expanded role as a professional nurse practitioner, provides under physicians’ supervision, whereas NPs and direct patient care to individuals, families and CNMs work under physicians’ supervision, or in other groups in a variety of settings. . . . The collaborative relationships with physicians and nurse practitioner engages in independent deci- other health professionals. Another major differ- sionmaking about the nursing needs of clients, ence lies in the training these practitioners un- and collaborates with other health professionals, dergo. NPs and CNMs are licensed registered such as the physician, social worker, and nutri- nurses 10 who have received advanced training be- tionist in making decisions about other health yond that of other registered nurses. NPs are needs. The nurse working in an expanded role trained as generalists in the provision of primary practices in primary, acute, and chronic health care services. They may choose to specialize at care settings. As a member of the health care team, the nurse practitioner plans and institutes the graduate level and deal with specific popula- health care programs. tions, as do geriatric or pediatric NPs. CNMs re- –GEMNAC, 1979 ceive advanced training in midwifery. PAs, how- The purpose of the physician assistant in pri- ever, are not required to be registered nurses, and mary care is to help the physician provide per- the great majority are not. They come from a va- sonal health service to patients under his care. riety of backgrounds and experiences before train- An assistant works with a supervising physician ing to become PAs. Most PAs have had 3 or more in performing clinical functions and tasks which years of college-level education or several years prior to the mid-1960s were reserved principally if not solely for performance by the physician. ‘“Three types of nursing education lead to registered-nurse licenses: –Allied Health Education Directory, 1985 2-year community-college programs; 3-year hospital-affiliated diploma [Nurse-midwifery practice is] the independent programs; and 4-year baccalaureate-degree programs. The trend to make nursing education more academic and uniform is reflected in management of care of essentially normal new- the discontinuation of many hospitals’ diploma programs, although borns and women, antepartally, intrapartally, this has not resulted in an increased demand for baccalaureate edu- postpartally and/or gynecologically [and] occurs cation for nurses. 13

Table 1-2.—Comparison of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives

Nurse practitioners Physician assistants – Certified nurse-midwives Date of first educational program ,. ., ...... 1966 1965 1931 C Approximate number trained .25,000 to 30,000a 18.116 b 3,500 Approximate number employed in field of trainingd . 15,433e 16,000f 2,0009 Services ...... Provide medical services within Provide medical services as Provide full range of prenatal, limits of competence; provide assistants to physicians labor, delivery, and postpartum counseling and health-promotion care; family-planning counseling, services and gynecological services Role ...... Provide advanced nursing Provide medical care under Provide midwifery services in services, including working with supervision of physicians consultation with physicians, clients having complex or multiple mainly serve low-risk women; needs; provide medical services in increasingly work administratively collaboration with physicians and independent of physicians other health providers Settings ...... Mainly primary care; trend toward Mainly primary care; trend Hospitals, trend toward birthing hospitals, long-term care facilities, toward hospitals, long-term centers, health departments. and and other settings care facilities, and other family planning clinics Education Registered nurse with additional Special academic and on-the- Registered nurse with additional training, increasingly at masters job training training. about half at masters level level Approximate average Income $25,975h $27,560’ $25,000J i?IEStlrn.at@ bY Denl~e GelOt, Dlvl~lOn of Nur~l ng, B“ rea” of Health pr~fes~i~n~, Health Resources and services Admi nlstratlon, Publ IC Health seWl Cfc3 U S be~artm-en-t of Health and Human Services, Rockvllle, MD, personal communication, Aug 20, 1966 bAmerlcan Academy of Physician Assistants, “AAPA Membership Statlstlcs by Graduation Date,” Arllngton, VA, May 13. 1986 cE~tlmated by American Col}ege of F4urse-Mldwlves, Washington, DC, PerSOnal communication Aug 20, 1986 dThe figures for NPs and CNMs are from 198rJ Later data from the U S Department of Health and Human services, Publlc Health SeTVICe Health Resources and Serv Ices Adml nl strat Ion, Bureau of Health Professions, Divlston of Nursl rig, ” 1984 National Sample Survey of Registered Nurses, ” Rockwlle, MD Indicates that the ag gregate number of employed NPs and CNMS is 18642 ‘Registered N Urse population eN at ,Onal sample Sumey of Registered Nurses, November 1980, i n and Overview, ’ U S Department of Health and Human Serv{ces Pub IIC Health Service, Health Research Services Adminlstratlon, Publication No HRS.P-OD-83-I, November 1982 fEstlmated by Gretchen Shafft, American AcademY of physlclan Assistants, Arlington, VA, personal communlcatlon, Sept 15, 1986 gEsttmate from Kathy Mlchels Ass!stant Director, Congressional and Agency Relatlons, American Nurses’ Assoclatlon, Washington DC personal commun!catlon June 17, 1986 hu s Department of Health and Human SewIces publlc Health Service, Health Resources and SWVlCt2S Admlnlstratlon, Bureau Of Health professions Dlvlslon of Nurs- ing, 1984 National Sample Survey of Registered Nurses, ” Rockvllle, MD IAmerlcan Academy of Physlclan Assistants, 1984 Rrys/c/an Ass/sfarrt Masterf//e Survey (Arl!nglon, VA 1984) jAmerlcan college of Nurse Mldwlves, Washington DC, personal communication Aug 20, 1986 SOURCE Off Ice of Technology Assessment 1986 of experience in health-related fields, although certified by the National Commission of Certifi- these are not entrance requirements for the train- cation of PAs. ing programs. CNMs are trained to provide care for essentially Certification is available to all three types of normal expectant mothers and to handle abnor- health practitioners and is required for CNMs. mal cases by referring the patients to physicians Certification is offered to registered nurses by the or by consulting physicians or working jointly American Nurses Association, by nurse-specialty with them. Specific functions include providing associations and by some academic nursing-edu- prepartum care, managing normal deliveries, pro- cation programs. An NP can be certified after viding postpartum care, providing gynecological completing either an NP-master’s program or an care, providing care to normal newborns and in- NP-certificate program. Master’s degree programs fants, and providing family-planning services. require applicants to have baccalaureate degrees and registered-nurse licenses, and such programs NPs are taught to perform functions beyond entail an average of more than a year of additional those of traditional nursing and to assume respon- training. Certificate programs are generally a year sibility for some of the care usually provided by long and require registered-nurse licenses. CNMs physicians (see box I-B). PAs are also trained to are certified according to the requirements of the provide some of the services typically provided American College of Nurse-Midwives. PAs are by physicians (see box I-B). PAs are trained in 14

interpersonal skills, but not to the extent that NPs tinctions between NPs and PAs exist; in other set- and CNMs are. Indeed, counseling and health tings the two types of providers function very education are traditional dimensions of nursing differently. NPs, as registered nurses, perform the practice. Although many PAs pursue medical and full scope of nursing practice in addition to per- surgical subspecialties, this study focuses on those forming medical tasks, whereas PAs only perform PAs who are primary-care practitioners in am- medical tasks. In reality, NPs and PAs often per- bulatory settings. form the same roles, and evaluations often focus The roles PAs and NPs play depend on their on NPs and PAs collectively, rather than on ei- work settings. In some settings, no functional dis- ther NPs or PAs alone. Chapter 2 Quality of Care Chapter 2 Quality of Care

Because health care encompasses both techni- diagnostic and therapeutic components of care; cal care and the art of care (146), the quality of the art of care refers to the environment in which both must be assessed in determining the quality care is provided and the provider’s manner and of the care provided by nurse practitioners (NPs), behavior in caring for and communicating with physician assistants (PAs), and certified nurse- the patient (146). midwives (CNMs). Technical care comprises the

INDICATORS OF QUALITY

Current methods of evaluating the quality of physicians do not usually perform are unrea- care provided by NPs, PAs, and CNMs are inex- sonable. act. Structure, process, and outcome of care are Comparison studies are biased against NPs, traditionally used to measure the quality of care 1 PAs, and CNMs because the studies assume the provided by physicians (70). The quality of care medical model as the standard—physician care provided by NPs, PAs, and CNMs is often evalu- 2 is considered the standard for care. This stand- ated by comparing the process and outcome of ard may be appropriate for measuring the tech- the care they provide with the process and out- nical quality of the tasks that NPs, PAs, CNMs, come of the care physicians deliver. Other accepted and physicians perform. But the medical model indicators of the quality of care provided by NPs, may be less suitable for measuring the interper- PAs, and CNMs are patients’ satisfaction and, sonal quality or art of care, which is more char- to a lesser extent, physicians’ acceptance. acteristic of care provided by NPs, PAs, and CNMs than of that provided by physicians. In- Comparisons With Physicians deed, health promotion, teaching, and counsel- The quality of care provided by NPs, PAs, and ing are the essence of nursing education and are CNMs can be compared to the quality of care pro- also stressed in the curricula for training NPs and vided by physicians with regard to only those CNMs. PAs also receive training in interpersonal functions that both physicians and NPs, PAs, and skills, but to a lesser extent. Physicians can legally CNMs usually perform. Comparisons based on provide health education and counseling, but the functions outside the scope of NPs’, PAs’, and training in these skills varies among medical spe- CNMs’ training and practice, or on functions that cialties and medical schools. Among physicians, only family practitioners and psychiatrists receive I Structural measures evaluate descriptive characteristics of facil- extensive training in interpersonal skills, although ities and providers, e.g., the soundness of a building and the board some physicians in all specialties provide personal certification of a physician. Process measures evaluate what a pro- care. vider does to and for a patient, e.g., order a cardiogram for a patient with chest pain. Outcome measures evaluate the result of pa- Some other comparison studies are biased in tient care, i.e., health status. Although outcome measures are the most accurate available measure of quality, they are difficult to ob- favor of NPs, PAs, and CNMs. In studies where tain. (For a discussion of the problems associated with measuring patients are not randomly assigned, patients as- the outcome of care, see OTA’s 1986 report, Payment for F’hysi- signed to NPs, PAs, and CNMs are, on the whole, cian Services: Strategies for Medicare (246). ) ‘The structural measures applicable to NPs, PAs, and CNMs in- healthier than patients who see physicians exclu- clude their certification, and the accreditation of their training pro- sively; and either the practitioners or patients can grams and of their continuing education programs. decide to consult physicians at any time. Of those 3Although acceptance and satisfaction are not synonymous, the literature uses the words interchangeably in describing positive re- patients who consult physicians, those who choose sponses to NPs, PAs, and CNMs and the care they provide. to remain exclusively under the physicians’ care

17 18 most likely are less healthy than those who re- pear to influence malpractice cases: physicians turn to the NPs, PAs, or CNMs. who maintain good relations with their patients tend to be sued less frequently than physicians Patients’ Satisfaction who lack rapport with their patients (185). Looking to patients’ satisfaction as an indica- Physicians’ Acceptance tion of quality of care reflects an increasing sen- sitivity to patients’ interests and concerns and a Some authorities reject the notion that physi- recognition that outcomes partly depend on pa- cians’ acceptance of NPs, PAs, and CNMs indi- tients’ attitudes. Little evidence, however, suggests cates that the care they provide is good. Other that patients’ satisfaction positively correlates with authorities believe that physicians’ acceptance of favorable technical outcomes (70). Patients’ judg- such providers indicates good care to the extent ments may be based less on the therapies’ success that physicians evaluate the care given by the than on the interpersonal aspects of care—for ex- providers against the standard of physicians’ care. ample, on how courteously patients felt they were Physicians’ evaluations of the care provided by treated, how they assessed the value of the ad- NPs, PAs, and CNMs in their employ, however, vice they received, on how much time they spent might be affected by the physicians’ fiscal inter- with the providers, and on how their emotional ests. Physicians pleased with the financial results states changed (267). Nonetheless, if patients are of employing NPs, PAs, or CNMs might view dissatisfied with the services they receive, part of these providers favorably, whereas physicians dis- the reason for their dissatisfaction may be that pleased with the financial results might show their their expectations have not been fulfilled. displeasure in negative assessments of the work of these providers. Other subjective factors, such Malpractice insurance premium rates and mal- as gender or personal acquaintance, might influ- practice claims can also be used to judge patients’ ence the degree to which physicians accept NPs, satisfaction. The comparison between physicians PAs, and CNMs. Competition from NPs and CNMs and NPs, PAs, and CNMs is crude because the in independent practice, for example, certainly in- number and scope of services provided by phy- fluences physicians’ acceptance of such practi- sicians differ from those provided by NPs, PAs, tioners. and CNMs. The interpersonal aspects of care ap-

METHODOLOGICAL PROBLEMS OF STUDIES

One or more common methodological prob- Study designs contain other weaknesses. Some lems affect most studies of the quality of care pro- studies compare the processes and outcomes of vialed by NPs, PAs, and CNMs. The problems care provided by NPs, PAs, and CNMs with the include using small samples, focusing on short- processes and outcomes of care provided by house term outcomes, using nonrandomized study pop- staff rather than by experienced physicians. Study ulations, applying single evaluation criteria, using designs that compare only medical tasks as per- incomplete and unstandardized medical records formed by physicians with tasks performed by data, and choosing nonrepresentative samples or NPs and CNMs are incomplete because they ig- sites. Some studies, because they were conducted nore the advanced nursing responsibilities that by educators and other proponents of NPs, PAs, NPs and CNMs also fulfill. and CNMs, might be biased in favor of the care 4 There are a few well-conducted, randomized, given by these providers. controlled trials that are valid within their own designs. The conclusions of these trials, as well

4 N0 bias against NPs, PAs, and CNMs was apparent in the studies as other less rigorous studies, can be generalized— examined for this review. applied to other populations and settings—but 19 only in a limited way. Many studies report on are more or less flawed. Problems include misin- only a few NPs, PAs, or CNMs in only one set- terpretation of questions by respondents, inves- ting, which limits the applicability of the findings tigators’ bias in framing questions, and reliance for other providers and other settings. on the respondents’ memories. Little attention has been given to the systematic development of the Some of the studies of patients’ satisfaction and questionnaires or measuring scales used by inves- physicians’ acceptance are opinion surveys that, tigators. depending on the rigor of design and execution,

QUALITY OF NURSE PRACTITIONERS’ CARE Comparisons With Physicians lowup of problems and therapies; completeness of physical examinations and interviewing skills, Reviews of comparison studies (230,242) and and patient knowledge about the management individual studies comparing NPs and physicians plan given to them by the provider (187). find that the quality of care provided by NPs func- Table 2-2 also suggests that NPs are especially tioning within their areas of training and exper- good at assisting ambulatory patients with chronic tise tends to be as good as or better than care pro- problems such as hypertension and obesity (189, vided by physicians (50,51,72,104, 186,199,231). 211). After clinic visits for chronic problems, NPs’ In some cases, the quality of NP care is equiva- patients are less likely than physicians’ patients lent to physician care (see table 2-1). For exam- to report that their activities are limited or that ple, NPs generally resolve patients’ acute problems they experience anxiety about their problems (139). as well as physicians (130,139), and the functional Whether NPs’ interpersonal skills contribute to status of patients treated by NPs and physicians their ability to care successfully for patients with is equivalent (212). Spitzer (231) found no differ- chronic problems has not been determined. Phy- ence between NPs and physicians in the adequacy sicians, however, appear to provide better care of their prescribing practices. Other researchers in managing problems that require technical so- found that NPs prescribe and use medications less lutions (104). frequently than do physicians, and that NPs tend to prescribe only well-known and relatively sim- Patients’ Satisfaction ple drugs (29,204,225). The studies did not ascer- Overall, patients are satisfied with the care they tain whether the differences in the prescribing receive from NPs (25,41,80,82,139,141,145,207, habits of physicians and NPs stem from differ- 231,265). Moreover, patients appear to be more ences in patient mixes, prescribing philosophies, satisfied with the care they receive from NPs than or other causes. with care from physicians, in regard to several The quality of NPs’ care differs from that of factors: personal interest exhibited, reduction in physicians’ care in other instances (see table 2-2). the professional mystique of health-care delivery, NPs appear to have better communication, coun- amount of information conveyed, and cost of care seling, and interviewing skills than physicians (41,145,190), have (84,104,178), a conclusion reinforced by one literature review citing a number of “variables for A few studies, however, indicate patients’ dissatisfaction with one or more aspects of NPs’ care which nurse practitioners received higher scores than physicians. ” These variables include: or show patient preference for physicians’ care. Patients are concerned about long waits to see NPs 5 . . . amount/depth of discussion regarding child (145), about how well NPs communicate with pa- health care, preventive health, and wellness; amount 5This finding was consistent across 10 settings, including solo prac- of advice, therapeutic listening, and support of- tices, university student-health centers, public health-department fered to patients; completeness of history, includ- clinics, private-hospital outpatient clinics, and a health maintenance ing the recording of previous problems and fol- organization. 20

Table 2-1 .–Equivalence in Quality of Care Provided by Nurse Practitioners (NPs) and Physicians (MDs) -. . Activity or measure Setting Study type Source Process measures: Adequacy of pediatric physical assessment ...... Health center, low-income Retrospective chart review Duncan, et al., 1971 neighborhood Adequacy of prescribing medication ...... Two MD family practice Randomized controlled trial Spitzer, et al., 1974 Adequacy of the management of episodes of care ...... HMO Prospective; chart review, timing of Spitzer, et al., 1974; segments of patient visits Salkever, et al., 1982 Management of hypertensive patients ...... Rural primary care center Retrospective chart review Watkins and Wagner, 1982 Similarity of treatment plans for pediatric patients ...... Military outpatient clinic Retrospective evaluation of NPs’ and DiGirol and Parry, 1983 MDs’ treatment plans Short- and long-term compliance by patients ...... Emergency room Prospective study with data collection at Powers, et al., 1984 emergency room visit, short-term followup, and long-term followup Outcome measures: Patient’s physical, emotional, and social functional status . . . . Two MD family practice Randomized controlled trial Sackett, et al., 1974 Resolution of acute problems . . . . Hospital ambulatory care Record review Komaroff, et al., 1976 clinics Resolution of acute problems Prepaid group practice Survey of providers and patients with Levine, et al., 1976 telephone followup of patients at 1 week Reductions in pain or discomfort among pediatric patients...... Prepaid group practice Survey of providers and patients with Levine, et al., 1976 telephone followup of patients at 1 week SOURCE Process meesures: M.T. DiGirol and W.H Parry, “Consultation to the Pediatric Automated Military Outpatient Systems Specialist (AMOSIST): A Comparison of Consultation by a Pediatric Clinical Nurse Specialist and by a Pediatrician, ” Mi/itary Med. 146(4):364-367, April 1963; B. Duncan, AN, Smith, and H.K. Silver, “Comparison of the Physical Assessment of Children by Pediatric Nurse Practitioners and Pediatricians, ” Am. J. Pub/ic Hea/th 60(6):1 170-1176, June 1971; M J. Powers, A Jalowiec, and PA. Reichert, “Nurse Practitioner and Physician Care Compared for Nonsurgery Emergency Room Patients,” Nurse Practitiorr- er 9(2):39-52, February 1984; W O Spitzer, D L. Sackett, J,C Sibley, et al , “The Burlington Randomized Trial of the Nurse Practitioner,” N. Eng/. J. Med. 290(5):251-256, Jan. 31, 1974; L.O. Watkins and E.H. Wagner, “Nurse Practitioner and Physician Adherence to Standing Orders Criteria for Consultation or Referral,” Am. J Pub/ic f-fea/t/r 72(1):22-29, January 1982. Outcome measures: D.M. Levine, L.L. Morlock, Al. Mushlin, et al., “The Role of New Health Practitioners in a Prepaid Group Practice: Provider Differences in Process and Outcomes of Medical Care, ” Med. Care 14(4):326-347, April 1976; A.L. Komaroff, K. Sawyer, M. Flately, et al., “Nurse Practitioner Management of Common Respiratory and Genitourinary Infections, Using Protocols, ” Nurs. Research 25(2) ’64-89, March-April 1976; D.L. Sackett, “The Burlington Randomized Trial of the Nurse Practitioner: Health Outcomes of Patients,” Annals /nt Med. 80(2):137.142, February 1974, D.S. Salkever, E A. Skinner, D.M. Steinwachs, et al., “Episode-Based Efficiency Comparisons for Physicians and Nurse Practitioners, ” Med Care 20(2):143-153, February 1982. tients (139), and about whether NPs can care for dissatisfaction with NPs’ care are likely to limit what patients perceive to be serious medical prob- the employment and use of NPs as the growing lems (131). Patients are dissatisfied with NPs who supply of physicians allows more consumers to do not consult with physicians about diagnostic choose between NPs and physicians. and treatment decisions (80,198). Some of these Malpractice insurance premiums and the inci- findings, particularly those having to do with dence of malpractice claims indicate that patients waiting time and communication, contradict those are satisfied with NP care. Although insurance of other studies (41,71,104,178,195), suggesting premiums for NPs are increasing, successful mal- that some aspects of NPs’ care may require fur- practice suits against them remain extremely rare. ther research. Not surprisingly, most of the estimated $1.4 bil- Additional research on patients’ satisfaction lion in malpractice claims paid in the United States would be especially timely now, when the Na- in 1984 (62) resulted from suits against physicians, tion’s supply of physicians is growing, and more particularly physicians in the surgical subspecial- physicians seem to be locating in small towns ties. Physicians, however, far outnumber other (36,39,68,174,264), where a relatively large pro- types of providers, generally deal with the most portion of NPs have been providing health serv- complex cases, and have more financial resources ices. Any factors that might contribute to patients’ than other providers. 21

Table 2-2.—Difference in Quality of Care Provided by Nurse Practitioners (NPs) and Physicians (MDs)

Relative quality of Activity or measure care by NPs and MDs Setting Study type Source Process measure: Number of diagnostic tests NP > MD Hospital outpatient clinic Random assignment of patients Flynn 1974 record review, time and motion studies, patient interviews Number of diagnostic tests NP > MD HMO Prospective, chart review timing of Salkever et al 1982 segments of patient visits Thoroughness of documentation NP > MD Preventive medicine department Cross sectional Patient survey and Brown et al 1979 of diagnosis and treatment of a multispecialty clinic chart review Information Adequacy of a telephone NP > MD University pediatric clinic Programed calls from a trained Perrin and Goodman 1978 management of common pediatric person about selected pediatric problems problems, calls recorded and analyzed Effectiveness of Interpersonal NP > MD University pediatric clinic Programmed calls from a trained Perrin and Goodman, 1978 management skills (Interviewing, person about selected pediatric Hastings et al 1980 communicating) problems calls recorded and content analyzed Management of problems NP < MD Jail health service Record review and audit Hastings et al 1980 requiring technical solutions Outcome measures: Rate of patient return to NP > MD University hospital medical clinic Random patient assignment Lewis et al 1969 employment interviews, chart reviews Reduction in number of NP > MD University hospital medical clinic Random patient assignment Lewis et al 1969 symptoms in patients interviews, chart reviews Level of patient awareness of NP > MD University hospital medical clinic Random patient assignment Flynn 1974 provider orders interviews, chart reviews Level of control of blood pressure NP > MD City hospital and health Record review Runyon 1975 in patients with hypertension department clinics Ramsay, et al 1982 Level of control of blood pressure NP > MD University hospital hypertension Prospective record review Ramsay et al 1982 in patients with hypertension clinic Level of activity limitation and NP < MD Prepaid group practice Survey of providers and patients with Levine, et al 1976 anxiety m patients with chronic telephone followup of patients at 1 problems week Amount of reduction in pain or NP > MD Prepaid group practice Survey of providers and patients with Levine et al 1976 discomfort in adult patients telephone followup of patients at 1 week Amount of weight reduction in NP > MD University hospital hypertension Prospective record review Ramsay et al 1982 obese— patients clinic SOURCE Process measures: J D Brown, M I Brown, and F. Jones, “Evaluation of a Nurse Practlt loner. Staffed Preventive Medlclne Program In a Fee.for.Service Multlspeclalty Cllnlc.” Prev Med 8(1) 53-64, January 1979, B C Flynn, “The Effectiveness of Nurse Cllniclans’ Serwce Delivery, ” Am J PIJMIC Hea/th 64(6) 604-611, June 1974, G E Hastings, L Vlck G Lee, et al “Nurse Practlt!oners in a Ja!lhouse Clinic, ” Med Care 18(7) 731.744, July 1980, E C Perrln and H C Goodman, ‘ Telephone Management of Acute Pedlatrlc Illnesses, ” N Errg/ J Med 298(3)130-135, Jan 19, 1978 Outcome measures: B C Flynn, “The Effectiveness of Nurse Cllnlcians’ Service Delivery, ” Am J Pub/fc Health 64(6)”604.611, June 1974; D M Levtne, L L Morlock, A I Mushlln, et al., “The Role of New Health Practltloners In a Prepaid Group Practice Provider Differences in Process and Outcomes of Medical Care, ’ &fed Care 14(4) 326-347 April 1976, C E Lewis, B A Resnlck, G Schmidt, et al , “Actlvltles, Events and Outcomes In Ambulatory Patient Care,’” N Errg/ J Med 280(12) 645-649 Mar 20, 1989, J A Ramsay, J K McKenzie, and D G Fish, “Physlclans and Nurse Practlt!oners” Do They Provide Equivalent Health Care?” Am J Pub/Ic F/ea/th 72(1 )’55.57, January 1982, J W Runyon, “The Memphis Chronic Disease Program Comparisons in Outcome and the Nurse’s Extended Roles, J A M A 231(3 )”264-270, Jan 20, 1975 D S Salkever, E A SkInner, D M Stelnwachs, et al , “Episode-Based Efflclency Comparisons for Physlclans and Nurse Practltloners Med Care 20(2) 143.153 February 1982

Physicians’ Acceptance Physicians who work with NPs express more satisfaction with NPs’ performance and more will- A variety of factors affect physicians’ opinions ingness to delegate higher level tasks than do phy- of NPs. For example, physicians are more inclined sicians whose contact is indirect or nonexistent to approve NPs’ performance of relatively sim- (21,134,223). This finding might indicate quality, ple tasks, such as history-taking, than to approve but it might also reflect physicians’ opinions about NPs’ performance of more challenging clinical such non-quality-of-care factors as the relatively tasks (84,108). Another major factor influencing low cost of NP care or the freeing of time for phy- physicians’ opinions of NPs is personal contact. sicians to see more patients or to spend in leisure. 22

Physicians in group practices and in institutional is available for the decrease in the employment settings are more supportive of NPs than are solo rate, although some observers have attributed the practitioners. The level of physicians’ satisfaction slight downward trend to lack of acceptance by increases with the degree of their control over the physicians, restrictive State licensing, and un- activities of NPs (21). favorable reimbursement practices (135).7 Further- more, the validity of these statistics is questiona- Many physicians who approve of the concept ble, because they are based on a very small number of NPs have expressed only limited interest in ac- of NPs. tually employing them (134,223), although NPs and PAs were introduced and established in the United States largely because a minority of physicians chose to support, train, and hire them. About 65 percent of the NPs in the United States 7Many other factors may also contribute to the lower employ- were employed as NPs in 1982, compared with ment rate. The number of Master’s programs preparing nurse prac- 6 69 percent in 1974 (237). No documented reason titioners has grown substantially (from 74 in 1977 to 124 in 1981), and the number of certificate programs has decreased (from 124 to — 84 during the same period) (262). The decrease in employment may bMore recent longitudinal, nationwide data on NP employment also partly reflect the increased number of NPs removing themselves are not available. from the work force and seeking doctoral degrees.

QUALITY OF PHYSICIAN ASSISTANTS’ CARE Comparisons With Physicians /

,! I\\ Within the limits of their expertise, PAs provide care that is equivalent in quality to the care provided by physicians (73,92,129,230,242). What little evidence is available about how the quality of PAs’ care differs from the quality of physicians’ care indicates that PAs provide more counseling of obese patients than physicians provide (129), that PAs spend more time educating patients than physicians spend (159), and that PAs’ patients generally are better able to resume their usual level of functioning than are patients of physicians (226).

Patients’ Satisfaction The few available studies that directly address patients’ satisfaction indicate that patients generally are as highly satisfied with the care they receive from PAs as with the care received from NPs

(127,173,179,207). One study found that patients’ Photo credit: American Academy of Physician Assistants satisfaction is tempered by the desire to see PAs The care provided by PAs functioning within their areas perform routine functions rather than make in- of training and expertise tends to be equivalent in dependent diagnostic and treatment decisions quality to the care provided by physicians for (227). comparable services. 23

Physicians’ Acceptance and honor previously scheduled appointments (57). Perry and Breitner (182) found that super- Physicians initiated and developed the concept vising physicians rate PAs higher than NPs on of PAs and serve as instructors in PA training pro- tasks involving educating, counseling, or instruct- grams. PAs function as their name implies—as as- ing patients. sistants to physicians. Thus, it is not surprising The high level of physicians’ satisfaction with that many physicians accept PAs and are satis- PAs may help account for their continued high fied with their work (125,129,179,208). employment rate. Employment rates provide the Physicians’ confidence in PAs extends beyond most consequential expression of physicians’ ac- routine care. One recent study found that al- ceptance, and nearly 86 percent of the Nation’s though physicians generally delegate routine, un- PAs were employed as PAs in 1981 (45). By 1984, complicated cases to PAs, physicians also permit the employment rate had increased slightly to ap- PAs to treat walk-in patients with urgent prob- proximately 88 percent; only 8.4 percent had not lems if the physicians cannot treat those patients been employed as PAs for more than a year (219).

QUALITY OF CERTIFIED NURSE-MIDWIVES’ CARE Comparisons With Physicians managed by physicians (253). Although this might seem to indicate that CNMs provide better care CNMs can manage normal pregnancies safely than physicians, it might reflect CNMs’ referral and can manage them as well as, if not better than, of high-risk pregnancies to physicians. In one re- physicians (65,148,190,193,226). Studies show cent study, the low-birth-weight rate for CNM- that, in accordance with their training, CNMs rec- managed deliveries was 28 percent less than the ognize deviations from the norm and seek medi- control group’s rate; the CNMs had also provided cal consultation promptly (65,210). The fact that prenatal care, whereas the control group received CNMs provide standard care has been documented prenatal care from State-supported maternal and in a variety of settings, including hospital inpatient child-care clinics (184). services, hospital clinics, migrant health centers, neighborhood health centers, and private prac- CNMs appear to differ from obstetricians in tices (67). some processes of care. CNMs order medications less frequently than do obstetricians (65), low-risk As measured by such short-term indicators as patients of CNMs have shorter inpatient stays for Apgar scores (a numerical expression of the con- labor and delivery than do low-risk patients of dition of a newborn infant) and birthweight, com- obstetricians (65), more obstetrical patients of parable outcomes of normal, low-risk pregnancies CNMs are tested for urinary tract infections and result from care by CNMs and care by physicians diabetes than are patients of house staff physicians (65,196,226). CNMs’ care and physicians’ care (226), and CNMs communicate and interact more also compare with regard to birth outcomes meas- with their clients than do physicians (66,181,190, ured by fetal, perinatal, neonatal, and maternal 265). The care given by CNMs differs from the mortality (65,181). A randomized clinical trial of usual care given by the physicians in the personal uncomplicated deliveries showed no significant attention patients receive throughout labor and difference in the outcome of care whether pro- delivery. Most physicians’ care is episodic, which vided by CNMs or by the obstetric house staff, may contribute to the fact that they rely more except that CNMs kept more appointments and heavily than CNMs do on technology, such as fe- performed fewer forceps deliveries (226). tal monitoring (265). Data on birth outcomes reveal that proportion- Although CNMs are trained to provide normal, ately fewer low-birth-weight infants result from low-risk maternity services, some of them col- deliveries managed by CNMs than from those laborate with physicians to participate in the care 24 of high-risk women during labor and delivery. CNMs would have preferred to have had them These CNMs perform such tasks as: delivered by obstetricians, although some of the physicians’ patients said that in retrospect they . . . applying internal uterine pressure monitor- (181). ing devices or fetal scalp electrodes, obtaining fe- would rather have been cared for by CNMs tal scalp blood samples, managing breech or mul- Patients in a large health maintenance organiza- tiple gestation deliveries, utilizing low or outlet tion expressed satisfaction with the care they re- forceps, or utilizing vacuum extractors (10). ceived from both obstetricians and CNMs, but the CNMs’ patients were significantly more likely to Little evidence exists about CNMs’ effectiveness express great satisfaction with, and great confi- in performing these tasks, although one researcher dence in, their providers (65). This study also concluded that CNMs “can render safe, effective found that patients of CNMs were more satisfied services to about one-third of the high-risk ob- than those of physicians with the promptness with stetric population” (210). Rooks and Fischman which they could obtain their first prenatal care (203) found that most CNMs working in collabo- visit and with the relatively short time they spent ration with physicians manage the care “of prena- waiting in reception rooms (65). tal patients with some complications. ” CNMs differ markedly from obstetricians with Patients’ Satisfaction respect to frequency of malpractice suits, a crude gauge of patients’ acceptance. The number of Women served by CNMs are satisfied with the CNMs who obtained malpractice insurance un- care they receive (65,181,190,209 ).8 Although ob- der the auspices of the American College of Nurse- stetric patients from all socioeconomic strata are Midwives (ACNM) grew from 625 in 1976 to satisfied with CNMs’ care, favorable feelings 1,400 in 1983. Between 1977 and 1982, 20 claims toward CNMs increase with patients’ age, educa- (not all successful) were made against ACNM tional background, and number of births (59). Pa- group policyholders (55). A 1982 national survey tients’ satisfaction has been recorded for a wide of CNMs found that 5.2 percent (55 of 1,065 re- range of family planning services and normal spondents) had been named in malpractice suits maternity care provided by CNMs in a variety during their careers (55). By contrast, of the 1,915 of ambulatory care and hospital settings (209). members of the American College of Obstetricians CNMs also appear to be readily accepted by new and Gynecologists responding to a recent survey, patients—90 percent of the patients seeking ob- 31 percent said they had been sued once, 16 per- stetric services from a group practice of obstetri- cent had been sued twice, and 20 percent had been cians accepted services from a CNM the practice sued at least three times (55). Interpreting these had recently employed (190). figures, however, is difficult, partly because they do not reflect case-mix differences. CNMs send When comparing their satisfaction with serv- patients with complicated or high-risk problems ices provided by CNMs and obstetricians, patients to physicians, especially in emergencies. That rela- of CNMs express preferences for the greater ease tively more obstetricians than CNMs are sued of communicating with CNMs and the chance may not reflect performance as much as the fact CNMs allow them to exercise more control dur- that obstetricians deliver many more babies than ing delivery (209). Perry found that none of the do CNMs and have higher incomes than CNMs. patients whose babies had been delivered by Physicians’ Acceptance ‘Perhaps the main problem with most studies of CNMs is the possible bias resulting from nonrandom assignment of patients to differ- CNMs may practice administratively and phys- ent types of providers. Self-selection suggests that those women who accept care from CNMs are inclined to be satisfied with CNMs’ care ically apart from obstetricians and gynecologists, (just as it suggests that those women who choose care from an ob- but by functioning “interdependently with” these stetrician are inclined to be more satisfied with physicians’ services). physicians, the CNMs retain the formal support Nevertheless, studies consistently find patient acceptance of CNMs and some studies find that patients express relatively greater satis- of the American College of Obstetricians and faction with CNMs’ care than with obstetrician’s care. Gynecologists. The American College of Obstetri- -- cians and Gynecologists has agreed with the Despite the reservations of many physicians as American College of Nurse-Midwives that: to whether CNMs are needed, their employment rate has been increasing in recent years. In 1976 . . . the appropriate practice of the certified nurse-midwife includes the participation and in- and 1977, only about half of the Nation’s CNMs volvement of the obstetrician/gynecologist as were employed in clinical midwifery practice (9), mutually agreed upon in written medical guide- but by 1982, approximately two-thirds (67.2 per- line/protocols (13). cent) of the CNMs in the United States were employed in nurse-midwifery practice (10). The The two colleges further agree that: CNMs' employment settings may better reflect the Quality of care is enhanced by the interdepen- extent of physicians’ acceptance. Although the dent practice of the obstetrician/gynecologist percentage of CNMs employed in private prac- and the certified nurse-midwife working in a rela- tice with physicians increased from 13 percent in tionship of mutual respect, trust and professional 1976 and 1977 to 20 percent in 1982, most CNMs responsibility. This does not necessarily imply in 1982 were employed in organizational settings the physical presence of the physician when care or in private nurse-midwifery practice (see table is being delivered by the certified nurse-midwife 2-3). (13). Nonetheless, CNMs have had difficulty in obtaining acceptance by practicing physicians, med- Table 2-3.—Percentage of U.S. Resident Certified ical societies, hospital departments of obstetrics Nurse-Midwives by Type of Organization, and pediatrics, companies that provide malprac- 1976-77 and 1982 tice insurance, State boards of health, and—not Type of organization 1976-77 1982 infrequently—nurses, themselves (196). Obstetri- Hospital ...... 45.6% 35.8% cians and gynecologists are thought to find com- Private practice with physicians ...... 12.9 19.8 petition from CNMs threatening to physicians’ po- Private nurse-midwifery practice ...... 2.4 14.4 Public health agency ...... 13.8 8.6 sition as the sole providers of a special type of Maternity service operated medical care (43,190). Opposition may also re- predominantly by nurse-midwives . . . 7.6 7.7 flect the tightening market conditions facing ob- Branch of the U.S. military ...... 8.2 6.2 stetricians and gynecologists in urban areas (196). Prepaid health plan ...... 3.4 6.0 University health service ...... 5.0 1.8 In addition, other physicians, particularly general SOURCES” American College of Nurse-Midwives, Nurse+ 4/dwivery In the Unfted and family practitioners, have resisted CNMs States’ 1976-77 (Washington, DC 1978); and American College of Nurse-M idwwes, Nurse-&f /dwivery In the Un/ted States 1982 (258). (Washington, DC” 1964)

SUMMARY

Within their defined areas of competence, NPs, The findings for NPs and PAs apply primarily PAs, and CNMs generally provide care that is to care provided in ambulatory settings, and the equivalent in quality to the care provided by phy- activities of CNMs have been documented in a sicians for similar problems. Considerable evi- variety of settings with favorable results. Al- dence exists, particularly for NPs and CNMs, that though the findings are qualified by the method- they are more adept than many physicians at com- ological limitations of the techniques used to in- municating effectively with patients and manag- dicate quality, the weight of the evidence seems ing patients who require long-term and continu- to show that the health-care services provided by ous care. Such patients include chronically ill these practitioners are equivalent in quality to patients and patients undergoing labor and deliv- comparable services provided by physicians. ery. Although the evidence is less voluminous concerning PAs’ supportive-care and health-promot- Although patients are generally very accepting ing activities, data indicate they overlap with NPs’ of care provided by NPs, PAs, and CNMs, pa- activities of that nature. tients are most satisfied with the services that re- 26 quire interpersonal skills. Patients seem to require tion with NPs and PAs tends to be highest when what might be called technical reassurance for access to other sources of care, particularly phy- serious conditions and to prefer that NPs, PAs, sicians, is limited. Patients’ satisfaction with and CNMs consult with physicians when techni- CNMs, however, appears to be independent of cal care is required. access to other sources of obstetrical care (201). Patients’ satisfaction with NP, PA, and CNM Based on historical data, physicians accept the care is affected by factors external to the actual concept of NPs and PAs but remain concerned care provided. Satisfying a particular patient de- about their practicing independently. Physicians pends partly on the physician’s conveying to the have been reluctant to accept CNMs, especially patient a sense of approval of the NP, PA, or those practicing independently. Physicians’ will- CNM (113). Patients’ judgments may also reflect ingness to delegate tasks depends on the particu- their past experiences with medical care and their lar tasks. Most physicians who hire NPs, PAs, socioeconomic status. One study, for example, or CNMs are satisfied with their performance. found that an upper-middle-class population Employment status, the most relevant indicator accustomed to receiving care from fee-for-service of whether physicians accept NPs, PAs, and physicians evaluated providers mainly on the ba- CNMs, is satisfactory; PAs, in particular, appar- sis of technical competence (35). Patients’ age, sex, ently enjoy a high level of appreciation by phy- and race also affect their opinions. Middle-aged sicians. Increasingly, CNMs’ employment is in- people, males, and blacks are more accepting of dependent of physicians. A growing supply of NPs (80); whites are more accepting of CNMs physicians and potentially heightened competition than are blacks, who are more likely to associate may decrease physicians’ acceptance of these the word midwife with untrained lay midwives health practitioners. Indeed, the American Med- (201). The American Nurses’ Association (21) con- ical Association resolved in 1985 to “oppose new cluded that trust in NPs and PAs varies with the legislation extending medical practice to non- options available to patients, and that satisfac- physician providers” (136). Chapter 3 Access to Care Chapter 3 Access to Care

In the late 1960s and the 1970s, health policy garding payment also significantly impede NPs, focused on making health care accessible to all PAs, and CNMs by restricting payment for med- Americans; much effort went toward helping peo- ical services to the supervising physician or insti- ple enter the health care system (1). A particular tution. The Rural Health Clinic Services Act (Pub- concern was geographic access to primary care, lic Law 95-210) waived the restriction for direct because the geographic maldistribution of physi- supervision of NPs, PAs, and CNMs practicing cians and their patterns of specialization had left in certified rural health clinics located in desig- many of the Nation’s inhabitants without ade- nated underserved areas (see app. B). quate access to primary care. Whether NPs, PAs, and CNMs are needed to Indeed, the creation and development of nurse improve access to primary medical care in under- practitioners (NPs) and physician assistants (PAs) served areas remains an issue, even though the occurred in large part in response to the limited supply of physicians has increased, and some phy- accessibility of basic medical services, especially sicians have moved away from urban areas (174, in rural and inner-city areas, where physicians 264). Some experts believe that competitive pres- were disinclined to practice (74,169,183 ).1 The sures will eventually remedy the maldistribution stated purpose of the early training programs for of medical manpower (222) but, the proportions NPs was to improve access to primary care for of physicians in urban and rural areas have re- people in areas without enough physicians (236). mained fairly constant since 1970 (255). Similarly, PAs were intended to “help remedy the Furthermore, large overall increases in physi- shortage of primary care physicians, particularly cian supply in a State may still leave some areas in medically underserved areas” (180). Much of in the State without adequate access to medical the impetus for the growth in the number of cer- care (112). The situation may worsen in those tified nurse-midwives (CNMs) during the 1970s areas as older physicians are not replaced by can be attributed to concern about the limited sup- younger ones. Indeed, the Bureau of Health Pro- ply of obstetricians in the United States (180). fessions has predicted that unmet needs for pri- The various barriers to providing care must be mary care will persist in many currently desig- considered in assessing the success of NPs and PAs nated shortage areas. Although the dispersal of in improving health care in medically underserved young primary-care physicians is expected to re- areas. Legislation and regulations vary widely from duce overall shortages, reducing shortages in all State to State but generally tie medical practice underserved areas may take an extensive period by NPs, PAs, and, to some extent, CNMs to asso- of time (250). ciations with physicians and limit such practice Although the need remains for NPs, PAs, and where physicians are not present. Although NPs CNMs to provide care to underserved populations may provide nursing services independently, for and in underserved areas, interest has increasingly the most part neither NPs nor PAs ‘can provide focused on these providers’ abilities to deliver medical services unless local physicians are will- good medical care in certain institutional settings, ing to hire them. Medicare and Medicaid rules re- such as jails, and to specific populations, such as elderly people and poor women and their infants. In addition, by functioning as case managers, these IOther factors, including improved integration of nursing and providers can help patients find appropriate care medicine, bolstered the NP movement, which signified a deliberate move to expand the nursing role and to meet the health-care needs in our increasingly complex health-care system. of many underserved populations. Other factors that contributed (The effect of NPs, PAs, and CNMs on access to to the success of NPs, PAs, and CNMs are the consumers’ and specific services, such as health education, coun- women’s movements, the new focus on self-help and self-care, and other pushes for social and personal change that emerged during seling, and health promotion, is addressed more the late 1960s and continue today (229). completely in chapter 2.)

29 30

NURSE PRACTITIONERS’ CONTRIBUTIONS TO ACCESS TO CARE

Although legal constraints (such as require- having satellite practices (most of which were ments for supervision by physicians) have hin- staffed by NPs; some by PAs) illustrates this de- dered NPs’ dispersal to isolated settings, NPs have cline. By 1979, only 24 of the centers were staffed helped improve geographic access to primary care by NPs or PAs alone (37). By 1984, 18 were staffed (31,86,160,168,261). In 1977, 23 percent of NPs only by NPs or PAs, 8 were staffed only by phy- worked in inner-city settings and 22 percent in ru- sicians, and 6 were staffed by a combination of ral areas (238)—the geographic areas of greatest physicians and NPs or PAs. In all but 4 of the need (120). In 1980, the proportion of NPs work- remaining 12 communities, where satellite clinics ing in these settings had increased to 47.3 percent had ceased functioning, physicians’ practices had in inner cities but decreased to 9.4 percent in ru- been established (38). ral areas (255). In both inner cities and rural areas, More recently, NPs’ contribution to access has more than half of NPs’ patients had annual in- been in nongeographic settings where not enough comes of less than $10,000 (255). physicians have been available. Case studies re- NPs alone cannot entirely resolve the problem port the satisfactory performance of NPs in a wide of provider maldistribution, because the profes- variety of settings. NPs act as team members in sional, social, and cultural attractions of the home health and nursing home care for elderly suburbs and cities that appeal to many physicians patients (220) and in correctional institutions also appeal to many NPs. An early survey of NPs (104), and in home health care for children with in six States found that generally they “do not chronic illness (234). NPs also provide terminal work in the inner city or in rural areas” (81), but care in patients’ homes (268); ambulator y care in a Pennsylvania NP-training program surveyed its large municipal teaching hospital units (30); and graduates through 1982 and found that 70 of the primary care in inpatient units (224), in normal 102 graduates worked in urban programs with newborn nurseries (188), and in occupational health low-income people (151). settings (26). NPs also deliver preventive care in the workplace (216), in retirement communities NPs tend to view themselves as being able to (109), and in industrial settings (47,162). These function effectively and appropriately not only descriptive reports are only a beginning; larger in settings with physicians, but also in practices scale studies are needed to evaluate the quality without physicians on the premises. Starting in of care NPs provide in these settings. the mid-1960s a significant minority of NPs worked in satellite settings as the sole providers Whether NPs can improve access to health care of services; they received medical supervision in schools has been carefully examined. A large- from physicians working in other communities. scale study, involving 18 school districts in 5 Often, the backup physicians would be available States, reports that NPs working as part of health- for telephone consultations, would visit the sat- care teams in schools can have highly favorable ellite settings, and would be responsible for en- effects on school children’s health (197). NPs are suring that the NPs adhered to the protocols guid- especially valuable in improving access to primary ing the provision of medical services. These NPs care and supplementary care in rural areas and increased access to care by working in places in health programs for the poor, minorities, and where physicians had not located. people without health insurance. NPs’ extension role is no longer as significant People over 65, a growing segment of the pop- as it was in the 1960s and 1970s. A national sam- ulation, suffer serious gaps in their ability to ob- ple of 44 rural communities identified in 1975 as tain health care. Many physicians lack the expertise or time required for managing all aspects of ‘Requirements for physicians’ supervision of NPs vary from State to State. In many States, physicians must be on the premises but elderly patients’ health problems. Although pri- not necessarily in the same rooms as the NPs providing the services. vate attending physicians provide most of the ‘The communities had populations of less than 10,000, with an medical care in nursing homes, many physicians average population of less than 2,000, and were at least ‘ 2 hour in travel time from communities that had populations of more than are unwilling to care for patients in nursing homes 10,000. (166). 31

NPs are trained to care for the older popula- health professionals, interservice transfers, and tion. Indeed, 40 of the approximately 200 NP- continuity of care, and in mobilizing family, in- training programs focus on geriatrics, and 31 other stitutional, and community resources (77). NP programs have gerontological components NPs also are particularly effective in improving (254). Furthermore, much of the care that institu- access to care for groups that, for a variety of rea- tionalized elderly people need is the kind that NPs sons, have difficulty in obtaining the care they can best give—health maintenance, personal assis- need. For example, NPs and PAs work well as tance, chronic-disease management, recognition members of multidisciplinary teams in improving of acute or exacerbating chronic conditions, on- access for chronically ill elderly people, whose going accurate and comprehensive health assess- needs for health services are great and whose abil- ment, appropriate and expeditious referral to other ities to manage the health-care system are limited team members, medication management and re- (155). The NPs and PAs facilitate linkages between view, coordination of daily services, family and the community and the nursing home. NPs, work- patient education and counseling, and so on. NPs ing as members of teams with physicians, are also have the assessment skills to recognize compli- effective in educating couples about the nature of cated acute illnesses or serious exacerbations of treatment for infertility and in providing emotional chronic diseases and to make medical referrals support to people seeking such treatment (175). (157). In general, NPs appear to improve continuity The few available studies show that NPs have of care. In institutional settings, their patients miss the professional ability to assist with the care of fewer appointments than do physicians’ patients (124,220,262). institutionalized elderly patients (30). Studies have generally shown that patients But of the more than 23,600 nursing homes in the of NPs in fee-for-service settings (34,84), as well United States, only approximately 250 have ger- as in clinics and health maintenance organizations iatric NPs on their staffs providing patient care (225), have higher rates of completed followup (76). Interest in the effectiveness of NPs in nurs- visits than do patients of physicians (213). These ing homes is growing rapidly, however, as evinced findings may explain the special success NPs have by the number and size of current studies of the 4 in caring for chronically ill patients and may re- issue. flect the adequacy (or inadequacy) of relationships NPs improve access for the general population between the practitioners providing care and the by acting as case managers, matching the needs patients. of patients with appropriate services (88). NPs are NPs affect access by expanding the scope of care effective in coordinating the care of many other for their patients into dimensions that physicians —.— might ignore. For example, some studies show ‘Ongoing studies include a large-scale research project measuring how geriatric NPs employed in nursing homes affect the qual- that NPs provide greater amounts of health edu- ity and costs of care. This project is being conducted by the Moun- cation than do physicians. NPs are more likely tain States Health Corp., the Rand Corp., and the University of than physicians to explain why medications are Minnesota School of Health Sciences and funded by the Health Care Financing Administration and the R.W. Johnson and the W.K. Kel- administered and what side effects are possible, logg Foundations. The faculties of the Geriatric Nursing Programs and to discuss health-promoting behaviors with at the University of Arizona, the University of California at San patients (34,84). Unfortunately, these studies do Francisco, the University of Colorado, and the Univeristy of Wash- ington are examining the role of the geriatric NP in concert with not say whether the need for health education is the study, and the Group Health Cooperative of Puget Sound has greater among the patients seen by NPs or among received funding from the Fred Meyer Charitable Trust to evaluate those seen by physicians. NPs employed by the health maintenance organization to serve elderly enrollees living in nursing homes—if a Medicare waiver of NPs spend about 50 percent more time than mandatory physician visits can be obtained (157), In addition, the Health Care Financing Administration has granted a waiver under physicians spend on each encounter with a pa- Medicare and Medicaid to permit fee-for-service reimbursement for tient (143). The time an NP spends over the course the provision of medical services to residents of nursing homes by of an illness, especially a chronic illness, may be physician-supervised NPs and PAs. A cost and utilization evaluation is being carried out by the Health Care Financing Administra- less than that spent by a physician, however, be- tion’s Policy Center at Rand. cause the NP has fewer encounters with the pa- 32

tient (143). The fact that NPs provide a more per- offering more advice to them (178). However, evi- sonal kind of care may account for the greater dence from other studies is insufficient to support time they spend with patients. One study found or refute this study’s finding, and other factors that pediatric NPs are as efficient as physicians may play a role. For example, the greater amount in gathering historical data and suggesting ther- of time NPs spend with patients might be due in apy, and attributed the NPs’ greater time per en- part to management. When NPs are used efficiently counter to greater communication with patients— in practices, physicians might be able to spend less gathering more information from patients and time with patients.

PHYSICIAN ASSISTANTS’ CONTRIBUTIONS TO ACCESS TO CARE

PAs have also contributed notably to improv- Figure 3-1 .—Distribution of Physician Assistants ing geographic access to care. A number of studies by Size of Community have shown that they are more interested than physicians in locating in nonaffluent, medically underserved areas with high percentages of non- white populations (90,137,147,169). This willingness is reflected in statistics on where PAs practice in the United States. Whereas about 27 percent of the general population and 14 percent of the Nation’s physicians are located outside standard metropolitan statistical areas (SMSAs), 32 percent of PAs practice outside SMSAs (49). And the percentage of PAs working in communities with populations of 10,000 or less has remained constant from 1974 to 1981 (45). The 1984 Masterfile Sur- vey of Physician Assistants reports that 6.5 percent of PA respondents were located in rural areas 0% 5% 10% 150/o 20 ”/0 Average community population = 980,235 of fewer than 10,000 people and that 40 percent 50,000 SOURCE American Academy of Physician Assistants, 1984 Physician Assistant were in communities of fewer than peo- Masterfile Survey (Arlington, VA 1984). ple (6) (see figure 3-l). More NPs than PAs have staffed rural satellite health centers (38), perhaps because some NP- Few physicians are trained in geriatric medicine training programs recruited students from rural (126), and the inadequacy of physician services areas hoping they would return there as NPs. for the growing population of institutionalized Nonetheless, in States that permit satellite clinics elderly patients is especially serious (122). Al- and permit PAs to practice apart from physicians, though more and better physician care for these a significant minority of PAs work in such set- patients may be available in the future (122), tings (45), whether physicians can satisfy all the health-care As members of health-care teams, PAs have im- needs of this group is questionable. proved access to care in settings where sufficient The potential of PAs in providing care for the physician care is not always available. PAs are elderly has been discussed in the literature (160, employed in industrial organizations; community 215,218). Nearly 5 percent of PAs now provide clinics; drug and alcohol abuse clinics; nursing care in nursing homes—the same proportion as homes and extended-care facilities; and Federal, in 1981 (6). The Federal Government has recog- State, county, and city prisons (25). nized this potential and requires an increased ger- 33 iatric content in the curricula of federally funded noted the need for more uniform teaching of ger- PA-training programs. A survey of 34 federally iatric medicine in training programs. (The report funded programs’ curricula, in fiscal year 1983, includes guidelines for standardizing geriatric cur- reported that three-fourths had varying degrees ricula during the training period and in continu- of geriatric content (254). Furthermore, the Fed- ing education programs for PAs. ) eral Government (through the Administration on PAs have also expanded the scope of care that Aging of the Office of Human Development Serv- most patients receive. PA training programs re- ices of the U.S. Department of Health and Hu- quire competence in interviewing, educating, and man Services) partly supported the American counseling patients (93). Although research is lim- Association of Physician Assistants in its report ited as to the interpersonal components of care on the assessment and improvement of PAs’ knowl- (215). that PAs provide, they appear to expand access edge and skills in geriatrics The report to patient education and counseling by mixing found a fivefold increase in the number of re- competence in technical care with interpersonal quired and elective experiences in geriatrics among skills (182). PA programs since 1980, which appear related to the Federal funding rules. However, the report

CERTIFIED NURSE-MIDWIVES’ CONTRIBUTIONS TO ACCESS TO CARE

Modern nurse-midwifery started in this coun- fees (53,65). One study, however, found that try in 1925, when Mary Breckenridge established CNMs’ fees exceeded physicians’ fees in urban the Frontier Nursing Service to serve rural Ken- locations (200), but nearly a year had clasped be- tucky. As of 1977, 10 percent of CNMs worked tween the measurement of physicians’ fees and the in communities with populations below 10,000 measurement of CNMs’ fees, which may account (9). CNMs still practice extensively in underserved for the finding. Also, a disproportionately large areas, such as the rural South, Indian reservations, number of CNMs practice in academic medical and inner cities, and significantly improve access centers, which have higher costs than community to health care in those areas. For example, in hospitals (200). Holmes County, Mississippi, the infant mortality rates dropped from approximately 38 per 1,000 CNMs affect access (as well as quality) by pro- live births to 20 per 1,000 live births 2 years after viding person-oriented services, such as commu- CNMs began providing primary care to pregnant nicating thoroughly with patients, counseling, women as part of a communitywide focus on the promoting self-help, and attending to patients’ health problems of mothers and babies (158). emotional needs (196). CNMs interact with patients more than physicians do (66,190,265). Pa- CNMs have also reduced financial barriers to tients feel more comfortable about asking ques- access by providing care at relatively low cost, tions of CNMs than of physicians (181,190). In particularly in short-stay, out-of-hospital births. addition, CNMs’ patients obtain care relatively Many such births occur in birth centers not af- early in their pregnancies and continue to receive filiated with hospitals. The number of these cen- prenatal care relatively frequently (140,193,226). ters increased from 3 in 1975 to more than 100 CNMs tend to increase the amount of prenatal in 1982 (33). They have made prenatal, labor and care their patients receive. delivery, and postnatal services increasingly accessible to poor patients (65,149,193). For example, In general, then, CNMs continue not only to 15 birth centers are accessible to families in New lower financial barriers to care, but to offer a con- York’s Lower East Side, a low-income area (150). siderable amount of care that includes both health The relatively low cost of CNMs’ services may advisory and health-promotion services. This ex- result from shorter inpatient stays as well as lower pertise is reflected in the valuable care CNMs on 34

multidisciplinary teams provide for high-risk pregnant adolescents (184), especially in clinic settings (42). Indeed, the Institute of Medicine’s report on preventing low birthweight calls for: . . . more reliance . . . on nurse-midwives . . . to increase access to prenatal care for hard-to-reach, often high risk, groups. This recommendation is based on the studies that indicate that CNMs can be particularly effective in managing the care of pregnant women who, because of social and economic factors are more likely to deliver low weight babies (121).

Photo credit American College of Nurse-Midwives CNMs are particularly effective in managing the care of pregnant women who are not at risk of having low-weight babies.

SUMMARY

NPs and PAs have long been recognized for in- more likely than older physicians to accept the creasing geographic access to primary health care, team approach to health care and to use the serv- particularly for residents of inner cities and rural ices of NPs and PAs. Furthermore, a small town communities. Although indications are that phy- might be able to support a physician-NP or a sicians are migrating to smaller communities (174), physician-PA team but not two physicians. Whether the growing supply of physicians appears to be these factors or others reduce the role NPs and affecting different communities differently (250). PAs play in improving geographic access to care, Overall increases in the supply of physicians in these practitioners will continue to be valuable, a State may still leave some areas in need of pri- especially in rural areas. mary care services (112). In those areas where ac- The evidence (primarily from case studies) is cess to physicians’ services remains inadequate to that NPs and PAs are improving access to pri- serve the population or has decreased (112), NPs mary health-care services in settings not adequately and PAs can continue to serve as a source of pri- served by physicians. For example, NPs and PAs mary care. In areas where access to physicians’ are trained to provide primary care for elderly pa- services has increased, employment opportunities tients in nursing homes, a growing population for NPs and PAs might decrease. But the employ- with poor access to standard health care. The ef- ment of NPs and PAs in rural areas has previously fectiveness of NPs and PAs in this role is under been limited by the scarcity of physicians willing scrutiny. They are also helping people to obtain both to practice in rural areas and to supervise primary care in an increasingly complex health- NPs and PAs. Thus, the growing numbers of phy- care system. sicians in previously underserved areas may well increase employment opportunities for NPs and Studies have shown that NPs are especially val- PAs. The physicians moving into smaller com- uable in providing primary care in school settings munities are mainly young physicians, who are to previously unserved or underserved children, 35 and in expanding the content of available care to care by the personal orientation of their services. include interpersonal and preventive care for all Studies have shown that CNMs’ communication patients. skills and attention to the social and psychological needs of pregnant adolescents, as- well as the CNMs have not only made care more accessi- technical care CNMs provide, have reduced the ble in underserved areas, they have also contrib- rate of low-birth-weight babies among this high- uted to making care financially available and have risk population. contributed to social and psychological access to Chapter 4 Productivity, Costs, and Employment Chapter 4 Productivity, Costs, and Employment

Several studies have examined the scope of practice must spend to employ an NP, PA, or practice and productivity of nurse practitioners CNM and how much society must spend to train (NPs), physician assistants (PAs), and certified these types of practitioners. Questions related to nurse-midwives (CNMs); how that scope relates employment compare productivity with the costs to the tasks usually undertaken by physicians; and of employment to ascertain whether medical prac- the implications of this evidence for the employ- tices could gain from employing more NPs, PAs, ment of these providers and for the costs of med- or CNMs, and whether society could gain from ical care. training more NPs, PAs, and CNMs. Because of the complexity of the issues involved and the lack Questions related to productivity include the of data, these questions are seldom addressed to- nature and size of the contributions NPs, PAs, gether. The literature does, however, permit the and CNMs make to medical practices’ outputs piecing together of some parts of this puzzle. (e.g., encounters between providers and patients). Questions related to costs include how much a

SCOPE OF PROFESSIONAL PRACTICE Services Provided by Nurse Practitioners and Physician Assistants The tasks NPs and PAs are trained to perform encompass a broad spectrum of primary care activities involving diagnosis and therapy (see ch. 1). Distinguishing between NPs and PAs on the basis of task descriptions is difficult. NP training may emphasize counseling and health promotion activities to a greater degree than PA training does, but the major difference lies in the practitioners’ relationships with physicians. By definition, PAs work under physicians’ supervision, whereas NPs have collaborative relationships with physicians and other health professionals. Most observers conclude that most primary Photo credit” Arner/can Nurses Assoclatlon care traditionally provided by physicians can be NPs are trained to perform a broad spectrum of primary-care activities. delivered by NPs and PAs. Hausner and others (105) conclude that 60 to 80 percent of the tasks normally performed by primary care physicians of the primary care responsibilities to be consid- can be provided by NPs and PAs without consul- ered viable alternatives in providing primary care, tation. Record and others (192) estimated that 90 even where direct supervision is unavailable. percent of pediatric care can be provided by NPs and PAs, and that NPs and PAs can substitute What NPs and PAs are trained to do and what for physicians in providing 50 to 75 percent of they do in practice maybe different. Their actual all primary care services. Hausner and others (105) roles depend on the settings in which they work. argue that NPs and PAs can safely perform enough Limited information exists as to how practicing

39 40

NPs and PAs actually spend their time. A 1979 of the ‘whole patient’ “ (160). These generalized review cites four reports indicating that “nurse characterizations do not apply universally, but practitioners, in particular, emphasize preventive they illustrate an important distinction between services, ” including one report concluding that PAs and NPs: PAs tend to function primarily as NPs can provide as much as 75 percent of the well- substitutes for physicians, generally providing person care for both adults and children (218). only services that physicians provide, whereas Other studies have found that NPs engage more NPs are likely to provide both services usually often than physicians in providing interpersonal provided by physicians as well as services gener- care (221) and chronic care (32). However, beyond ally provided by nurses. these sorts of indications and references to the NP orientation to health education, counseling, and Services Provided by preventive and chronic care, accurate descriptions Certified Nurse-Midwives of the actual specific tasks performed by NPs do not exist. Indeed, such information would be dif- In 1982, the American College of Nurse-Mid- ficult to obtain, because the range of primary care wives (ACNM) (10) conducted a survey of its mem- services provided by NPs in outpatient settings bers which obtained detailed information about is so broad. the specific tasks performed by CNMs in clinical practice. Of the approximately 1,000 CNMs re- Little information exists concerning trends in sponding, over 75 percent delivered prenatal, la- the freedom of NPs to function independently of bor, delivery, and postpartum care as well as fam- physicians. Nearly two-thirds of the pediatric NPs ily planning and normal gynecological services. responding to a national survey in 1978 said that The CNMs’ responses to detailed questions about a physician was always physically present when tasks showed that they provide the full range of they worked. Only 39 percent of the respondents services within their areas of expertise and they to a similar survey in 1983 noted that a physician assume specific responsibility for many of the was always present (44). These findings suggest tasks which they perform without physician direc- some movement toward administrative independ- tion and supervision. CNMs clearly can substi- ence, but more data on other types of NPs work- tute for physicians in performing a significant ing in a variety of settings are required in order share of the tasks normally carried out by physi- to establish whether the trend is significant. cians. A major difference between CNM care and Although PA training programs also include physician care is that CNMs are less likely than health education and counseling, relatively little physicians to prescribe drug treatments, which empirical evidence exists on how much health- may reflect both philosophical differences and le- promotion and disease-prevention services PAs gal restrictions. CNMs also tend to use less high- actually provide. In general, PAs tend to focus priced technology than physicians, and CNMs do more than NPs on providing acute care services not perform major surgery. In collaboration with (138). PAs place less emphasis on preventive serv- physicians, however, CNMs manage high-risk pa- ices (218) and “provide selective patient services, ” tients during the prenatal, labor, and delivery whereas NPs are oriented more “toward treatment stages.

PRODUCTIVITY

If the tasks performed by NPs, PAs, and CNMs by providing some services, such as counseling overlap substantially with those performed by or health education, not currently provided by physicians, an obvious potential exists for these many physicians or not carried out to the same providers to substitute for physicians in the sense extent. of performing tasks typically and characteristically carried out only by physicians. NPs, PAs, Whether a service is a substitute or a comple- and CNMs can also complement physician care mentary service is often difficult to determine. 41

Technically, empirical measurement of substituta- Nurse Practitioners’ and bility is complicated by the need for large amounts Physician Assistants’ Productivity of accurate data on the prices and utilization levels of resources used in the production process as well Studies of NPs’ and PAs’ productivity have gen- as on the output of the production process. There- erally taken one of three approaches: fore, studies of the role of NPs, PAs, and CNMs 1. time per visit (comparing how much time have taken the more straightforward approach of physicians and NPs or PAs take to complete productivity analyses based on small samples, office visits); case studies, or simulations. 2. average number of visits per unit of time Productivity, simply stated, is output per unit (comparing how many visits different types of input. The productivity of medical practitioners of providers handle in a given period of is frequently expressed in terms of the number of time); and 3. patients seen per week or per hour of the practi- marginal product (assessing the effect of add- tioners’ time. In comparing physicians with NPs, ing an NP or PA on a practice’s total num- PAs, and CNMs, the appropriate method of meas- ber of patient visits). uring productivity depends on whether the NPs, Most studies of NPs and PAs indicate that these PAs, or CNMs are working under direct super- providers spend more time per office visit than vision by physicians or working interdependently do physicians (242). For example, Mendenhall and with physicians. For example, studies of PAs others (160) found in a national survey of physi- directly supervised by physicians examine how cian practices that NPs averaged 19.4 minutes per employing PAs marginally affects total practice direct encounter with a patient, PAs averaged 13.3 output (e. g., the additional number of patients minutes per encounter, and physicians spent slightly seen per week). Or time-and-motion studies of the more than 11 minutes per encounter. A study by production process might examine the tasks per- Charney and Kitzman (52) yielded similar results, formed by PAs and how long they take, as com- but studies are not unanimous on this issue. In pared with the time physicians would take. To a large health maintenance organization (HMO) evaluate the productivity of practitioners work- —a special setting—Record and others (191) re- ing in collaboration with physicians, as CNMs ported that PAs spent less time per routine visit work, studies could compare the number of pa- (an average of 7.1 minutes) than physicians did tients seen per week in collaborative practice with (8.9 minutes). The study noted, however, that: the number of patients seen for the same service by an obstetrician. Physicians could also be com- . . . a sampling of medical charts revealed that even where the presenting morbidity was the same, pared with NPs, PAs, or CNMs with regard to physicians tended to get somewhat older patients the number of minutes required per encounter for with a greater number of associated morbidities, a particular type of patient or medical service. including chronic diseases, which might easily ex- This approach attempts to control for case mix. plain the time difference. Comparing the productivity of physicians and Also, Kane and others (129) found little differ- PAs is facilitated by the fact that the tasks they ence in the amount of time physicians and physi- perform overlap significantly. Indeed, PAs tend cian assistants spent per visit. These data support to provide essentially the same services physicians the conclusion reached by Record and her col- perform. The need to understand differences in leagues (192) in a review of more than a decade content of care, therefore, is not as great in com- of experience and studies, that “there is more of paring physicians with PAs as in comparing phy- a tendency for NPs than for PAs to vary from sicians with NPs, who generally provide a much physicians in the average amount of time spent wider range of services. on an office visit. ” 42

The shorter average time physicians, as com- to estimate that the productivity gain maybe only pared with NPs and PAs, spend with patients 20 percent after calculating the “offsetting changes translate into greater productivity over time. In in measures such as provider time available for other words, the number of encounters with pa- nondirect patient care activities, patients’ wait- tients per hour or per work week is higher for phy- ing time, waiting room congestion, practice hours, sicians than for NPs or PAs. Mendenhall and and supervisor y requirements. ” The findings of others (160) reported the following: Mendenhall and others (160) indicate that even ● NPs average 7.9 direct encounters and 2.4 though direct encounters between patients and the telephone encounters with patients per day; supervising physician decline when an NP or PA ● PAs average 14.2 direct encounters and 2.6 is hired, the practice’s total output increases. Rec- telephone encounters with patients per day; ord and others (192) reported “greatly varying re- ● physicians who supervise NPs or PAs aver- sults” in studies of how adding an NP or a PA age 18.9 direct encounters and 3.4 telephone to a practice affected its productivity. Some studies encounters with patients per day; and found NPs and PAs to have greatly increased ● physicians who do not supervise NPs or PAs productivity, and other studies found that add- average 21.4 direct encounters and 5.7 tele- ing PAs or NPs actually decreased the number of phone encounters with patients per day. patients seen. The one fact about which researchers appear to agree is that the potential for increas- Data from a recently completed national suring productivity is greater in large practices than vey of rural health care delivery organizations in- in small ones (111,192). dicated that primary care physicians saw an average of 105.6 patients per week and worked 48.6 Three major problems arise in assessing produc- hours per week, whereas NPs and PAs saw an tivity in terms of length of encounter or number average of 75.0 patients per week and worked 40.7 of patients seen per unit of time. First, these units hours per week (107). On the average, then, these of measure do not reflect the content of the care physicians, saw 2.2 patients per hour, compared provided or the severity of the patients’ illnesses. with 1.8 patients per hour for NPs and PAs. Romm Because some visits require more skill than other and others (199) found that, compared with PAs, visits Holmes and others (114) applied a relative- NPs spent more time per patient and, therefore, value measure of productivity, considering both saw fewer patients per week. Because physicians the number of visits and the complexity of those work more hours per week than do PAs and NPs, visits. The researchers found that although phy- these productivity comparisons are best made on sician-NP teams handled only 5.7 patient visits a per-hour basis, i.e., adjusting for the number more than physician-nurse teams handled each of hours worked per week. Overall, the findings day, the teams with NPs were 26 percent more indicate that, in terms of patients seen per unit productive in terms of total value-weighted serv- of time, NPs are less productive than PAs, who, ices (114). The difference in content of care is an are less productive than physicians. However, this important consideration because NPs provide more result does not adjust well for severity of illness time-consuming services, such as health educa- (i.e., case mix), nor does it necessarily mean that tion and counseling, than do physicians and phy- physicians are relatively cost-effective. For exam- sicians are capable of providing some medical ple, physicians might be three times more produc- services that NPs cannot provide. Measures un- tive than NPs and PAs are, but cost six times as adjusted for content and complexity of work may much as they do. yield biased estimates of relative productivity. The extent to which hiring an NP or PA in- The second major problem in basing produc- creases the output of a practice has been the sub- tivity estimates on numbers of patients or lengths ject of some debate (110,111,153). LeRoy (138) of visits is that these measures inadequately re- reported increases of between 20 and 90 percent flect the ultimate objective of medical care. The in the productivity of physicians’ practices that purpose of medical care is to treat and prevent added NPs. Hershey and Kropp (110) used a model health problems rather than to provide individ- 43 ual services. Recognizing this fact, Salkever and would be too demanding and that patients pre- others (213) examined the productivity of physi- ferred to receive care from physicians (125). cians and NPs in terms of episodes of care, because episode-based assessments account for dif- In addition to reflecting physicians’ willingness ferences in referral, and because “the episode is or unwillingness to delegate responsibilities, the also a more appropriate unit for measuring differ- productivity of NPs and PAs depends on many ences in effectiveness of care, since the outcome factors, including practice type (solo or group), of the care process may be causally related not practice setting and size, case mix, how long the only to a service received at a single visit, but to NPs or PAs have been practicing, practice regu- any services received over the course of the epi- lations, and how much autonomy the NPs or PAs sode.” The researchers found that the per-episode have. Many of these factors are beyond the con- costs were about 20 percent lower when NPs were trol of NPs and PAs, however, which means that the initial providers than when physicians were the potential or capacity of NPs and PAs has a the initial providers. limited effect on their productivity and, consequently, on their ability to affect the cost of care. A third major problem in ascertaining produc- Indeed, most productivity analyses consider NPs tivity is that existing studies reflect current sub- and PAs as part of physicians’ practices. Little evi- stitution practices, which may not fully exploit dence exists as to the productivity and cost-effec- the potential for using NPs and PAs cost-effectiveness of NPs and PAs as autonomous practi- tively. The fact that NPs and PAs can safely per- tioners. form numerous medical-care services suggests that these practitioners have the capacity to be highly In sum, the studies of the productivity of NPs productive as individuals and to contribute sub- and PAs suggest that: stantially to the productivity of the organizations ● physicians can substantially increase their in which they work. But a key factor affecting practices’ output by employing NPs or PAs the productivity of NPs and PAs is the extent to who operate under the supervision of phy- which their employers—often physicians—are sicians; willing to delegate tasks to them. ● although PAs, and, especially, NPs see fewer The evidence about what physicians actually patients per hour than physicians see, these delegate as opposed to what they can safely del- practitioners are capable of carrying substan- egate is limited. A recent study of physicians in tial proportions of the workloads of primary- care physicians; and a large HMO (125) found that physicians did not ● practice setting may be an important factor delegate as many tasks as they thought NPs and in NPs’ and PAs’ productivity, as evidenced PAs could handle safely. General internists, pediatricians, and obstetrician/gynecologists indicated by the differences in the use and productivity of NPs and PAs in HMOs and traditional that 49, 46, and 29 percent, respectively, of their settings. total office visits could be shifted safely to PAs and NPs. The internists and pediatricians, how- The potential suggested by these studies is lim- ever, were willing to shift only about 28.5 per- ited by the reluctance of physicians to delegate cent of their visits to NPs and PAs, and obstetri- tasks. Evidence shows that physicians are reluc- cian/gynecologists were willing to shift only about tant to use NPs or PAs even to the extent that 14 percent of their visits. Most pediatricians and physicians think feasible and safe, basing their obstetrician/gynecologists cited their patients’ reluctance on patient preferences. preferences for being treated by physicians and the physicians’ own needs to maintain overall pro- Certified Nurse= Midwives’ Productivity ficiency by seeing a full range of patients as the primary reasons for not delegating more. The pri- Compared to the many studies of NPs and PAs, mary reasons most internists cited for not delegat- much less information is available on the producing more were that seeing only complex cases tivity on CNMs. Furthermore, “it is characteris- 44

tic of the nurse-midwifery studies that they con- interpersonal aspects of care, such as counseling, centrate on outcome” (67). This almost exclusive health education, and patient interaction (103, focus on outcome rather than process limits in- 184). Such an understanding is necessary in or- formation about CNMs’ involvement in produc- der to specify what facet of the care provided by ing services. CNMs contributes to the positive outcomes their patients experience (226). One study (253) indicated that CNMs were only “about 23 percent as productive as obstetricians Data from the ACNM survey (1984) suggest when the number of deliveries was used as the substantial possibilities for CNMs to substitute for output measure. ” But the same study reported physician care. Many CNMs are already assum- when the volume of patient visits was used as the ing responsibility for a wide variety of complex output measure, CNMs were 98 percent as pro- tasks in prenatal, labor, delivery, and postpar- ductive as obstetricians. tum care. As with NPs, the content of care provided by CNMs must be understood because they stress the

COSTS AND EMPLOYMENT Although considerable scope exists for substi- reimbursement policies, legal barriers, and prac- tuting of NPs, PAs, and CNMs in providing some tice setting characteristics. Furthermore, NPs, of the care traditionally provided by physicians, PAs, and CNMs sometimes compete with profes- the resulting increases in productivity are not sionals other than physicians or operate independ- enough, by themselves, to justify greater employ- ent practices. Nonetheless, given the large over- ment of these practitioners in private practices. lap of their practices, primary care physicians From the standpoint of a private firm, the mar- provide an appropriate comparison group for con- ginal value (as measured by the amount patients sidering the employment of NPs, PAs, and CNMs. would pay for the additional output) must com- Although some information is available about sal- pare favorably with the marginal cost (i.e., the aries, the figures are imprecise enough that the salary and related expenses) of hiring an NP, PA, discussion must be carried out in approximate and or CNM. From the perspective of a long-run in- qualitative terms. vestment in training, either by society or by the trainees, the value (i.e., compensation) placed on Costs and Benefits of Training Nurse the output of the NPs, PAs, or CNMs must com- Practitioners, Physician Assistants, pare favorably with the costs of training to justify and Certified Nurse-Midwives expending the resources. Estimates of the social and private rates of re- In 1983, annual salaries for NPs, PAs, and turn to investments in training and education in- CNMs averaged about $25,000, compared with dicate the value placed on these investments by the $60,000 to $80,000 median salaries of primary- society and private individuals, respectively. The care physicians (18). This wage gap raises several best of such computations require large amounts questions. What are the costs and benefits to so- of data on earnings over the career of the indi- ciety of using NPs, PAs, and CNMs rather than vidual. However, some conceptual issues can be physicians? And if NPs, PAs, and CNMs are cost- addressed qualitatively. In theory, the rate of re- effective substitutes, why isn’t their employment increasing relative to the employment of physicians? turn on investment in the training of NPs, PAs, or CNMs can be calculated without reference to NPs, PAs, and CNMs, clearly could not com- the training or earnings of physicians. Society pletely replace physicians, because the scope of must expend a certain amount to train a person the NPs’, PAs’, and CNMs’ professional activi- to be an NP, for example, and this investment ties is constrained by their more limited training, yields a return of about $25,000 per year (plus 45

fringe benefits) minus what the person would have A substantial portion of these direct costs are earned otherwise. borne by taxpayers, rather than by the trainees. Society, through government support, has in- An alternative approach would be to consider vested heavily in the training of NPs as well as the costs and benefits of training someone to be physicians. For example, between 1975 and 1982, an NP, PA, or CNM instead of training the per- the Federal Government spent $65.9 million on son to be a physician. The costs to society of train- educating NPs. These funds supported approxi- ing an NP, PA, or CNM are much less than the mately half the NP training programs in the costs of training a physician. The direct costs re- United States (251). lated to education such as payments for instructors, supplies, and facilities, are greater for phy- The indirect costs—primarily foregone earn- sicians than for NPs, PAs, and CNMs, probably ings—are substantial, but they are difficult to esti- on a yearly as well as overall basis. The indirect mate with any precision. Because a physician costs, primarily what the individual would have spends about 6 more years in training than does earned during the time spent in training, are also an NP, the indirect costs an individual must pay greater for physicians, because more years of school- to become a physician are much greater. Deter- ing are required. mining the value of the foregone earnings for those individuals who become doctors versus Differences between the social and private rates those who become NPs is a more complex em- of return primarily reflect differences in the costs pirical task. Clearly, however, several NPs could of education. The more that government subsi- be trained for the cost of educating one physician. dizes training, the higher will be the private rate of return, compared with the social rate. Little evi- Extrapolating from’ CBO’s estimates of PA-train- dence exists as to what either rate of return is or ing costs (242), the total direct costs of training what the differential between the two is, but edu- a physician assistant would have been $16,900, cational subsidies over the years have been con- compared with $86,100 for training a physician siderable. Scheffler (217) provides an estimate of as of 1985. The indirect costs for PAs are about the private rate of return as of the early 1970s, the same as for NPs. Thus, the total costs of train- arguing that “. . . the private rate of return is ing are higher for PAs than for NPs, but the aver- sufficient to produce a relatively strong demand age earnings of PAs are higher than those of NPs for PA training; therefore, an increase in govern- ($24,500 versus $23,500) (44,237). Although, a ment support is unwarranted. ” He finds high rates more precise comparison would require some ad- of return—over 20 percent—comparable to those justment for the sex compositions of the two received by physicians. The available data are groups, the chief implication of the studies is that probably insufficient to allow distinctions between PAs, like NPs, are much less costly to train than these two types of investment, but thinking about physicians. them qualitatively is useful. Certified Nurse-Midwives Nurse Practitioners and Physician Assistants The tuition charges for nurse-midwifery edu- The most recent estimates of the costs of education vary considerably among programs, but cating physicians and NPs, PAs, and CNMs were an estimated average of the annual cost of edu- made in 1979 by the Congressional Budget Of- cating a nurse-midwifery student is approximately fice (CBO). CBO estimated the mean total costs $12,000 (78). The total cost of training is increas- of educating NPs and physicians at that time to ing with the growing trend toward master’s de- be $10,300 and $60,700, respectively. Assuming, gree programs, which last 2 years and are usually conservatively, that these costs increased at an twice as long as certificate programs. Approxi- average annual rate of 6 percent, the total educa- mately 40 percent of the Nation’s CNMs have tional costs would have been $14,600 for NPs and graduated from master’s degree programs. The $86,100 for physicians as of 1985. average total training cost for certificate and 46 master’s programs combined is about $16,800, question is whether the total value of the prac- compared to the $86,100 cost of physician train- tice output increases enough (i.e., would there be ing as of 1985. enough additional revenue) to cover the additional cost of the NP? Costs and Benefits of Private Denton and others (61) examined the effect of Employment of NPs, PAs, and CNMs the additional costs in a hypothetical calculation of the savings that would have resulted in Can- Because physicians or group practices some- ada in 1980 “had nurse practitioner time been sub- times must choose between hiring additional phy- stituted for physician time in the provision of all sicians and hiring NPs, PAs, or CNMs, the per- services for which such substitution has been dem- spective of the physician as employer should be onstrated to be safe and feasible. ” The research- considered in any attempt to understand the em- ers concluded that the savings from this widespread ployment levels of these nonphysicians. Using use of NPs would have been from 10 to 15 per- NPs, PAs, and CNMs to provide services that cent for all medical costs (or from $300 million would otherwise be provided by physicians can to $450 million) and that the savings would have benefit society with lower fees if the cost of pro- amounted to between 16 and 24 percent of the viding services by the nonphysicians is less than total costs for ambulatory care. Furthermore, the that of providing services by physicians and if the researchers determined that their “estimates are savings are passed on to patients. The costs of employing an NP, PA, or CNM include salary, fringe quite insensitive to demographic changes and will be as valid in the future as they are today. ” benefits, supervisory expenses, costs of any ex- pansion necessitated by adding another provider These findings are supported somewhat by the to the staff and costs of resources used by the ad- findings of Salkever and others (213), who com- ditional provider. These costs must be compared pared patterns of treatment for otitis media and with the costs that would be incurred if a physi- sore throat by three types of prepaid group prac- cian were added to the practice, The benefits a tices—NP only, NP-physician team, and physi- practice receives by hiring an additional provider cian only. With respect to otitis media, the find- are the additional fees the provider’s services gen- ings support the contention that NPs’ services are erate for the practice. less expensive than those of physicians. Services provided by NPs alone are less costly than those Nurse Practitioners provided by physicians alone or by NP-physician How employing a nurse practitioner would af- teams. The researchers found no difference, however, between the cost of treatments for otitis me- fect the cost of a practice cannot be determined dia by physicians alone and NP-physician teams. with any precision, but the following simple cal- The findings were similar for care of sore throats. culation provides a rough picture of the effect. These results confirm earlier studies (81,141) com- The median salary of NPs in clinical practice in paring the costs of specific medical tasks conducted 1983 was approximately $23,500. If fringe bene- by nurse practitioners with the costs of the same fits averaged 25 percent of salaries, total costs would be about $29,500 per year. This is far be- tasks conducted by physicians. low the $82,000 net income of young physicians Physician Assistants (19). Hiring a nurse practitioner or another physician might also result in indirect costs for such The average salary of a PA is $24,500 and fringe things as new office space, new equipment, ad- benefits probably amount to about 25 percent of ditional support staff, and additional resources. their salaries, making the average direct cost of employing a PA approximately $30,600 per year Total practice costs would change in composi- a sum much lower than the average income of tion because physicians would spend some time young primary-care physicians. supervising the NP instead of providing visits, or the NP might order more or fewer lab tests than Accurately estimating the relative cost of em- the physician would have. However, the basic ploying a PA versus that of employing a physi- .-, 4/ cian requires an examination of the indirect costs Dickstein (53) found that clinic prenatal and post- that result from the resources expended by the ad- partum costs in a large HMO were higher for ditional employees. Little information exists about CNMs than for obstetricians, “primarily because the extent of the costs PAs generate by using a midwifery visits are longer and more frequent, use practice’s resources. For example, Wright and others more RN educational time, and include the cost (266) found that PAs generate more laboratory of OB consultations and referrals. ” Generally, al- costs than medical residents but fewer than med- though existing data do not allow precise quan- ical faculty. The calculations that Denton and tification of the costs of CNM care and physician others (61) employed for determining that using care, the salary differential probably ensures that NPs would save 10 to 15 percent of the total cost the total costs are considerably less for CNMs for medical care in Canada could apply to using than for physicians. PAs, as well, because the researchers used the term nurse practitioner in a broad sense to encom- Costs Versus Benefits of Private Employment pass “several different types of intermediate health The private physician’s firm that employs an professionals. ” NP, PA, or CNM incurs extra costs for salary, fringe benefits, capital improvements, and other Certified Nurse-Midwives items. Productivity studies have shown that the time a physician spends supervising the NP, PA, The average salary of CNMs was $24,800 in or CNM reduces the number of patients the phy- 1983. If their fringe benefits were 25 percent of sician sees, although the reduction is more than their salaries, the average direct cost of employ- offset by the overall increases in practice volume ing a CNM was approximately $31,000 that year. generated by the additional provider. Studies have The mean net income of obstetricians in 1983 was not, however, directly addressed whether the $119,900 (before fringe benefits) but because most value of the additional output exceeds the addi- CNMs have been practicing fewer than 15 years, tional cost. In terms of rough magnitudes, the the most appropriate figure for comparison would comparison is between a $25,000 salary (plus be the average salary of young—rather than all– other costs) and a 20- to 50-percent increase in obstetricians. The average income of young ob- the practice’s revenues, from a base of $150,000 stetrician/gynecologists is $100,000 per year plus to $200,000 annually. In view of the uncertainty $25,000 or more for fringe benefits. about the extent to which an NP, PA, or CNM As with the other types of health-care provid- would increase marginal revenues, the marginal ers, the indirect costs a CNM generates by using revenues do not clearly exceed the marginal costs. a practice’s resources need to be calculated to de- But the careful accounting by Denton and others termine the full costs of employment. Evidence (61) in Canada suggests that significant savings exists that clients of CNMs have shorter hospital are possible for private practices that hire an NP, stays than do clients of obstetricians (53,65). But PA, or CNM rather than an additional physician.

CURRENT EMPLOYMENT: SETTINGS AND TRENDS

The productivity studies suggest that hiring Nurse Practitioners’ and NPs, PAs, and CNMs may provide private prac- Physician Assistants’ Employment tices a cost-effective alternative to hiring additional physicians. And although private markets Most of the pertinent studies have addressed may be functioning as expected under existing le- the employment of NPs and PAs in primary-care gal and market institutions, unexploited social settings, although NPs and PAs work at all levels benefits may be available from the greater em- of health care in a wide variety of settings (154). ployment of NPs, PAs, and CNMs. A 1982 national survey of pediatric NPs, for ex- 48 ample, revealed that 22 percent of the respondents ple, the proportion of PAs employed in hospitals worked in hospitals, 20 percent in community- grew from about 10 percent in 1974 to more than health agencies, 17 percent in private pediatricians’ 30 percent today. offices, 10 percent in specialty clinics, 8 percent Increasing numbers of NPs, as well as PAs, are in schools, 6 percent in HMOs, and the rest mainly finding work in hospitals. This development may in nursing schools and military clinics (167). not be due to the implementation of prospective NPs are increasingly being employed in home payment for hospitals based on diagnosis-related health agencies (155,196,220,268), and finding groups (DRGs) and, in fact, maybe occurring de- work in nursing homes (87,262). NPs are also spite DRGs. Instead, the trend is probably related working in industrial settings (216), correctional in part to the growth in the supply of physicians. institutions (104), and schools (156,228). As the number of physicians increases in cer- Different types of practice settings have differ- tain specialties, e.g., surgery, residency positions ent implications for any economic analysis of the are being decreased to contain the numbers and benefits of hiring NPs or PAs. For example, com- PAs [are being] employed as ‘junior house staff’ paring NPs with other nurses might be more ap- to supplement patient care (262). propriate than comparing NPs with physicians in New employment opportunities for NPs and PAs such settings as home health agencies, HMOs, may also stem from the trend for hospitals to schools, and businesses, where NPs might be em- establish community-based, ambulatory-care cen- ployed instead of, or in addition to, registered or ters in order to broaden their patient bases and licensed nurses. In these settings, the NPs—the to assure themselves of solid sources of inpatient more costly alternative—might be selected be- referrals. Hospital managers recognize that their cause they could provide a wider range of serv- best interests are served by providing these services. NPs employed in schools, for example, can ices as efficiently as possible and, consequently, serve as liaisons among the various health-care by employing NPs and PAs. providers serving schools; NPs can also provide backup support and in-house education to school nurses and provide educational services to teach- Certified Nurse= Midwives’ Employment ers, parents, and students (228). According to the 1982 ACNM survey, 36 per- Because of increases in the variety of settings cent of the Nation’s CNMs worked in hospitals, in which NPs work, their employment rates might 20 percent were in private practice with one or reasonably be expected to be higher than ever. more physicians, 14 percent were in private nurse- But, proportionately fewer NPs are working as midwifery practice, and the remainder worked in nurse practitioners in the 1980s than were doing public-health agencies, prepaid groups, and other so in the 1970s (237). The extent to which this de- settings (10). Nearly 35 percent of the respond- crease reflects increased competition from the grow- ents to this survey revealed that they were not ing supply of physicians is unknown. working as nurse-midwives, and about half of these said the reason was that “no nurse-midwif- PAs also work in a wide variety of settings and ery positions are available in my community. ” in every level of health care from primary to ter- tiary. Of all the Nation’s PAs, about one-third The data in table 2-3 indicate the changes that work in office-based practices (about half of these have taken place in how CNMs are distributed PAs work with physicians in solo practices); another among the types of organizations in which they one-third or so are based in hospitals; and the re- work. In general, the shift has been away from maining one-third work in prepaid groups, pub- employment in hospitals, public health depart- lic health departments, drug and alcohol rehabili- ments, and university health services and toward tation centers, industrial settings, nursing homes, private practice (9,10). In contrast to NPs and prisons and jails, and military facilities (45). Con- PAs, proportionately fewer CNMs practice in hos- siderable change has occurred in the proportion pitals now than did so in the 1970s: in 1984, only of PAs employed in various settings. For exam- 6.7 percent of the Nation’s hospitals had CNMs 49 on staff (171). More than 14 percent of the Na- 10 CNMs in private practice with physicians were tion’s CNMs worked in private nurse-midwifery supervised directly by physicians, whereas ap- practice in 1982, compared with 2.4 percent in proximately one-third of hospital-based CNMs 1976 to 1977 (9,10). were under the supervision of physicians. Almost half the CNMs in private nurse-midwifery prac- CNMs are finding increased employment where tice were not administratively responsible to any- they are not administratively responsible to phy- one other than themselves, and an additional 22 sicians. Administrative independence must not be percent reported to other nurse-midwives. In all, confused with clinical independence, because nearly 36 percent of the respondents noted that CNMs do not aspire to clinical independence. they were supervised directly by other CNMs (10). They highly value their professional interdepen- dence and collaboration with physicians (13). The evidence suggests that CNMs–-especially those in private nurse-midwifery practice—tend Although most NPs and PAs in primary care to function organizationally more independently are supervised directly by physicians, only 48 per- of physicians than do NPs or PAs. Because of the cent of the CNMs practicing in the United States sixfold increase in the percentage of CNMs work- who responded to the 1982 ACNM survey indi- ing in private nurse-midwifery practices between cated that their immediate supervisors were phy- 1976-77 and 1982, the organizational independ- sicians. All the responding CNMs, however, col- ence of CNMs has increased markedly. This trend laborated on clinical matters with physicians (10). shows no signs of slowing down, although all The proportion varied considerably depending on obstetrics-related care may be decreased by the the type of practice. For example, about 9 of every liability-insurance crisis.

SUMMARY

Studies show that NPs, PAs, and CNMs can have not grown and diffused to a greater extent. provide services that both substitute for and com- Although the private markets for NPs, PAs, and plement physicians’ services, depending on the CNMs as employees in physicians’ practices do particular service or type of practice. Moreover, not suggest a current shortage, the removal of hiring an NP, PA, or CNM increases a practice’s payment barriers and limitations could greatly in- total output and costs less than employing an ad- crease the demand for these alternative practi- ditional physician. Because training is less costly tioners. Unless the barriers are altered, the poten- for these practitioners than for physicians, using tial savings from a greater use of NPs, PAs, and NPs, PAs, and CNMs rather than physicians for CNMs will probably remain unexploited. certain services would presumably be cost-effec- Continuing research and analysis is needed to tive from a societal point of view, given that the ascertain the cost savings that would result from quality of care is equivalent to that provided by increased employment of NPs, PAs, and CNMs. physicians for comparable services (see ch. 2). Al- Many productivity studies have been conducted, though additional cost savings might result from but few attempts have been made to compare how greater employment of these providers, the evi- NPs, PAs, or CNMs affect the revenues of indi- dence suggests that current employment levels and vidual practices with how they affect the prac- practices more or less reflect existing market con- tices’ costs. Changing market circumstances cre- ditions. ate a need for both types of studies, but those that The abilities and cost-effectiveness of NPs, PAs, compare revenues and costs are especially im- and CNMs raise a question as to why their ranks portant. ------

Chapter 5 Payment Issues Chapter 5 Payment Issues

In their areas of expertise, nurse practitioners apply even if payment were indirect, i.e., to the (NPs), physician assistants (PAs), and certified NPs’, PAs’, or CNMs’ employer. ] These modifi- nurse-midwives (CNMs) can provide safe care cations would have several implications for em- that meets generally recognized standards of qual- ployment and the scope of practice of these prac- ity, care that emphasizes personal and preventive titioners2 and for the costs borne by third parties, dimensions often underemphasized by physicians, patients, and society. and care that would otherwise be unavailable in Some Federal health programs and private in- inner cities, remote areas, and certain settings surers provide coverage and direct payment for where demand or ability to pay are insufficient the services of NPs, PAs, and CNMs in some set- to support physicians’ practices. NPs, PAs, and tings (see app. B). For purposes of analysis, this CNMs could also reduce costs in certain settings. case study assumes that coverage and direct pay- Nonetheless, professional attitudes and restric- ment for such services would be offered by all the tive statutes, regulations, and policies have hin- programs and insurers and that any new Federal dered the ability of NPs, PAs, and CNMs to ob- legislation would not override State laws or reg- tain employment in some settings and to practice ulations governing the licensing and practice of at levels commensurate with their training (see box NPs, PAs, and CNMs. l-A). One major constraint is that many third- The effect of the modifications would vary, de- party payers, including many Federal programs, pending on the setting in which the provider prac- do not cover (authorize payment for) services pro- ticed and on the method of payment. Because vided by NPs, PAs, and CNMs in certain settings, these two factors are interdependent—in that pay- if the services are typically and characteristically ment method is usually typical of a type of prac- provided by physicians nor do they pay them tice setting—they are considered together. directly for such services (see app. B). Although most third-party payers usually do not look be- The effect of these modifications also depends yond a physician’s claim for payment as to whether on the health-care environment, which is chang- the physician or NP, PA, or CNM have provided ing. The supply of physicians and the organiza- a particular service, uncertainties about coverage tion and financing of health care are changing in are partly responsible for some physicians’ reluc- ways that are likely to bring about a more com- tance to hire NPs, PAs, or CNMs. Lack of direct petitive market for health-care services.3 These payment limits the independent practice of NPs trends have implications for the future of NPs, and CNMs. Third-party payers have been more generous in covering and directly paying for the ] During the publication of this case study, the Omnibus Recon- services of CNMs than NPs. Although PAs, as ciliation Act of 1986 (Public Law 99-509) was enacted. The act mod- well as NPs and CNMs, have actively sought cov- ifies Medicare and authorizes payment for (covers) services of phy- erage for their services, they differ from NPs and sician assistants working under the supervision of physicians in hospitals, skilled nursing facilities, intermediate-care facilities, and CNMs in not wanting direct payment. as an assistant at surgery. The payment is indirect and at levels lower than physicians would receive for providing comparable services. Observers have suggested modifying the cur- 2Many other factors affect the employment and practice patterns rent rules for payment of such services by requir- of NPs, PAs, and CNMs. Several issues, especially malpractice in- ing coverage for NP, PA, and CNM services and surance, are critical, but a discussion of them would be beyond the scope of this case study. by paying NPs and CNMs directly and not through ‘The fact that the U.S. population is aging and consequently need- the employing physician. Requiring coverage would ing more health-care services would also affect the employment of NPs and PAs and, to the extent that the provide gynecological serv- be both an independent modification and a pre- y ices, CNMs. The aging of the population has been discussed in de- liminary step toward direct payment. A third tail in a number of previous OTA reports, notably in Technology modification —establishing a payment level—could and Aging in America (245).

53 54

PAs, and CNMs, regardless of whether payment ing health-care environment, however, would cer- for their services changes. Modifying payment for tainly affect their employment and use and might the services of NPs, PAs, and CNMs in a chang- alter the costs of health care.

EFFECTS OF MODIFYING PAYMENT FOR SERVICES OF NURSE PRACTITIONERS, PHYSICIAN ASSISTANTS, AND CERTIFIED NURSE-MIDWIVES

Modifying the method of payment could be ex- joint statement, because it permits interdependent pected to have varying effects on the employment practice without calling for physicians to be pres- and scope of practice of NPs, PAs, and CNMs, ent whenever CNMs are caring for patients (13). depending on whether they were in independent In addition, the American College of Nurse Mid- practices or worked in physicians’ practices, health wives requires that CNMs agree to work in clini- maintenance organizations, hospitals, nursing cal collaboration with physicians in order to ob- homes, or other settings. Modifying the method tain certification. of payment might also affect costs. In addition to professional restraints, State laws and regulations that limit NPs’ and CNMs’ scope Effects on Independent Practices of of practice and specify requirements for supervi- Nurse Practitioners and sion by physicians serve as a formal control on Certified Nurse-Midwives clinical independence. NPs and CNMs in independent practice are also accountable for their Mandated coverage and direct payment to NPs mode of practice by the malpractice insurance and CNMs for providing services typically and they carry. characteristically performed by physicians would dramatically increase NPs’ and CNMs’ ability to Although a few NPs have attempted to estab- establish fee-for-service practices that were ad- lish administratively independent practices, most ministratively independent from physicians. In- NPs in such practices provide traditional nursing deed, direct payment would be the most advan- care rather than primary medical care (138), Among tageous payment method for NPs or CNMs in the barriers NPs face in undertaking independent independent practices. As autonomous providers, practices are the necessity of making substantial NPs and CNMs could provide the full range of financial investments and the lack of coverage and services for which they were trained and licensed. direct reimbursement for their services. The Amer- ican Nurses Association (ANA) believes that Such practices would be administratively inde- many NPs would establish such practices if cov- pendent but according to current modes of prac- erage and direct payment were more widely avail- clinically tice, they would not be independent able (256). from physicians when NPs and CNMs were performing delegated medical tasks. ’ The nursing CNMs are highly interested in administratively profession has agreed to clinical collaboration. For independent practice. Indeed, the proportion of example, a joint statement of “practice relation- CNMs in private midwifery practices increased ships” calls for obstetrician/gynecologists and from 2.4 percent in 1976 to 1977, to 14 percent CNMs to adhere to clinical-practice arrangements in 1982 (9,10). During that period, the number that include the participation and involvement of of third-party payers that provided coverage and obstetrician/gynecologists with CNMs as mutu- direct payment for CNMs’ services increased. If ally agreed on in written medical guidelines or additional third-party payers were to cover and protocols. CNMs in administratively independ- pay for these services, more CNMs probably would ent practice believe that they are adhering to the be interested in independent practices ‘Problems with obtaining malpractice insurance coverage and high 4NPs and CF/Nls may legally be c1 in ical 1 y independent from phy malpractice premium costs are significant limitations on independ- sicians when performing nursing tasks. ent practice by CNMs. 55

How coverage and direct payment for NPs’ Patients’ costs might be lower if the NPs and services would affect the establishment of adminis- CNMs charged their patients lower fees than phy- tratively independent fee-for-service practices by sicians charged for comparable services. For most NPs partly depends on the extent to which NPs primary care services, e.g., office visits, savings seek and obtain direct payment. The impetus for to most patients would be small, because fees for direct third-party payment of nurses, an ANA pri- such services are not high and third-party pay- ority since 1948, increased for organized nursing ments cover a large part of them. Savings for with the establishment of NPs as health practi- maternity care could be appreciable however, be- tioners (22). Indeed, the ANA has been actively cause charges and patient liability for such serv- involved in seeking and sometimes obtaining such ices are high. Coverage and direct payment would payment at the State and national levels (23,232). allow patients to choose NPs and CNMs as providers without being penalized financially by lack Little information is available as to how many of reimbursement. practicing NPs receive direct payment. A 1983 survey of NPs, conducted 4 years after the pas- Any savings to third parties and patients might sage of a Maryland law providing direct third- be decreased or negated by duplicative visits. Pa- party payment for services not directly supervised tients who sought care from NPs or CNMs in in- by physicians, found that fewer than 1 percent dependent practices might also see physicians for were paid directly (99). In 1986, however, 7 years the same or related care, on their own initiative after the passage of similar legislation in Oregon, or on referral by NPs or CNMs. Seeing both phy- a survey of NPs in that State found that 25 per- sicians and nonphysicians could result in dupli- cent were receiving direct third-party payment; cation of examination and laboratory procedures. 42 percent had been issued provider numbers; and Although NPs and CNMs in independent prac- 38 percent were signing the claims forms for the tices could lower societal costs for health care, the services they provided (102). The researcher who extent of the savings is difficult to estimate. So- conducted both surveys suggests that the disparate cietal costs would reflect, among other things, any findings might reflect the fact that more time had decreases in program costs and beneficiary costs elapsed between the passage of the legislation and and any savings resulting from NPs’ and CNMs’ the survey in Oregon than had elapsed in Mary- care that reduced the need for care in the future. land (101). For example, although CNMs might not find it feasible to charge patients lower fees than physi- The establishment of independent fee-for-service practices by NPs and CNMs could affect the cians charge (because CNMs spend so much more time with patients than physicians spend), CNMs costs of third-party payers. If the total volume might lower societal costs by decreasing the need of services by all providers did not increase, set- for expensive neonatal intensive care for infants ting payment levels for services provided by NPs and CNMs lower than levels for comparable serv- of women whose socioeconomic status puts them (193). ices provided by physicians might decrease the and their infants at high risk costs of third-party payers. Of course, the size of Scant evidence is available as to how much NPs any savings to third-party payers would depend in independent practices charge their patients. In on the size of the gap between payment levels for an exploratory phase of a survey of Maryland physicians and payment levels for NPs and CNMs. NPs, Griffith (99) found that the median fees Paying NPs and CNMs 10 percent less than phy- charged by NPs in independent practice were lower sicians are paid would have a minimal effect on than the median fees charged by physicians for third-party costs in the immediate future, in part most services. However, 59 percent of NPs’ fees because the number of NPs and CNMs is so much were the same as physicians’ fees for all types of smaller than the number of physicians. Savings visits (99). Charging lower fees than physicians to third-party payers would also depend on the charge for similar services appears to be the norm extent to which patients chose to patronize NPs for NPs in many types of settings other than in- and CNMs in independent practices. dependent practice. Brooks (36) reported that the 56 fees charged by NPs in rural satellite settings are Charges for CNM services in independent prac- lower than those charged by a sample of rural tice appear to vary by region—in some areas their physicians. Several national studies of NPs in fees are lower than those of physicians, and in organized settings confirm this finding (256). Pa- other areas they are about the same (79). CNMs tients were generally charged less for visits to Ore- charge slightly less than obstetricians for normal gon NPs who received direct payment either in maternity care (98) when services are provided independent practices or in physicians’ fee-for- in independent birthing centers (103,149). The to- service practices than for visits to salaried NPs tal costs of maternity care by CNMs may also be (102). The difference between the charges for short less than total costs for care by physicians for sim- initial visits and brief followup visits was statis- ilar cases, not necessarily because CNMs have tically significant. Furthermore, charges for visits lower fees, but because the care they provide is to NPs were lower than for visits to physicians usually technologically less complex than physi- in both Oregon and Maryland. The difference be- cian care (98,201). tween charges for NPs and those for physicians Costs to patients, third-party payers, and so- was greater in Oregon than in Maryland, perhaps ciety would also be influenced by changes in the because the proportion of NPs receiving direct volume of services provided as a result of cover- payment was greater in Oregon than in Maryland (102). age and direct reimbursement for new providers. Historically, insurance companies have contended Whether NPs would increase their fees if they that covering and directly paying additional pro- were in independent practice and received direct vider groups in fee-for-service settings increases payment is unclear, although some evidence in- the volume of services provided by the new pro- dicates that other groups that provide services viders, the physicians, or both and, consequently, typically provided by physicians have gradually increases costs for third-party payers, benefici- increased their fees to the level of physicians’ fees aries, and society. The evidence to prove or re- after receiving direct payment. The American Psy- fute this argument is equivocal (246). The recent chiatric Association (APA) has reported two emphasis that public and private third-party payers studies that found this phenomenon to be true of have placed on monitoring the volume of health- psychologists and clinical social workers (256). care services may help to control potential increases in volume. Some private insurers report that their total Direct evidence is unavailable as to how cov- costs from CNMs for maternity care are lower erage and direct payment would affect the volume than those from physicians. Of course, physicians’ of services provided by NPs and CNMs. Indirect care includes care for complex cases that require information, which consists only of anecdotal more resources than normal maternity care. How- reports of private insurers’ experiences with other ever, Mutual of Omaha has noted that CNMs groups, is conflicting. Mutual of Omaha and other provide a “valuable service at a reduction in costs insurers report that chiropractors increased their from that charged by medical doctors or osteo- provision of services to consumers after being au- paths, ” and the Blue Cross and Blue Shield Asso- thorized for direct reimbursement but that psychi- ciation found that CNMs were less costly than atric social workers did not increase theirs (256). physicians in normal maternity care (256), Indeed, based on the current status of direct payment for Whether coverage and direct payment for serv- services, insurers of CNMs appear to be less resis- ices by NPs and CNMs would increase the pro- tant to coverage and direct payment than do in- vision of services by physicians is unclear. Phy- surers of NPs (see table B-l). Insurers, such as sicians might change their behavior in response Mutual of Omaha and Blue Cross, perceive that to competitive providers. If NPs and CNMs charged NPs would provide services in addition to those their patients lower fees, some physicians might normally provided by a physician, whereas CNMs decrease their fees in order to compete but, to provide services that substitute for physicians’ maintain their incomes, might increase the num- services (256). ber of services they provided to their patients (in- 57 ducing demand for services). Although research sicians. If the payment levels set for NPs’, PAs’, on physicians’ influence on the volume of serv- or CNMs’ services were lower than those set for ices has been conducted for many years, none of physicians’ services, the costs to third-party payers the studies positively proves the magnitude or would be lower if NPs, PAs, or CNMs, rather even the existence of induced demand for serv- than physicians, were employed. ’ ices (246). In the past, however, physicians in the However, authorizing payment for NPs’, PAs’, United States and Canada have maintained their and CNMs’ services would not necessarily increase income level even with substantial increases in the the opportunities for these providers to become supply of physicians (28). salaried employees in physicians’ practices. Alle- gations have been made that many physicians’ Effects on Physicians’ Practices practices, knowingly or unknowingly, submit bills In the 1970s, a major reason cited by physicians under the physicians’ provider numbers for un- as a disincentive to employing NPs, PAs, and covered NPs’, PAs’, and CNMs’ services. The bills CNMs was that Federal payment policies did not are seldom challenged by third-party payers. If authorize payment for services provided by NPs, the payment levels were the same for the serv- PAs, and CNMs (138). Whether mandating cov- ices of NPs, PAs, and CNMs as for the employ- erage for such services would increase incentives ing physicians, coverage of NPs’, PAs’, and CNMs’ for physicians in fee-for-service practices to em- services would not affect the revenues of physi- ploy these practitioners and delegate more serv- cians’ practices that were already billing for such ices to them depends on several factors, includ- services. In these practices, coverage probably ing physicians’ billing practices and the payment would affect neither the employment opportuni- levels for NPs’, PAs’, and CNMs’ services. The ties for NPs, PAs, and CNMs nor the services higher the payment level, the greater the mone- physicians delegated to such practitioners. tary incentive a physician would have to employ The revenues of these practices would decrease, an NP, PA, or CNM, but simultaneously the cost- however, if the payment levels were significantly saving potential to the third-party payer would lower for NPs’, PAs’, and CNMs’ services than decline. for physicians’ services, if the volumes of serv- Providing coverage and payment for the services remained the same for the practices, and if ices of NPs, PAs, and CNMs (at any level) would the physicians billed for the services of NPs, PAs, increase practice incomes for physicians who have or CNMs under the NPs’, PAs’, or CNMs’ pro- employed these practitioners without billing for vider numbers. How physicians would respond their services. Such physicians might increase the to decreases in their practices’ revenues is unclear, range of services they delegate to NPs, PAs, and but employment opportunities for NPs, PAs, and CNMs. Third-party payers’ costs would probably CNMs might be jeopardized. The physicians increase, regardless of whether the practices’ vol- might increase the volumes of services provided umes of services increased. Whether increases in by their practices. practice income would be passed on to patients Coverage of NPs’, PAs’, and CNMs’ services in the form of lower fees is unclear. would not affect third-party costs if the number If services by NPs, PAs, and CNMs were au- of services provided by practices remained sta- thorized for payment, physicians’ practices that ble; i.e., if the practices had billed for services un- currently do not employ such practitioners might der the physicians’ provider numbers before cov- be more inclined to employ them rather than hire erage was expanded, and if the payment levels additional primary-care physicians. If the pay- were the same for NPs, PAs, and CNMs as for ment level was 100 percent of what a physician the employing physicians. If the payment levels would receive for providing a comparable serv- were lower for NPs, PAs, and CNMs than for ice, third-party payers probably would incur higher ‘It is not clear whether or not NPs would accept payment levels costs for such practices regardless of whether the lower than those of physicians. As noted earlier, PAs are willing new employees were NPs, PAs, CNMs, or phy- to accept levels of compensation lower than those of physicians. 58 physicians, third-party payers’ costs for such prac- It is to their [HMOs] financial advantage to tices might decrease. For physicians’ practices, as produce services with the most efficient combi- for NPs’ and CNMs’ independent practices, the nation of inputs, substituting lower priced phy- size of the difference between the payment levels sician extenders for higher priced physicians when- for services provided by NPs, PAs, and CNMs ever possible (138). and for comparable services provided by physi- Furthermore, past experience with HMOs has shown cians would partly determine how lowering the that: payment level would affect the costs of third-party 7 payers. . . . capitation plans do care for [non-Medicare] enrollees at lower costs while maintaining qual- Because data do not exist as to how physicians ity at levels equal to or better than comparison bill for the services of NPs, PAs, and CNMs, the practices (246). overall effect that required coverage would have on NPs’, PAs’, and CNMs’ employment oppor- Effects on Hospitals tunities in physicians’ fee-for-service practices is Payment for services delivered in inpatient hos- uncertain. Coverage might influence employment pital settings by NPs, PAs, and CNMs who are indirectly. NPs have argued that coverage estab- hospital employees is most commonly made ei- lishes a collegial professional relationship. Further- ther retrospectively on the basis of cost or pro- more, they claim that coverage can cause physi- spectively on the basis of diagnosis-related groups cians to see that NPs’, PAs’, and CNMs’ services (DRGs). There is no statutory permission or lack generate revenue as well as costs (98). This per- of permission under Medicare or Medicaid for spective might increase the employment potential payment of NPs’, PAs’, and CNMs’ services as of these practitioners (98). inpatient hospital services when the providers are Direct payment would only indirectly affect the employed by the hospitals. Most other third-party employment of NPs and CNMs as salaried em- payers are also silent on this issue. Moreover, hos- ployees of physicians. Direct payment would al- pitals usually pay a salary to NPs, PAs, and low NPs and CNMs to choose to work as salaried CNMs that they employ. employees, to undertake independent practices, Medicare, Medicaid, and most other third-party or to enter into joint practices with physicians payers pay hospitals for total operating costs, and (i.e., partnership arrangements by NPs or CNMs most hospitals’ accounting systems simply lump with physicians). Paying NPs in physicians’ prac- the costs of NPs’, PAs’, and CNMs’ services to- tices directly, rather than indirectly, could be ex- gether with other types of operating costs. Nurses pected to decrease the fees for patients’ visits to contend that coverage and direct payment as well NPs (102). as the identification of the services that coverage and direct payment would require, would influ- Effects on Health Maintenance ence hospitals interest in them as employees. De- Organizations lineating the costs of these services might facilitate internal management decisions. Nurses have Because most third-party payers in the public advocated the identification of the costs of nurs- and private sectors currently provide coverage for ing services in institutional settings, believing that the services of these practitioners in health main- identification would increase nurses’ autonomy, tenance organizations (HMOs) (see table l-l), ex- encourage economic decisionmaking, enhance tending coverage is largely irrelevant to their em- nursing efficiency, and spur hospital administra- ployment in this setting. Also, most HMOs pay tors to recognize that nurses generate revenue as NPs, PAs, and CNMs a direct salary, which makes the issue of direct payment of little importance ‘Cavitation is a method of paying for medical care, in which a per capita amount is paid prospectively for all services received by in the HMO setting. an enrollee or beneficiary during a given period of time. The payment is not related to the quantity of service provided. Cavitation The data suggest that NPs, PAs, and CNMs payment provides financial incentives to use resources more effi- save costs for HMOs: ciently and even to underuse services. 59

well as costs (22,98,162). Nurses believe that recognition of their revenue-producing abilities could increase their employment opportunities in hospitals (161).

Extending coverage and direct payment for the services of NPs, PAs, and CNMs as hospital employees in the inpatient hospital setting most likely would require that the costs of the services be paid for as professional services, the category under which Medicare and other third-party payers currently pay for physicians’ services. Such a move would run counter to most current thinking, es- poused in both the public and private sectors, which is focused on containing costs by aggregating services. For example, some observers have expressed interest in aggregating physician services by adapting the DRG approach,8 particularly for hospital-based physicians (63,165). The Om- nibus Reconciliation Act of 1986 (Public Law 99- Photo credit: Geisinger Medical Center and the 509), however, has extended direct payment for American Academy of Physician Assistants anesthetic services rendered by certified registered PAs provide post-operative care as well as pre-operative nurse anesthetists in hospitals. These services were care and assisting in performing surgical procedures. originally to be paid for under Medicare as a component of a DRG but were passed through as a expressed concern that the lack of coverage has hospital cost. restricted PAs’ employment and the delegation of appropriate services to PAs at surgery. Using PAs Coverage of their services would affect the em- rather than physicians as surgical assistants re- ployment of PAs who are employees of physicians duces practices’ costs, but whether the savings are or physicians’ practices but who work as surgi- passed on to patients is unclear. cal assistants in hospitals.9 PAs assist in performing surgical procedures and also provide preoper- Effects on Nursing Homes ative and postoperative care (7). Medicare does not cover PAs’ provision of such procedures and Because virtually all NPs and PAs working in care, although Medicare currently covers and nursing homes are salaried employees, their em- pays at amounts equivalent to 20 percent of the ployment would not be necessarily affected by surgeons’ fees for the services of physicians who coverage of their provision of services typically act as assistants at surgery. Some observers have provided by physicians .’” With coverage, NPs and PAs could supply primary-care services in nurs-

8 ing homes as employees of physicians’ practices Under the DRG approach, Medicare pays a fixed amount for the operating costs associated with treating patients in each diagnostic or as team members in group practices provid- category. In applying the DRG approach to physicians, the payment unit would be a bundle of services rather than an individual —. ——— service. This approach could control both costs and utilization by IOSevera] other Medicare and Medicaid regulations specific to nurs- reducing the number of service units billed and encouraging the ju- ing homes limit the role of NPs and PAs and specify services that dicious use of services within packages. must be performed by physicians in order for the nursing homes’ 9During the publication of this case study, the Omnibus Recon- services to be covered (see app. B). Many States have passed laws ciliation Act (Public Law 99-509) was enacted. The act modifies to “permit the delegation of these services by a physician to a phy- Medicare and authorizes coverage of a physician assistant services sician assistant or nurse practitioner” (116). However, strict inter- furnished under the supervision of a physician as an assistant at sur- pretation of these and similar rules prohibits the appropriate use gery. The payment to the employer will be 65 percent of the rea- of NPs and PAs in nursing homes. In addition to permitting cover- sonable charge for a physician when acting as an assistant at sur- age under Medicare and Medicaid, amendments to these regulations gery and will be effective after Jan. 1, 1987. would be required in order for NPs and PAs to be used appropriately. 60

ing visits to nursing homes.11 If NPs were paid Except when more intensive care can be sub- directly, they could supply primary-care services stantiated, the number of physician visits to nurs- to nursing homes as independent practitioners, ing homes is limited under the Medicare program. similar to physical therapists. Extending coverage, therefore, might not increase the costs attributable to nursing-home visits for Many nursing homes have difficulty supplying third-party payers, assuming payment levels were primary-care services because few physicians are the same, or lower, for the NPs and PAs as for interested in visiting patients in nursing homes to the physicians. When physician-NP teams, rather (166). provide services Furthermore, most phy- than physicians alone, visited nursing homes, sicians are poorly prepared to care for seriously however, total costs to third-party payers were ill elderly patients. The growing number of elderly shown to decrease, mainly because of lower rates people in our society, particularly those over 85 of hospitalization and fewer visits to physicians who most frequently need nursing-home care, has or clinics (128). A 1980 and 1982 study found that, increased concerns about the quality and costs of as compared with physicians alone, a group prac- such care. Many residents are medically stable but tice of salaried physicians, NPs, and PAs showed functionally impaired by chronic physical or men- substantially lower overall medical costs for nurs- tal conditions. Other residents are admitted from ing home residents even though the number of hospitals for recuperation and rehabilitation fol- visits to the homes were not limited. Savings were lowing surgery, or are terminally ill and do not realized from decreases in expensive hospital- (245). require hospital care NPs and PAs are based emergency and outpatient services and in uniquely suited to provide the types of care needed the numbers of hospital days used (155,257). Fur- by nursing home residents with chronic conditions thermore, the quality of care increased, and the and their associated disabilities (see chs. 2 and 3). NPs acted as patients’ advocates. 1 IDU~i~~ the Publication of this case study, the Medicare Iafi was Although payment changes are a necessary changed as a result of the enactment of the Omnibus Reconcilia- tion Act of 1986 (Public Law 99-509) during October 1986. The act step, innovative approaches to improving the care authorizes the coverage of the services of PAs furnished under the and reducing the costs associated with nursing supervision of a physician in skilled nursing facilities and interme- homes need to include modifications of regula- diate-care facilities in States where the physician assistant is legally authorized to perform the services. The payment to the employer tions concerning visit limitations and changes in is to be at 85 percent of the prevailing charge of physician services other Medicare and Medicaid regulations that for comparable services provided by a nonspecialist physician. limit the role of NPs and PAs in nursing homes.

THE CHANGING CONTEXT OF HEALTH CARE Financing riod. The health-care organization receives its payment, the amount of which is not related to A growing trend is to set payment rates for the quantity of services provided, and must then health services before, rather than after, they are pay physicians and other providers. Cavitation delivered. Prospective payment has been adopted payment provides financial incentives to prevent in response to rapidly rising health-care costs and high-cost problems and to deliver services at low the recognition that cost increases have been partly cost. Acceptable standards of care, or at least pa- caused by retrospective reimbursement. One of tient satisfaction, are essential if capitated plans the most innovative approaches is Medicare’s are to maintain enrollment at sufficiently high method of paying for beneficiaries’ inpatient care levels to maintain financial viability (246). on the basis of DRGs. The other major trend is increased interest in Supply of Physicians the use of cavitation, in which a per capita amount is set prospectively for all medical services received In the mid-1960s, public policy in the United by an enrollee or beneficiary during a given pe- States began to focus on counteracting the shortage and maldistribution of physicians. As a re- 88,290 in 1980 to 140,213 in 1984 (4). Some phy- sult, the number of medical schools increased sicians join group practices because the practices from 89 in 1965 to 127 in 1984 (255), and the num- are established, they entail less financial risk than ber of first-year medical students nearly doubled solo practices, and they provide access to the cap- (240,255). Expected increases in the numbers of ital required for purchasing and using sophisti- graduates from U.S. medical schools, combined cated medical technology (16). Group practices with graduates of foreign medical schools, are re- may be even more attractive to physicians in the sulting in physician surpluses, which the Gradu- future for a number of reasons including the cap- ate Medical Education National Advisory Com- ital required to purchase expensive technology and mittee predicts will be significant by 1990. Since increased competition. 1982, enrollment in medical schools has declined The types of organizations in which physicians slightly, as the Federal Government has reduced practice—with or without other health-care pro- both its funding of subsidized loans for medical viders—have also increased. HMOs have been students and its support of medical schools (58). growing rapidly in recent years. Enrollment in The growth rate in the supply of foreign medical HMOs grew by 25.7 percent in 1985 to a total en- graduates also is expected to decrease (255), but rollment of 21 million (123). Although Individ- the effect of past efforts to increase the supply of ual Practice Association (IPA) models outnumbered physicians will be felt well into the next century. all other kinds of HMOs combined, group-model Observers expect increases in the number of plans retained the lead in enrollment (123). That physicians to significantly outpace population enrollment is expected to increase rapidly in the growth. For every 100,000 people in the United next 5 years. Estimates of total enrollment in States, there were 148 physicians in 1970 and 218 HMOs range between 25 and 50 million for 1990 in 1983 (255). Estimates for 1990 range from 215 (241). Part of the growth in HMOs has been at- (240) to 224.4 (255) per 100,000. Estimates for the tributed to the increased willingness of physicians year 2000 range from 240 (240) to 245.2 (255) per to be employed in them (240). Recent changes that 100,000. ’2 From 1981 levels, the numbers of phy- might affect the employment and use of NPs, PAs, sicians in primary-care specialties, including ob- and CNMs in HMOs are the increasing involve- stetrics and gynecology, are expected to have in- ment of for-profit corporations in HMOs, and the creased 28 percent by 1990 and 53 percent by joint purchasing and other cost-saving ventures 2000, outpacing the growth in the total supply of undertaken by groups of HMOs (246). (255). physicians Although the need for physicians Preferred-provider organizations (PPOs) in- is expected to increase, the supply of physicians clude several types of arrangements between third- is expected to exceed the need by 1990, accord- party payers and health-care providers, including to all estimates (94,240,251,255). ing physicians, hospitals, or both. In these arrangements, providers contract with insurers or Delivery Sites and Organizations employers to deliver care at reduced prices. The 1978; 1985, In 1983, for the first time, the main practice ar- first PPO was organized in by June 334 had been organized and 229 were operating rangement of less than half (48.9 percent) of all (118). physicians in the United States was solo practice. Although PPOs were designed to reduce Only 8 years previously, more than 54 percent expenditures, no evidence currently exists that the of the Nation’s physicians practiced individually. care they deliver costs less than that delivered by In 1984, the number of group practices (three or other types of organizations. more physicians) was over 15,000—up 44 percent The delivery of health services is also affected since 1980 (16). The number of physicians in group by the growth of the multihospital system—two practices during the same period increased from or more hospitals owned, leased, controlled, or managed by a single for-profit or not-for-profit “The total number of physicians in 1970 was 334,028 and in 1983 corporation. Indeed, the multihospital system has was 519,546 (255 ). Estimates for 1990 range from 537,750 (240) to 555,300 physicians (255). Estimates for 2000 range from 642,950 (240) become an important component in the chang- to b55, Q20 physicians (255), ing health-care-delivery system. Some 35 percent 62 of the Nation’s hospitals and 38 percent of all com- Effects of Changes in the Health-Care munity hospital beds are now in multihospital sys- Environment on Nurse Practitioners, tems (14). Since 1976, the number of multihospi- Physician Assistants, and 60 tal systems has increased by more than percent Certified Nurse= Midwives (2). A few observers believe that the growth of the for-profit component will eventually result in How changes in the health-care environment most services being provided by a few nationwide will affect the integration of NPs, PAs, and CNMs suppliers that might appropriately be labeled in the health-care system is unclear. The changes, “megacorporate health care delivery systems” which generally reflect trends toward cost-con- (85). tainment and increased competition, are inter- Another trend is toward increasingly diverse dependent. For example, the increasing supply of sites for providing care (see table 5-1) .13 For ex- physicians has heightened competition among ample, the first free-standing center was estab- medical-care providers (19,176,205,206), leading lished in Delaware in 1973. By July 1984, there many young physicians to accept salaried posi- were an estimated 1,800 such centers in the United tions and to enter into contractual arrangements States and the total is projected to grow to ap- with third-party payers (19,240). The number of proximately 4,500 by 1990 (152). In late 1983, physicians in salaried positions is twice as great about 9 percent of the Nation’s physicians worked for those in practice 5 years or less as for those an average of about 13 hours per week in free- in practice 6 years or more (18). In effect, the in- standing centers providing primary or emergency creasing supply of physicians is an important fac- care. Some of these centers were operated by hos- tor in changing medical practice arrangements in pitals or chains and others operated independently the United States and in fostering a willingness (16). to practice in fee-for-service groups and in capitated and institutional settings, which many phy-

1315ee Medjca] Technology and Costs of the Medicare program sicians avoided only a few years ago. (244) for a more detailed description of alternative sites of care. Competition in the health-care system could either limit or expand employment opportunities Table 5-1 .—Selected Alternatives to Traditional for NPs, PAs, and CNMs. Competition resulting Health-Care Delivery from the growing supply of medical-care providers might reduce such opportunities, especially in 1. Alternative sites: physicians’ office-based, fee-for-service practices. Alcohol and drug abuse centers Ambulatory care centers Physicians with declining patient bases might not Ambulatory surgical centers have enough patients to justify employing addi- Birthing centers tional providers (97). However, the American Diagnostic imaging centers Freestanding emergency centers Medical Association (15) notes that, faced with Hospices increasing competition, rising practice costs, and Mammography centers cost-conscious patients, physicians are concerned Nurse-managed centers Nutritional dietary centers about the cost-effectiveness of their practices and Oncology centers might attempt to improve the practices’ produc- Pain management centers tivity and increase the practices’ income by em- Psychiatric centers Rehabilitation centers ploying NPs, PAs, and CNMs. Compared with Sports rehabilitation centers practices that do not employ NPs and PAs, phy- Student health centers sicians’ practices that do employ NPs and PAs Wellness programs have higher numbers of patient visits per hour and 11. Alternative organizations: Competitive medical plans per week and higher incomes for the employing Extensive provider organizations physicians (17). Because such practices charge Health maintenance organizations lower fees per office visit (17), they might be more Independent practice associations Preferred provider organizations competitive with other practices. Physicians might Social health maintenance organizations also attempt to attract more patients by expand- SOURCE Office of Technology Assessment, 1986 ing the range of the services provided by their 63 offices, which could enable NPs and PAs to prac- cause the “plan is primarily organized around tice the full range of services for which they were solo/single specialty group practices, ” (123) which trained. do not benefit as much from employing and using NPs, PAs, and CNMs as do larger practices. Some physicians, however, might find it eco- nomically more advantageous to hire new phy- The trend toward alternative providers, most sicians rather than NPs, PAs, or CNMs. The rate of whom are profit-making entities, suggests pos- of growth in physicians’ incomes has started to sible new sources of employment. Anecdotal evi- decline, a trend that is expected to continue (20). dence indicates that ambulatory care centers are If new physicians’ incomes decline sufficiently, employing PAs and NPs. A survey of 250 indi- and if their interest in salaried positions continue vidual ambulatory care centers, owned by 142 pri- to increase, they might be more attractive than vate organizations, found that PAs’ salaries ranged NPs, PAs, or CNMs to established physicians from $20,784 to $35,000, with an average of who want to expand their practices. $25,946 (172). Humana, Inc., owns 150 ambulatory care centers (Medfirst) and employs NPs Competition among different types of health- only in its high-volume centers, about 5 percent care organizations might increase the employment of the total (163). NPs, who receive salaries or and responsibilities of NPs, PAs, and CNMs (15, hourly wages, have been found to provide stand- 143,144). For example, the growth of risk-sharing ard care and to cost Humana one-third as much HMOs—which have used the services of NPs, as physicians. Nonetheless, the organization per- PAs, and CNMs extensively in the past—would ceives a demand from its clients for physician care seem to ensure a larger role for these providers and does not intend to change its staffing patterns. in the health-care system. But like physicians’ practices, HMOs could turn instead to physicians, The effects of payment changes, such as the if their incomes are reduced enough. Anecdotal DRG approach, on the employment and use of reports from California note “that clinics that had NPs, PAs, and CNMs in hospitals have not yet intended to employ NPs and PAs were having been well documented. From individual reports, physicians arrive on their doorsteps saying they the effects appear to vary among hospitals. Some would work for $30,000 or $40,000” (263). Clinic hospitals have reportedly cut their nursing staffs administrators, then, must consider whether to and reduced the nurses’ work schedules because hire NPs or PAs at $25,000 or to hire physicians of DRGs (163). Other hospitals reportedly have for only $10,000 more. In addition to salary, how- hired PAs to increase efficiency (48). The differ- ever, other factors might enter into such decisions. ent responses were to be expected and might be NPs, PAs, and CNMs save costs for capitated en- attributed to differences in patient mix (and thus tities and provide the types of services—health differences in DRGs), in the costs of the hospi- education, counseling, and preventive care—that tals with respect to specific DRGs, and in DRG HMOs emphasize. Indeed, observers generally rates (based on geographic location—urban or ru- agree that the opportunities for employment and ral). The aggregate effect on the employment and full use of NPs, PAs, and CNMs are highest in use of NPs, PAs, and CNMs is thus difficult to capitated systems. ascertain. The increase in the numbers of IPA-model Reports also indicate that, as a result of DRG HMOs is another trend that might adversely af- payment, some hospitals are dismissing NPs and fect the employment and use of NPs, PAs, and PAs and shifting portions of their operations to CNMs. Large group- and staff-model HMOs usu- their outpatient departments, where fee-for- ally provide care at primary HMO sites and em- service physicians deliver care (117). PAs’ advo- ploy NPs, PAs, and CNMs because they are cost- cates suggest that eventually hospitals might seek saving, and because they provide health educa- more efficient outpatient operations and use PAs tion and preventive services that meet standard in an attempt to contain their costs (48). New roles levels of quality. The IPA model is less likely than could also emerge for PAs as utilization review other models to employ these practitioners, be- specialists or DRG coordinators (48). 64 —

Nurses expect that prospective payment and its ated incentives for maximizing the costs of pro- related cost management will bring about increas- viding services. The adoption of prospective paying attention to the contribution of nursing serv- ment by Medicare, some Blue Cross plans, and ices in critical care and transplant units and will some State Medicaid programs has created incen- result in a much more realistic allocation of dol- tives for minimizing such costs. In addition, pri- lars for nursing services (233). Also, because pro- vate sector groups—HMOs, PPOs, employers, spective payment may result in the early discharge and insurers—are contracting with selected hos- of patients into the community, followup serv- pitals on the basis of price. ices for patients after they are discharged are assuming increasing importance. Nurse-managed Hospitals, especially investor-owned hospitals, and nurse-owned organizations are emerging to will need to lower their costs of production in re- provide nursing services in the community, and sponse to the increasingly competitive new envi- nurses are attempting to establish a mechanism ronment (194), but investor-owned hospitals are of payment for community, nursing services (233). not hiring lower priced personnel, such as NPs, NPs are also assuming new roles in managing PAs, and CNMs, to substitute for physicians in cases and reviewing the use of hospital services inpatient settings (95). Indeed, investor-owned (96). hospitals are not employing many physicians, either (170). Investor-owned chains are using de- Studies are not available to show how the growth partment managers, who for fixed-price contracts of investor-owned hospitals and multi-hospital provide services, including personnel, for hospi- systems has affected the employment and use of tal departments (95). Because the managers are NPs, PAs, and CNMs. Studies on the differences at risk financially, however, they have incentives in economic performance based on ownership (in- to save costs and, therefore, might employ appro- vestor-owned or not-for profit) and system affili- priately trained NPs and PAs. ation (affiliated or free-standing) found no significant difference in costs for delivering comparable The growth of investor-owned hospitals might care to patients (260). Compared with other types signal fewer opportunities for CNMs to be em- of hospitals, investor-owned chain-hospitals had ployed in hospital settings. Both system-affiliated fewer employees per bed, but paid employees— and free-standing hospitals treated proportion- except nurses—more (260). The years studied ately fewer maternity patients than not-for-profit were 1978 and 1980, when payment methods cre- hospitals treated (260).

SUMMARY

The employment and use of NPs, PAs, and payment for the services of NPs and CNMs would CNMs would be affected by changes in the meth- significantly help them in administratively inde- ods of payment for their services and by other pendent practices, could stimulate the growth of changes in the health-care system. Examining how such practices to the extent permitted by State particular changes in payment would interact with laws and regulations, and would increase oppor- the other changes provides some indication of tunities for NPs and CNMs to provide the full what roles NPs, PAs, and CNMs might play in range of services for which they are trained and particular health-care settings and how costs might licensed. change for health-care providers, patients, and As independent providers, IPA-model HMOs society. might engage NPs as contractors for primary-care Despite anticipated changes in the methods of services (100) and CNMs as contractors for mater- paying for physicians’ services, fee-for-service will nity services, PPOs also might treat these practi- probably remain a major form of payment in the tioners as contractors who agreed to provide serv- foreseeable future. Allowing coverage and direct ices at a discounted fee. The opportunities for NPs 65 and CNMs to become contractors might be lim- merous variables could affect physicians’ decision ited, however, by the increasing supply of pri- to employ and appropriately use these providers. mary-care physicians, including obstetricians, and Such variables include the physicians’ billing prac- by competition from physicians, who are lower- tices; the payment levels for services of NPs, PAs, ing the amounts for which they are willing to and CNMs; the cost differentials between hiring work. physicians or hiring NPs, PAs, or CNMs; the competitive position of the physicians’ practices; NPs’ and CNMs’ employment and the full use of their skills in administratively independent the practices’ interests in expanding the range of services they provide in order to improve their practices could decrease costs for programs, beneficiaries, and society. If the numbers of services competitive positions; the abilities—as well as the NPs and CNMs and physicians provided did not physicians’ perceptions of the abilities—of NPs, greatly expand, and if the payment levels for NP PAs, and CNMs to improve the practices’ produc- and CNM services remained lower than those of tivity and income, and the physicians’ perceptions of the noneconomic benefits these providers could physicians for comparable services, lower program costs would be likely, Furthermore, if the bring to the practices. fees to patients reflected the lower payment level, Coverage might encourage fee-for-service prac- costs to beneficiaries and society might be lower. tices, particularly group practices to use NPs and In any fee-for-service practice, including one PAs in settings and for certain populations and operated by NPs or CNMs, the degree to which settings where appropriate care currently is un- costs would decrease would depend on how much available or inadequate. For example, physicians lower the level of payment was for these practi- have been reluctant to make nursing home visits, tioners than for physicians and on the particular and there is no evidence that an increased supply of physicians will decrease their reluctance. The service. For example, the Congressional Budget Office found that covering the services of PAs at increases in the elderly population and the growth rates 10 percent below those of physicians would of nursing homes have exacerbated an unmet need have negligible effects on costs or savings for the for services in this setting. Not only does the training of NPs and PAs enable them to provide the Medicare program or for society (177). Even if older population with care whose quality is com- the savings occasioned by the lower payment level parable to that of the care provided by physicians, were passed on to beneficiaries, they would have only small incentives to seek treatment from lower but evidence shows that teams of physician, NPs, priced PAs. At the margin, patients would pay and PAs visiting patients in nursing homes pro- coinsurance of ‘only 20 percent. A reduction in vide standard care and reduce total expenditures. 14 the charge for an office visit from $30.00 to $27 .00 Elderly people and children with disabling con- would save a Medicare patient only $0.60, an ditions and other individuals with chronic conditions would also benefit from NP and PA care amount that might well be paid by Medicaid or in the home setting. a private Medi-Gap policy and would not provide an incentive to use such services. Similarly, The employment practices of HMOs, the health- most of the services provided by NPs are primary care setting with significant growth potential, care services, such as visits, and would likely not would not be directly influenced by changes in provide much saving for a patient. Maternity the current methods of paying for the services of care, however, is costly and patients’ out-of- NPs, PAs, and CNMs because most public and pocket costs could be high. If CNMs would ac- private third-party payers cover such services in cept lower payment levels than those of physi- HMO settings. Furthermore, whether payments cians, any savings passed on to the expectant were direct or indirect to the NP, PA, and CNM, mother would be considerable. How covering their services would affect the “The Omnibus Reconciliation Act (Public Law. QQ-50QI enacted employment and use of NPs, PAs, and CNMs in during the publication ot this case study provides colerage t[~r wr\’- physicians’ fee-for-service practices is unclear. Nu- ices of PAs provided i n n u rsing homes under hled Ica re. 66

would not be an issue for organizations paid pro- In order for coverage and direct payment to af- spectively by a capitated amount. fect the employment of NPs, PAs, and CNMs by hospitals for providing inpatient services, the costs However, the increase in the number of IPA- of their services would be billed as professional model HMOs does affect the employment of NPs, services. If the payment levels for the services they PAs, and CNMs. In 1985, although group model provided were lower than those for physician’s HMO plans retained the lead in total enrollment, services, and if the volume of services were not IPA model plans outnumbered all other kinds of increased, savings might be likely for Medicare HMO plans for the first time (123). Because they and—if fees were lowered accordingly—for so- are primarily solo or single-specialty practices, ciety. However, if Medicare paid NPs or CNMs IPAs are less likely than group model HMOs to for providing services for which hospitals were employ these practitioners. also paid under the DRG rate, paying for them The data suggest that NPs, PAs, and CNMs separately might increase program costs, if DRG save costs for HMOs. In an increasingly competi- payment rates were not changed. Reducing DRG tive environment, the financial incentives promote rates to account for eliminating the costs associ- passing onto consumers the savings generated by ated with the NPs’ or CNMs’ services would be the employment and full use of NPs, PAs, and extremely difficult because of the lack of data. In CNMs. Thus, as the environment becomes more any case, because the proportion of the DRG rate competitive, the employment of these providers ascribed to nursing costs is unknown, the effects in capitated HMOs could benefit society finan- of direct payment on organizational, program, or cially. To the extent these providers are used to societal costs cannot be determined. provide interpersonal care and preventive services, the types of services traditionally incorpo- A major change in health-care delivery is the rated into the practice of these providers and of growth of investor-owned hospitals, particularly HMOs, the quality of care will also benefit. investor-owned chains of hospitals. These orga- Third-party payers pay hospitals an aggregate nizations are currently focusing their efforts on sum for operating costs, and the hospitals are re- attracting medical specialists to their staffs and sponsible for paying salaried employees. There- have evinced no interest in employing NPs, PAs, fore, coverage and direct payment for inpatient and CNMs. The advantages of coverage for the hospital services provided by NPs, PAs, and CNMs services of these providers do not appear to be would not directly affect their employment pos- sufficiently significant to spark such interest. sibilities. This is especially applicable to Medicare, which pays for inpatient services on a DRG-rate In the final analysis, it seems that extending basis. This payment method creates incentives for coverage for the services of NPs, PAs, and CNMs lowering the cost of resources, and the costs of in at least some settings could benefit the health NPs, PAs, and CNMs are included in calculating status of certain segments of the population cur- the costs of resources. Although coverage and sep- rently not receiving appropriate care. The imme- arate billing for their services could clarify their diate effects on third-party costs are unclear, al- revenue-producing abilities as well as their costs though long-term effects could be a decrease in to the employing hospital, the use of these prac- total costs. The advantages of direct payment for titioners to provide patient care as hospital em- the services of NPs and CNMs are less obvious. ployees is likely to decline under DRG-based pay- Direct payment might encourage qualified NPs ment. PAs and NPs could be used in new roles, and CNMs to move into unserved and under- such as DRG coordinators. served areas to expand access to health care. Appendixes —

Appendix A Methods and Acknowledgments

The study is based on an analysis of information obtained from an extensive review of the literature and from individuals and organizations with relevant experience. An advisory panel of experts with backgrounds in health policy, medical economics, health insurance, medicine, nursing and consumer advocacy defined the goals for the study and suggested source material, subject areas, and perspectives to consider in presenting the material. The drafts of the report were revised to reflect the thoughtful comments of the panel. OTA thanks the panel for its assistance and the following people and organizations for supplying information and reviewing drafts. Joel J. Alpert Carl Fasser Kerry Kemp Boston City Hospital Baylor University Office of Technology Assessment Boston, MA Houston, TX Washington, DC American Nurses Association William Finefrock Cynthia P. King Washington, DC American Academy of Physician American Medical Association American College of Nurse- Assistants Chicago, IL Midwives Arlington, VA Karl Kronebusch Washington, DC Loretta C. Ford Office of Technology Assessment American Academy of Physician University of Rochester Medical Washington, DC Assistants Center William Larson Arlington, VA Rochester, NY Health Care Financing David Banta Louis P. Garrison Administration The Netherlands Project HOPE Baltimore, MD Millwood, VA James D. Campbell Kenneth Lease University of Missouri Archie Golden U.S. Office of Personnel Columbia, MO Chesapeake Health Plan–South Management Side Washington, DC James F. Cawley Baltimore, MD George Washington University Charles E. Lewis Medical Center Linda Golodner University of California Washington, DC National Consumers League Los Angeles, CA Washington, DC Katherine H. Chavigny Joan Lynaugh American Medical Association Bradford Gray University of Pennsylvania Chicago, IL Institute of Medicine Philadelphia, PA Washington, DC James Crouch Nancy March Utah Department of Health Marie Hawk American College of Nurse- Salt Lake, City, UT Harvard Community Health Plan Midwives Boston, MA Washington DC M.L. Detmer American Medical Association Anita Hegster Lynn May Chicago, IL Health Care Financing American Academy of Physician Administration Assistants Karen Ehrnman Baltimore, MD Arlington, VA American College of Nurse- Midwives Martha Hill Kathy Michels Washington, DC Johns Hopkins School of Nursing American Nurses Association Baltimore, MD Washington, DC E. Havey Estes, Jr., Duke University Medical Center Ada Jacox Evelyn Moses Durham, NC University of Maryland Health Resources and Services Baltimore, MD Administration Claire M. Fagin, Rockville, MD University of Pennsylvania Jean Johnson Philadelphia, PA George Washington University Washington, DC

69 70

Norbert Nelson Gretchen Schafft Marlent Ventura New York University Medical American Academy of Physician Veterans Administration Hospital School Assistants Buffalo, NY New York, NY Arlington, VA Judith Wagner Ronald Nelson Sherry Shamansky Office of Technology Assessment American Academy of Physician Yale University Washington, DC Assistants New Haven, CT Jerry Weston Arlington, VA Jane Sisk National Center for Health Robert Oseasohn Office of Technology Assessment Services Research University of Texas Washington, DC Rockville, MD San Antonio, TX Julie Sochalski Judith Willis Henry B. Perry Ann Arbor, MI Health Care Financing Mountain Medical Center Sally Solomon Administration Clyde, NC National League for Nursing Baltimore, MD Elaine Power New York, NY Sidney Wolfe Office of Technology Assessment Brenda Splitz Health Research Group Washington, DC George Washington University Washington, DC Robert Ranney Washington, DC Susan Yates National Rural Health Care Margetta Styles American College of Nurse- Association American Nurses Association Midwives Kansas City, MO Kansas City, MO Washington, DC Ginette Rodger Dan Thomas Canadian Nurses Association Health Insurance Association of Ottawa, ON America Washington, DC Appendix B Payment for the Services of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives

Health-care services are paid for by individuals and The fourth medical-benefits program provided by by third-party payers. Third-party payers in the pri- the Federal Government is the Federal Employees vate sector include commercial insurance companies; Health Benefits Program (FEHBP), a voluntary health- hospital and medical plans, such as Blue Cross and care program that provides health insurance for ap- Blue Shield; prepaid group medical plans, such as proximately 10 million Federal employees and their de- health maintenance organizations (HMOs); and others, pendents. Enrollees receive health-insurance services such as labor unions or employers of insured individ- from more than 300 health-benefit plans under con- uals (106). Specific benefits, exclusions, and limitations tracts negotiated with the Office of Personnel Man- on financial coverage vary from one third-party payer agement of the U.S. Government (256). to another and differ even among the policies and plans As table B-1 shows, payment for the services of offered by a particular payer. However, State and, to nurse practitioners (NPs), physician assistants (PAs), a lesser extent, Federal laws and regulations require and certified nurse-midwives (CNMs) varies consider- private third-party payers to offer some benefits and ably, in part because of variations in the State laws do not permit them to offer others. and regulations that govern these providers’ practices The Federal Government plays a significant role in and payment. Table B-1 provides a generalized over- paying for health-care services under four primary- view of the payment practices of the major third-party health-care programs. The government acts as a third- payers in the public and private sectors. These prac- party payer for health care under the Medicare and tices are described in greater detail below. the Medicaid programs. Although the Health Care Financing Administration (HCFA) is the Federal agency Nurse Practitioners and responsible for both Medicare and Medicaid, the two Physician Assistants programs differ considerably in their payment practices and covered populations. Medicare is a nation- Government-Sponsored Programs wide health insurance program for the 27.5 million Americans who are at least 65 years of age and for Medicare.—Under Part B of the Medicare program, 2.9 million disabled Americans, Part A, the Hospital coverage and payment for NPs’ and PAs’ services are

Insurance Program helps pay for hospital services, re- restricted to services not traditionally performed by lated institutional services, and other services. Part B, physicians, to services normally delegated by physi- the Supplementary Medical Insurance Program cov- cians, and to services performed under the direct su-

ers physicians’ services and many other medical serv- pervision of physicians. This provision is commonly ices. Medicaid is a joint Federal-State program for 22 termed the “incident to” provisional million low-income persons. The program is admin- Under this provision, services of nonphysicians may

istered by individual States under general Federal be covered where they are of types which are commonly guidelines, which include mandatory minimum bene- performed by physicians’ office personnel, and are per- fits that all States must provide to eligible recipients formed by employees of the physician under his or her and optional benefits that individual States may elect direct supervision, e.g., giving injections, taking tem- peratures and blood pressures, performing blood tests, to provide to recipients. etc. Payment cannot be made, however, for services The Civilian Health and Medical Program of the performed by nonphysicians where the services are of Uniformed Services (CHAMPUS), the third medical- benefits program provided by the Federal Govern- ‘The relevant Medicare Part B regulation prohibits payment for ment, is administered by the Department of Defense medical services rendered by someone other than a physician ex- (DOD) (245), CHAMPUS covers nearly 8 million de- cept for services that are “furnished as an incident to a physician’s pendents of military personnel, retirees, and depen- professional services of kinds which are commonly furnished in phy- dents of retirees inside and outside the United States sicians’ offices and are commonly either rendered without charge or included in physician’s bills. ” Sec. 1861(s)(2)(A) of the Social Secu- (60). rity Act, 42 U, S.C. Sec. 1395(s)(2)(A), 20 CFR 405-231(b).

71 72 —.

Table B-1 .—Coverage and Direct Payment for Servicesa of Nurse Practitioners, Physician Assistants, and Certified Nurse-Midwives — Nurse practitioners Physician— assistants Certified nurse-midwives— Direct Direct Direct Third-party payer Coverage payment Coverage payment Coverage payment Medicare: Part A ...... No No No No No No Part B ...... No No Nob No No No HMOs C . Yes NA Yes NA Yes NA State Medicaid programsd . . . .Some A few Some None Almost all Almost all programs programs programs programs programs Medicare and Medicaid: Rural Health Clinics. ., . . Yes No Yes No Yes No CHAMPUS e ...... Yes Yes No No Yes Yes FEHBP f ...... 7 plans 7 plans 6 plans 6 plans 20 plans 20 plans Private insurance ...... In some In some No No In some In some States States States . States NA = not available aservices that are typically and characteristically provided by Physicians.

bDuring the Publication of this case study, the Omnibus Rec~ricili~tlon Act of 19~ (publlc Law 99.509) was enactfjd The act modifies part B Of Medicare and authorizes payment for (covers) services of physician assistants working under the supervision of physiciansIn hospitals, skilled nursing faclllties, Intermediate.care facllltles, and as an assistant at surgery. The payment is indirect and at levels lower than physicians would receive for prowdlng comparable services cHealth maintenance organizations.

dstate Medicaid programs have the option of irl~luding NF’ and PA se~ices lfl their state MediCald plans. Congress rnarlctated coverage Of CNMS’ SerVICeS In 1980 As of January 1985, all States in which CNMS practiced either were complying with the law (Public Law 96-499) or were considering changes In their Medical plans to comply with the law. ecivilian Health and Medical Program of the Uniformed services fFederal Employees Health Benefit program. FEHBP has 21 fee-f or.service plans, some of which authorize PaYment to NPs PAs, and Cf’Jfvf S

gwhether State laws and regulations require or permit Insurance coverage and direct payment for the serwces of NPs, PAs, and CNMS SOURCE Office of Technology Assessment, 1986

the kinds which are typically and characteristically ren- only in areas that do not have certified home-health dered by physicians, e.g., prescribing medications, set- agencies. In 1984, there were 5,247 Medicare certified ting casts on fractures, assisting at surgery, and other home-health agencies (164), and the number is growing activities that involve an independent evaluation or (115). Presumably, therefore, NPs and PAs provide treatment of the patient’s condition even if the attending physician is directly supervising these services (64). services to homebound patients only to a limited extent The “incident to” provision was partly intended to and only in areas where home-health agencies do not reduce the possibility of physicians’ making excessive find it economical to function. profits by employing large numbers of assistants (162). The Tax Equity and Fiscal Responsibility Act of 1982 The provision has been refined over time, and its com- (Public Law 97-248) allows for Medicare coverage of plexity has led to varied interpretation by physicians. NPs’ and PAs’ services in HMOs and competitive medical plans (CMPs) that have entered into certain con- Strictly interpreted, the provision means that Medicare 2 only pays for physicians’ typical services when they tractual risk-sharing arrangements with HCFA. The are actually provided by physicians. Knowingly or implementing regulations permit NPs and PAs in HMOs and CMPs to furnish services without the direct per- unknowingly, however, some physicians bill for serv- ) ices irrespective of who performs the service. Unless sonal supervision of physicians. The NPs and PAs audits are performed, Medicare contractors have dif- essentially can provide whatever services State law au- ficulty determining who has rendered services from the thorizes, including supervising or ordering services and Medicare billing form. One of the “incident to” pro- supplies incidental to the services. vision’s effects has been to sharply limit the adminis- During the publication of this case study, the Om- tratively independent practice of NPs who cannot bill nibus Reconciliation Act of 1986 (Public Law 99-509) Medicare for medical services. This provision was modified in 1980 (248) to permit ‘Calculations of cavitation rates do not include NPs’ or PAs’ sal- generally supervised nurses and other paramedical aries but are determined by the average adjusted per capita costs personnel—such as NPs and PAs—to provide certain which are based on the costs of past services received by beneficiaries who fall into particular sets governed by such factors as geo- services to the homebound in some medically under- graphic location, age, sex, and eligibility. served areas. The “incident to” provision is waived 3Federal Register, vol. 50, No. 7, Thursday Jan. 10, 1985, p. 1351. 73 — was enacted. The act modifies Medicare and author- The scope of services covered for PAs also varied from izes payment for (covers) services of PAs working un- the general (e.g., all the services cited in the PA law der the supervision of physicians in hospitals, skilled governing scope of services) to the specific (e.g., ex- nursing facilities, intermediate-care facilities, and as aminations under the program Early and Periodic an assistant at surgery. The payment is indirect and Screening, Diagnosis, and Treatment; services in com- at levels lower than physicians would receive for pro- munity health centers; and services in family planning viding comparable services. agencies). Three States specified that only “incident to” Medicare’s payment for inpatient hospital services services (i.e., services not traditionally performed by under Part A does not specify coverage or payment physicians) were covered for payment (25). for NPs’ and PAs’ services, either under Medicare’s Medicaid payment for inpatient hospital services former cost-based reimbursement method or the differs by State. Although 41 State Medicaid programs current prospective-payment system. Hospitals usu- paid for hospital inpatient services on a retrospective ally pay for NPs’ and PAs’ services by salaries; the sal- cost basis at the beginning of 1980, 34 State Medicaid aries and other costs of employing or contracting with programs had some form of prospective-payment NPs and PAs are included in the hospitals’ formulas system as of December 1985 (133). Each State Medic- for calculating operating costs. Under cost-based reim- aid program pays for operating costs—including sal- bursement, Medicare pays the hospital the total oper- aries and other costs associated with NPs and PAs— ating costs associated with Medicare beneficiaries. according to its unique payment method for inpatient Under the prospective-payment system, Medicare pays services (40). a fixed amount for each patient admitted; the aggre- Rural Health Clinics.—Access to primary-care gated amount is intended to cover the hospitals’ total services by NPs and PAs in satellite settings in isolated operating costs for Medicare beneficiaries. areas was hindered by the fact that payment for such services was available under Medicare and Medicaid Medicaid.—Under Medicaid, each State has consid- only if a physician was on the premises when the erable discretion to design its program within broad services were delivered. The Rural Health Clinic Federal guidelines. Covering and paying for the Services Act of 1977 (Public Law 95-210) waived such services provided by NPs and PAs is one of the bene- restrictions for NPs and PAs practicing in certified ru- fits a State may choose to include in its Medicaid Plan. ral health clinics located in designated underserved Data on the number of State Medicaid programs that areas. The act permits payment for the services of NPs cover NPs’ services are not collected by HCFA’s central and PAs even when they are not directly supervised office. Although the available data conflict, they by physicians at all times. This allows rural clinics indicate that State Medicaid programs are cautious staffed only by NPs and PAs backed up by physicians about extending payment to NPs. A 1985 study noted to provide reimbursable primary care typically pro- that NPs were authorized to receive direct payment vided by physicians, so long as written plans of or indirect payment—i.e., to bill directly or through treatment are periodically reviewed and approved by physicians—in 21 State Medicaid programs (60). An physicians. Payment, which is based on reasonable earlier study found that of the 26 State Medicaid costs, is made to the employing clinic, not to the NP programs that covered NPs’ services, most paid in- or PA, and is restricted to services that State legisla- directly. Nineteen of the twenty-six States adopted the tion authorizes NPs and PAs to perform. Medicare approach of allowing payment only for NPs’ Nursing Homes.—Various Medicare and Medicaid services that were incidental to physicians’ services regulations, in addition to coverage and payment (22). provisions, limit the provision of certain services by A preliminary survey of State Medicaid programs PAs and NPs in nursing homes. In some States, the found that 26 of the 36 State Medicaid programs cov- laws permit physicians to delegate such services to NPs ered PAs’ services (5). Of those 26 programs, 18 re- and PAs. imbursed for PAs’ services at the same rates as physi- Only physicians can provide certain services if a fa- cians’, 4 reimbursed at lower rates, 2 reimbursed on cility is to: a cost basis, and the remainder did not respond to the 1. be certified as a skilled nursing facility (SNF) in question. Most of the State Medicaid programs’ re- the Medicare and Medicaid programs (42 CFR quirements for supervision by physicians were simi- 405.1123,1124,1125,1126, and 1128); lar to the requirements contained in State laws gov- 2. be certified as an intermediate-care facility (ICF) erning PAs’ practice. (In most States, the scope of PAs’ in the Medicaid program (42 CFR 311, 334, 343, practice is controlled under medical-practice acts and and 346); regulations. ) Other State Medicaid programs require 3. obtain certification and recertification of a patient’s that physicians review patients’ charts every 7 days, need for care in an SNF in the Medicare program that physicians be onsite, or that physicians be present. (42 CFR 456.260, 270, and 280); or 74

4. obtain certification of a patient’s need for care in tract year 1986, 7 cover and offer direct payment for an SNF and ICF in the Medicaid program (42 CFR services of NPs and 6 cover and offer direct payment 456.360, and 380). for the services of PAs4 (256). Only 14 percent of The specific services that must be performed by physi- enrollees in FEHBP are enrolled in plans that cover cians vary according to the type of certification and NPs’ services and 11 percent of enrollees in FEHBP are the program. Under the Medicare and Medicaid pro- enrolled in plans that that cover PAs’ services. Direct grams, for example, patients can be admitted to SNFs payment for NPs and other providers is now under based only on physicians’ medical findings, diagnosis, consideration by Congress.5 and orders. Patients’ care must be supervised by physicians, and patients must be seen by physicians at least Private Insurance every 30 days for the first 90 days after admission. Only physicians can prescribe drugs and order diag- Private third-party payment for NPs’ and PAs’ serv- nostic and specialized rehabilitative services and ther- ices is subject to State laws and health insurance reg- apeutic diets. ulations. Increasing numbers of States have passed Unlike Medicare, Medicaid allows NPs and PAs to laws and regulations concerning payment for the serv- recertify patients’ needs for institutional care. NPs and ices of NPs and PAs. Such laws and regulations must PAs are authorized to recertify the necessity of accord with the States’ requirements governing the continuing medical care in SNFs (42 CFR 456.260) and scope of practice of these providers and, in some cases, ICFs (42 CFR 456.360) where general supervision is of physicians. provided by physicians. The State payment laws vary in a number of dimen- Civilian Health and Medical Program of the Uni- sions, including the types of insurers affected (for- formed Services. -The Federal Government, through profit, nonprofit, or both) and the types of insurance the Department of Defense’s CHAMPUS, has taken policy (22). Some laws affect the services of all nurses; the lead in treating NPs as autonomous and independ- others affect only special groups of nurses, such as ent providers of care for payment purposes. CHAMPUS NPs. Some States require insurers to include nurses’ began billing and paying for NPs’ services on an ex- services as a reimbursable benefit (mandatory bene- perimental basis in fiscal year 1980. When the experi- fit), whereas other States require insurers to offer reim- ment ended 2 years later, CHAMPUS continued cover- bursement for nurses’ services as an option in their pol- age and direct fee-for-service payment of NPs, thereby icies (mandatory option) (232). recognizing them as a distinct group of providers de- 4 300 prepaid compre- serving direct compensation for services (60). Although The numbers do not include the more than hensive medical plans in the FEHBP, because the organization of CHAMPUS does not cover PAs’ services, PAs are not medical delivery systems under these plans makes the issues of di- seeking coverage under CHAMPUS, because DOD has rect access, payment, supervision, and referral largely irrelevant. indicated that CHAMPUS will begin contracting out 5In early 1986, President Reagan vetoed H. R. 3384 which con- its services and cease paying on a fee-for-service basis tained a provision requiring direct reimbursements to nurses and nurse-midwives who provide services to employees covered by the (83). FEHBP. Congress then passed new legislation, Public Law 99-251, Federal Employees Health Benefit Program.—Like directing the Office of Personnel Management (OPM) to study and CHAMPUS, FEHBP experimented with direct payment report to Congress on the advisability of amending the law governing and required that all FEHBP plans directly pay health FEHBP to provide mandatory recognition of additional health-care practitioners, including NPs and PAs, who were li- practitioners, such as nurse-midwives, nurse practitioners, chiropractors, and clinical social workers. The legislation extended direct reim- censed under applicable State law in those States where bursement for nonphysician providers in medically underserved at least 25 percent of the population was located in areas, which are determined by the Department of Health and Hu- formally designated primary-medical-care manpower man Services to have at least 25 percent of the population living -shortage areas (60). After the experimental period of in areas with inadequate numbers of medical providers. OPM’s study advised against mandatory coverage on grounds specific to FEHBP January 1980 to December 1984, FEHBP did not require (e.g., mandating coverage would not increase the choice of practi- plans to compensate NPs and PAs directly. tioners available to plan members, nor would it necessarily increase Payment to providers of covered services currently competition among the plans). Nonetheless, the Subcommittee on depends on the terms of the FEHBP’s contract with Compensation and Employee Benefits of the House Committee on each health-benefit plan and thus varies among the Post Office and Civil Service remains interested in the topic. The subcommittee held hearings on direct reimbursement for nonphy- plans. There is no statutory requirement that all plans sicians on Apr. 15, 1986, and indicated its intention to continue offer payment to NPs and PAs, but some plans cur- studying the issue. H.R. 4825, introduced on May 14, 1986, would rently authorize NPs and PAs to receive direct pay- authorize direct payment for services performed by NPs and CNMs ment or reimbursement for covered services without and other health-care providers. As of June 1986, the bill had been reported favorably by the House Committee on Post Office and Civil referral or supervision (see table B-1). Of the 21 fee- Service and was awaiting floor action. The bill did not pass the 99th for-service plans participating in FEHBP for the con- Congress. 75

Although direct third-party payment is the excep- under Medicaid. The Federal statute recognizes CNMs’ tion rather than the rule, 13 States currently permit autonomous practice expressly stating that the man- direct payment for NPs’ services (24). The wide vari- dated benefit shall be provided “whether or not he is ation in conditions for payment of NPs’ services is under the supervision of, or associated with, a physi- apparent in the laws of Mississippi, Maryland, and cian or other health care provider” (60). HCFA issued Oregon regarding supervision by physicians. In all the regulations that implemented this law in May 1982. three States, insurers must pay for any service that is As of January 1985, all States in which CNMs prac- within NPs’ lawful scope of practice, but Mississippi ticed either were complying with the statute and the requires the NPs to work under the supervision of phy- regulations or were considering changing their Med- sicians, whereas Maryland prohibits direct payment icaid plans to bring them into compliance. Currently to NPs who work under the direct supervision of phy- only four States and the District of Columbia do not sicians (101). In Oregon, supervision by physicians is provide for direct Medicaid payment to CNMs, and not a condition for reimbursement (2 I). HCFA’s regional offices are working with these juris- No State laws mandate coverage of PAs’ services. dictions to bring them into compliance (235). Further- Except in Wisconsin, State laws are silent even about more, the Medicaid statute was amended by Public optional coverage of PAs’ services (83). None of the Law 98-369 to ensure that birthing centers operated States mandate direct reimbursement for PAs’ services; by CNMs need not be administered by physicians to indeed, 16 States explicitly prohibit it. Although there be eligible for coverage as Medicaid clinic services. is anecdotal information concerning third-party payers Rural Health Clinics.--CNMs are treated differently who cover PAs’ services, sometimes under physicians’ from NPs and PAs under the Rural Health Clinics Act. billing, information concerning the extent of coverage Only rural clinics employing NPs or PAs are eligible is not available. for certification under the act (Title 42, Section 481.4). Businesses in the United States are beginning to pro- Once a clinic is certified, however, it can receive pay- vide insurance that pays directly for NPs and PAs (as ment for the services of the CNMs it employs. well as CNMs). The Washington Business Group on Civilian Health and Medical Program of the Uni- Health recently conducted a national survey of its formed Services.—CHAMPUS singled out CNMs for member organizations, all of which are large firms. special consideration before it experimented with di- Of the approximately 200 respondents, 43 percent are rect payment for NPs’ services starting in 1980. The paying directly for the services of NPs, and 39 per- Defense Appropriations Act of 1979 (Public Law 95- cent are doing so for PAs (91). The proportion of mem- 457) was the first Federal law to pay directly for serv- ber companies reimbursing NPs and PAs (and CNMs) ices provided by CNMs without either referrals or di- has increased steadily over the past decade (91). rect supervision by physicians. In many States, NPs’ and PAs’ services still must Federal Employees Health Benefit Program.—Of the be “incident to” physicians’ services, for payment pur- 21 FEHBP fee-for-service plans, 20 cover CNMs with- poses, and compensation for NPs’ and PAs’ services out a contractual requirement for physicians’ referrals must be made to their employing physicians or orga- or supervision. In addition, many prepaid plans in the nizations. Nevertheless, the recent changes in some FEHBP employ CNMs. Roughly 90 percent of all Fed- States’ laws and in the policies of major corporations eral enrollees are in plans that cover CNMs (256). suggest a movement away from requirements for di- Many of the insurance companies in the FEHBP offer rect supervision by physicians. Increasingly, NPs and the same coverage of CNMs for their private sector PAs can function administratively independently of business. physicians and qualify for direct payment. Also, more States are likely to pass legislation providing for the Private Insurance direct compensation of NPs and PAs. Private third-party payment for CNMs’ services has Certified Nurse-Midwives also been mandated in a growing number of jurisdic- tions. As of 1983, 14 States had mandated direct reim- Government-Sponsored Programs bursement by private insurers for CNMs’ care (55), By April 1986, the number of States had increased to 17 Medicare and Medicaid.—Medicare’s policies con- (11). In most States, direct supervision by physicians cerning payment are the same for the services of CNMs is not a condition of reimbursement (22). In addition, as for the services of NPs and PAs. Medicaid’s pay- “in many other States insurers voluntarily have cho- ment policies are much more permissive for CNMs’ sen to pay for nurse-midwifery care” (55). Fifty-seven services than for NPs’ and PAs’ services. In 1980, percent of the large corporations surveyed by the Wash- Congress enacted legislation (Public Law 96-499) to ington Business Group on Health provide direct reim- require that CNMs’ services be a mandatory benefit bursement to CNMs (91). References References

1. Aday, L. A., and Andersen, R., Development of 16, American Medical Association, Center for Health Zndices of Access to Medical Care (Ann Arbor, Policy Research, SMS Reports 2(7), November IMI: Health Administration Press, 1975). 1983. 2. Alexander, J. A., Lewis, B. L., and Morrisey, 17. American Medical Association, Center for Health M. A., “Acquisition Strategies of Multihospital Policy Research, Socioeconomic Characteristics Systems,” Health Affairs 4(3):49-66, fail 1985. of Medical Practice 1983, R.A. Reynolds and J.B. 3. Allied Health Education Directory, 13th ed. Abram (eds. ) (Chicago, IL: 1983). (Washington, DC: American Medical Associa- 18. American Medical Association, Center for Health tion, 1985). Policy Research, Socioeconomic Characteristics 4. Alsofrom, J., “Number of Group Practices Ris- of Medical Practice (Chicago, IL: 1984). ing in U.S., ” American Medical News 27(45):19, 19. American Medical Association, Center for Health Dec. 7, 1984. Policy Research, SMS Reports 3(8), November 5. American Academy of Physician Assistants, 1984. “Survey of Chapters, ” Arlington, VA, unpub- 20. American Medical Association, Center for Health lished document, September 1984. Policy Research, Socioeconomic Characteristics 6. American Academy of Physician Assistants, 1984 of Medical Practice (Chicago, IL: 1985). Physician Assistant Masterfde Survey (Arlington, 21. American Nurses Association, Nurse Practitioners: VA: 1984). A Review of the Literature (Kansas City, MO: 7. American Academy of Physician Assistants, 1983). Hearings Before the Subcommittee on Health, 22. American Nurses Association, Third-Party Re- Senate Committee on Finance, U.S. Congress, imbursement for Services of Nurses (Kansas City, Washington, DC, Apr. 25, 1986. MO: 1983). 8. American Academy of Physician Assistants, 23. American Nurses Association, Obtaining Third- “AAPA Membership Statistics by Graduation Party Reimbursement: A Nurse’s Guide to Meth- Date, ” Arlington, VA, May 13, 1986. ods and Strategies (Kansas City, MO: 1984). 9. American College of Nurse-Midwives, Nurse- 24. American Nurses Association, Washington, DC, Midwifery in the United States: 1976-1977 (Wash- personal communication, June 1986. ington, DC: 1978). 25. Association of Physician Assistant Programs/ 10. American College of Nurse-Midwives, Nurse- American Academy of Physician Assistants, Na- Midwifery in the United States: 1982 (Washing- tional Physician Assistant Survey (Arlington, ton, DC: 1984). VA: May 1982). 11 American College of Nurse-Midwives, “Federal 26. Atherton, R. A., and LeGendre, S. T., “A Descrip- Employees Health Benefits Program, ” Hearings tion of Nurse Practitioners’ Practice in Occupa- Before the Subcommittee on Compensation and tional Health Settings, ” occupational Health Employee Benefits, House Committee on Post Of- Nursing 33(1):18-20, January 1985. fice and Civil Service, U.S. Congress, Washing- 27, Bailet, H., Lewis, J., Hochheiser, L., et al., ton, DC, Apr. 15, 1986. “Assessing the Quality of Care, ” Nurs. Outlook 12. American College of Nurse-Midwives, Washing- 23(3):153-159, March 1975. ton, DC, personal communication, Aug. 20, 28. Barer, M. L., Evans, R. G., and Labelle, R., “The 1986. Frozen North: Controlling Physicians Costs 13. American CoI1ege of Nurse-Midwives and Amer- Through Controlling Fees, ” prepared for the Of- ican College of Obstetricians and Gynecologists, fice of Technology Assessment, U.S. Congress, “Joint Statement of Practice Relationships Be- Washington, DC, November 1985. tween Obstetricians/Gynecologists and Certified 29. Batey, M. V., and Holland, J. M., “Prescribing Nurse-Midwives, ” unpublished mimeo, Washing- Practices Among Nurse Practitioners in Adult ton, DC, Nov. 1, 1982. and Family Health, ” Am. J. Public Health 14. American Hospital Association, Directory of 75(3):258-262, March 1985. Multihospital Systems (Chicago, IL: American 30. Becker, D, M., Fournier, A. M., and Gardner, Hospital Publishing, Inc., 1985). L, B., “A Description of a Means of Improving 15. American Medical Association, Center for Health Ambulatory Care in a Large Municipal Teach- Policy Research, SMS Reports 2(3), June 1983. ing Hospital: A New Role for Nurse Practi-

79 80 — .

tioners, ” Med. Care 20(10): 1046-1050, October Nurse-Midwives, ” Nurs. Outlook 23: 453-456, 1982. July 1975. 31, Bellet, P. S., and Leeper, J. D., “Effectiveness of 44 Butler, C., “The 1983 NAPNAP Membership the Pediatric Nurse Practitioner Well-Baby Survey, ” Pecf. Nurs. 10(3):187-190, May/June Clinics in West Alabama, ” Ala. J. Med. Sci. 1984. 19(2):126-128, April 1982. 45 Carter, R. D., Perry, H. B., and Oliver, D., “Sec- 32, Bennett, S., and Biener, L., Executive Summary: ondary Analysis: 1981 National Survey of Phy- Determinants of Effectiveness of Physician-Nurse sician Assistants, ” final report to the National Practitioner Teams, Grant No. I RO1 HSO 4104 Center for Health Services Research, Public (Washington, DC: National Center for Health Health Service, U.S. Department of Health and Services Research, March 1983). Human Services, Washington, DC, June 25, 33. Bennetts, A. B., and Ernest, E. M., “Freestanding 1984, Birth Centers, ” Research Issues in the Assessment 46 Caserta, J. E., and Addiss, S., “Economics of of Birth Settings, Pub. No. (IOM) 82-04 (Wash- Community Health: Dealing With Realities, ” ington, DC: National Academy Press, 1982). American Academy of Nursing, American 34. Bibb, B. N., “Comparing Nurse-Practitioners and Nurses’ Association, 1985. Physicians: A Simulation Study on Processes of 47, Caward, J. F., “Economics of the N.P. Role in an Care, ” Eva]. & Health Prof. 5(1):29-42, March Industrial Setting, ” Nurse Pract. 6(6):17, Novem- 1982. ber/December 1981. 35. Breslau, N., “The Role of the Nurse-Practitioner 48, Cawley, J. F., “Health Policy Issues Facing PAs, ” in a Pediatric Team: Patient Definitions, ” i’kfed. Physician Assistant 9(2):133-162, February 1985.

Care 15(12):1014-1023, December 1977. 49< Cawley, J. F., “The Physician Assistant Profes- 36. Brooks, E. F., “Nurse Practitioners’ Changing Im- sion: Current Status and Future Trends, ” J. Pub- pact on Health Care, ” presented at the 12th An- lic Health Policy 6(1):78-99, March 1985. nual Regional Conference on Maternal and Child 50. Chambers, L. W., Bruce-Lockhart, P., Black, Health, Family Planning, and Crippled Childrens’ D. P., et al., “A Controlled Trial of the Impact Services, Chapel Hill, NC, Apr. 23, 1985, of the Family Practice Nurse on Volume, Qual- 37. Brooks, E. F., Bernstein, J. D., DeFriese, G. H., et ity, and Cost of Rural Health Services, ” Med. al., “New Health Practitioners in Rural Satellite Care 15(12):971-981, September 1977. Health Centers: The Past and Future, ” J. Com- 51. Chambers, L. W., Burke, M., Ross, J,, et al., munity HeaZth 6(4):246-256, summer 1981. “Quantitative Assessment of the Quality of Med- 38. Brooks, E. F., and Johnson, S., “Nurse Practi- ical Care Provided in Five Family Practices Be- tioner and Physician Assistant Rural Satellite fore and After Attachment of a Family Practice Health Centers: The Pending Demise of an Or- Nurse, ’rJ. Can, Med. Assoc. 118:1060-1064, May ganizational Form?” A4ed. Care 24(10):881-890, 6, 1978. October 1986. 52. Charney, E,, and Kitzman, H., “The Child Health 39. Brooks, E. F., and Whitsel, J. R., “Recent Changes Nurse (Pediatric Nurse Practitioner) in Private in the Supply and Distribution of Physicians in Practice: A Controlled Trial, ” N. Engl. J. Med. North Carolina, ” N,C. A4ed. J. 44(8):477-480, 285(24):1353-1358, Dec. 9, 1971. August 1983. 53. Cherry, J., and Foster, J. C., “Comparison of Hos- 40. Brown, C., Division of Alternate Reimbursement pital Charges Generated by Certified Nurse- Systems, Health Care Financing Administration, Midwives’ and Physicians’ Clients, ” J. Nurse- U.S. Department of Health and Human Services, Midwifery 27(1):7-11, January/February 1982. Baltimore, MD, personal communication, Apr. 54. Cohn, S. D., “Introduction,” J. Nurse-Midwifery 30, 1986. 29(5):57-62, September/October 1984. 41. Brown, J. D., Brown, M. I., and Jones, F., “Evalu- 55. Cohn, S. D., “The Nurse-Midwife: Malpractice

ation of a Nurse Practitioner-Staffed Preventive and Risk Management, ” J. Nurse-Midwifery Medicine Program in a Fee-for-Service Mul- 29(5):316-321, September/October 1984. tispecialty Clinic, ” Prev. Med. 8(1):53-64, Janu- 56 Comptroller General of the United States, Report ary 1979. to the Congress, Progress and Problems in Train- 42. Brucker, M. C., and Muellner, M., “Nurse-Mid- ing and Use of Assistants to Primary Care Phy- wifery Care of Adolescents, ” J. Nurse-Midwifery sicians, Department of Health, Education, and 30(5):277-279, September/October 1985. Welfare, Washington, DC, Apr. 8, 1975. 43. Burosh, P., “Physicians’ Attitudes Towards 57. Crandall, L. A., Santulli, W. P., Radelet, M. L., 81

et al., “Physician Assistants in Primary Care: Pa- 70, Donabedian, A., Explorations in QuaZity Assess- tient Assignment and Task Delegation, ” Med. ment and Monitoring, Volume I: The Definition Care 22(3):268-282, March 1984. of Quality and Approaches to Its Assessment 58. Crowley, A. E., Etzel, S.1., and Petersen, E. S., (Ann Arbor, MI: Health Administration Press, “Undergraduate Medical Education, ” J. A.M.A. 1980). 252(12):1525-1532, Sept. 24, 1984. 71. Draye, M. A., and Stetson, L. A., “The Nurse 59. Davis, L., and Katz, B. M., “Knowledge and Atti- Practitioner as an Economic Reality, ” Nurse tude Toward Nurse-Midwifery Taking Childbirth Pract. 1(2):60-63, November/December 1975. Education Classes, ” -1. Nurse-Midwifery 24:18-26, 72. Duncan B., Smith, A. N., and Silver, H. K., Ju]y August 1979. “Comparison of the Physical Assessment of Chil- 60. DeLeon, P. H., Kjervik, D. K., Kraut, A. G., et al., dren by Pediatric Nurse Practitioners and Pedi- “Psychology and Nursing: A Natural Alliance, ” atricians, ” Am. J. Z%blic Health 60(6):1170-1176, An-t. F’sychol. 40(11 ):1153-1164, November 1985. June 1971. 61 Denton, F. T., Gafni, A., Spencer, B. G., et al., 73. Duttera, M. J., and Harlan, W. R., “Evaluation of “Potential Savings From the Adoption of Nurse Physician Assistants in Rural Primary Care, ” Practitioner Technology in the Canadian Health Arch. Intern. Med. 138(2):224-228, February Care System, “ Socio-Econ. Plan. Sci. 17(4):199- 1978. 209, 1983. 74. Eastaugh, S. R., “An Efficient Solution to Primary 62. Dentzler, S. and Tsuroka, D., “Malpractice In- Care Maldistribution, ” Hosp. Progress 62(2):32- surers Are Ill, ” IVewsweek 105:58, Apr. 29, 1985. 35, February 1981. 63. Desmairais, H. R., Acting Administrator, Office 7.5. Ebersole, P., Smith, A., Dickey, E. W., et al., of the Administrator, Health Care Financing “Roles and Functions of Geriatric Nurse Practi- Administration, U.S. Department of Health and tioners in Long Term Care as Viewed by Physi- Human Services, Washington, DC, Medicare cian, GNP and Administrator, ” American Health Reimbursement for Physician Services, Hearings Care Association Journal 8(2):2-7, March 1982. Before the Subcommittee on Health, House Com- 76. Ebersole, P., “Gerontological Nurse Practitioners mittee on Ways and Means, U.S. Congress, Apr. Past and Present, ” Geriatric Nursing, 219-222, 14, 1986. July/August 1985. 64. Desmairais, H. R., Acting Administrator, Office 77. Eggert, J., “Direct Assessment vs. Brokerage: A of the Administrator, Health Care Financing Comparison of Case Management Models, ” Administration, U.S. Department of Health and Monroe County Long-Term Care Program, Human Services, Washington, DC, personal Rochester, NY, Mar. 28, 1986. communication to The Honorable Robert W. 78. Ehrnman, K., Government Relations Coordina- Kasten, April 1986. tor, American College of Nurse-Midwives, Wash- 65 Dickstein, D., “Evaluation of the Group Health ington, DC, personal communication, May 28, Cooperative Demonstration Midwifery Service, ” 1985. report to The Henry J. Kaiser Family Foundation, 79. Ehrnman, K., Government Relations Coordina- Menlo Park, CA, 1983. tor, American College of Nurse-Midwives, Wash-

66< Diers, D., “Future of Nurse-Midwives in Amer- ington, DC, personal communication, Oct. 20, ican Health Care, ” Nursing in the 1980’s: Crises, 1986. Opportunities, and Challenges, L.H. Aiken and 80, Enggist, R. E., and Hatcher, M. E., “Factors Influ- S.R. Gortner (eds. ) (Philadelphia, PA: J.B. Lip- encing Consumer Receptivity to the Nurse Prac- pincott, 1982). titioner, ” J. Med. Syst. 7(6):495-512, December 67. Diers, D., and Burst, H. V., “Effectiveness of Pol- 1983. icy Related Research: Nurse-Midwifery as a Case 81. Feldbaum, E. G., “Will Nurses Alleviate Health Study, ” Image: The Journal of Nursing Scholar- Service Maldistribution Problems?” Nursing ship 15(3 ):68-74, summer 1983. Administration Quarterly 4(1):61-66, fall 1979. 68. Diggs, W. W., “Rural Distribution of Physicians 82. Feldman, R., Taller, S. L., Garfield, S. R., et al., in Tennessee, ” J. Term. Med. Assoc. 76(2):79-85, “Nurse Practitioner Multiphasic Health Check- February 1983. ups, ” Prev. Med. 6(3):391-403, September 1977. 69. DiGirol, M. T., and Parry, W. H., “Consultation 83, Finefrock, Williams, American Academy of Phy- to the Pediatric Automated Military Outpatient sician Assistants, Arlington, VA, personal com- Systems Specialist (AMOSIST): A Comparison munication, May 6, 1986. of Consultation by a Pediatric Clinical Nurse Spe- 84. Flynn, B. C., “The Effectiveness of Nurse Clini- cialist and by a Pediatrician, ” Nfilit. Med. 148(4): cians’ Service Delivery, ” Am. J. Publjc Health 364-367, April 1983. 64(6):604-611, June 1974. —- 82

85. Freedman, S. A., “Megacorporate Health Care: Providers?” Am. J. Nurs. 85(5):538-542, May A Choice for the Future, ” N. Engl. J. Med. 1985. 312(9):579-582, Feb. 28, 1985. 101. Griffith, H. M., Consultant, Lecturer, Researcher 86. Gairola, G. A., “Physician Assistant Graduates: on Health Policy and Nursing Economics, Fort Factors Related to Rural-Urban Practice Loca- Hood, TX, personal communication, July 15, tion, ” J. Community Health 8(1):23-32, fall 1982. 1986. 87. Gambert, S. R., Rosenkrantz, W. E., Basu, S. N., 102. Griffith, H. M., “Direct Third-Party Reimburse- et al., “Role of the Physician-Extender in the ment for Nursing Services, ” Nursing Economics Long-Term Care Setting, ” Wis. Med. J. 82(9):30- (forthcoming). 32, September 1983. 103. Harsham, P., “Midwifery: The Latest Growth In- 88< Garfield, S. R., Cohen, M. F., Feldman, R., et al., dustry, ” Soc. Sci. Med. [c) 60(10):232-245, May “Evaluation of an Ambulatory Medical Care- 16, 1983. Delivery System, “N. Engl. J. Med. 294:(8) 426- 104. Hastings, G. E., Vick, L., Lee, G., et al., “Nurse 431, Feb. 19, 1976. Practitioners in a Jailhouse Clinic, ” Med. Care 89 Gelot, D., Division of Nursing, Bureau of Health 18(7):731-744, July 1980. Professions, Health Resources and Services Ad- 105. Hausner, T., Lee, A. J., Venable, A., et al., “The ministration, Public Health Service, U.S. Depart- Urban Health Clinics Demonstration: A Review ment of Health and Human Services, Rockville, of the Relevant Literature, ” working paper, Ar- MD, personal communication, Aug. 20, 1986. thur D. Little, Inc., Cambridge, MA, May 1983. 90. Gladhart, S. C., Crespo, M., and Anderson, V., 106. Health Insurance Association of America, Source “The Physician Assistant in Kansas, ” J. Kans. Book oflfealth Insurance Data: 1982-83 (Wash- Med. Soc. 79:20, 1978. ington, DC: no date). 91. Goldbeck, W. B., President, Washington Business 107. Health Services Research Center, University of Group on Health, Washington, DC, personal North Carolina at Chapel Hill, The NationaZRu- communication, Apr. 18, 1985. ral Primary Care Evaluation Project: Issue- 92. Goldberg, G. A., Jolly, O. M., Hosek, S., et al., Oriented Data Analysis, final report to the Robert “Physician’s Extenders’ Performance in Air Force Wood Johnson Foundation (Chapel Hill, NC: Clinics, ” Med. Care 19(9):951-965, September April 1985). 1981. 108. Heiman, E. M., and Dempsey, M. K., “Independ- 93 Golden, A. S., and Cawley, J. F., “A National Sur- ent Behavior of Nurse Practitioners: A Survey of vey of Performance Objectives of Physician’s As- Physician and Nurse Attitudes, ” Am. J. Public sistant Training Programs, ” J. Med. Educ. Health 66(6):587-589, June 1976. 58(5):418-424, May 1983. 109. Henderson, M., “A GNP in a Retirement Com- 94. Graduate Medical Education National Advisory munity, ” Geriatric Nursing 5(2):109-112, March/ Committee, U.S. Department of HeaIth and Hu- April 1984. man Services, GMENAC Summary Report to the 110. Hershey, J. C., and Kropp, D. H., “A Re- Secretary (Washington, DC: 1980). appraisal of the Productivity Potential and Eco- 95. Gray, B., Project Director, Institute of Medicine, nomic Benefits of Physician’s Assistants, ” Med. National Academy of Sciences, Washington, DC, Care 27(6):592-606, June 1979. personal communication, May 1986. 111. Hershey, J. C., and Kropp, D. H., “Response to 96. Greaney, F. J., “Opportunities for Nurses in Uti- Ms. Major,” Med. Care 28(6):687-689, June 1980. lization Review, ” Nursing Economics 4(5):245- 112. Hicks, L. L., “Social Policy Implications of Phy- 246, 257, September/October 1986. sician Shortage Areas in Missouri, ” Am. J. Pub- 97. Greene, S., Senior Director, Health Economics lic Health 74(12):1316-1321, December 1984. Research, Blue Cross/Blue Shield of North Caro- 113. Hoekelman, R. A., Patterson, P. K., Bergman, lina, Durham, NC, personal communication, A. B., et al., “Parent Reaction to the Concept of Apr. 2, 1985. Pediatric Assistants, ” Pediatrics 44:69-75, July 98. Griffith, H. M., “Strategies for Direct Third-Party 1969. Reimbursement for Nurses,” Am. J. Nurs. 82(3): 114, Holmes, G. C., Livingston, G., Basset, R. E., et 408-411, March 1982. al., “Nurse Clinician Productivity Using a Rela- 99< Griffith, H. M., “A Case Study of the Implemen- tive Value Scale, ” HeaZth Serv. Res. 12(3):269- tation of Legislation Providing Direct Third Party 283, fall 1977.

Reimbursement for Services of Nurse Practi- 115< Hoyer, R., National Association for Home Care, tioners in Maryland, ” unpublished doctoral dis- Washington, DC, personal communication, May sertation, 1984. 16, 1986. 100. Griffith, H. M., “Who Will Become the Preferred 116, Huntington, C. G., General Partner, Hermon 83

Medical Group, Hermon, NY, personal commu- 131. Kubala, S., and Clever, L. H., “Acceptance of the nication, Apr. 9, 1985. Nurse Practitioner, ” Am. J. Nursing 74(3):451- 117 Huntington, C. G., General Partner, Hermon 452, March 1974. Medical Group, Hermon, NY, personal commu- 132. LaBar, C., Third-Party Reimbursement Legisla- nication, May 12, 1986. tion for Services of Nurses: A Report of Changes 118. Institute for International Health Initiatives, in State HeaZth Insurance Laws (Kansas City, Directory of Preferred Provider Organizations MO: American Nurses’ Association, 1984). and the Industry Report on PPO Development 133. Laudicina, S. S., “A Comparative Survey of Med- (Bethesda, MD: American Medical Care and Re- icaid Hospital Reimbursement Systems for In- view Association, June 1985). patient Services, State by State, 1980-1985,” 119. Institute of Medicine, National Academy of Sci- Intergovernmental Heal Project, George Wash- ences, Primary Care in Medicine: A Definition ington University, Washington, DC, 1985. (Washington, DC: National Academy Press, 134 Lawrence, R. S., DeFriese, G. H., Putnam, S. M., 1977). et al., “Physician Receptivity to Nurse Practi- 120 Institute of Medicine, National Academy of Sci- tioners: A Study of the Correlates of Delegation ences, Nursing and Nursing Education: Public of Clinical Responsibility, ” Med. Care 15(4):298- Policies and Private Actions (Washington, DC: 310, April 1977. National Academy Press, 1983). 135. Lazarus, W., Levine, E. S., Lewin, L. S., et al., 121, Institute of Medicine, National Academy of Sci- Competition Among Health Practitioners: The ences, Preventing Low Birth weight (Washington, Influence of the Medical Profession on the Health DC: National Academy Press, 1985). Manpower Market, Volume II: The Childbear-

122< Institute of Medicine, National Academy of Sci- ing Center Case Study, report prepared for the ences, Committee on Nursing Home Regulation, Federal Trade Commission (Washington, DC: Improving the Quality of Care in Nursing Homes Federal Trade Commission, February 1981). (Washington, 13C: National Academy Press, 1986). 136. Legislative Network for Nurses, “AMA Resolves 123. Interstudy, Inc., National HMO Census 2985 (Ex- to Fight Expanding Practices of Nurses, ” Legis- celsior, MN: Interstudy, Inc., 1986). lative Network for Nurses 2(17):1, Aug. 22, 1985. 124. Jean, G. L., Brovender, S. R., Freeland, R., et al., 137. Leiken, A. M., “Factors Affecting the Distribution “Geriatric Nurse Practitioners Impact Long Term of Physician Assistants in New York State: Pol- Care, ” presented at the XII International Congress icy Implications, “J. Public Health Policy 6(2): of Gerontology, Hamburg, Germany, July 12-17, 236-243, June 1985. 1981. 138. LeRoy, L., and Solkowitz, S., The Costs and Ef- 125. Johnson, R. E., Freeborn, O. K., Lee, G., et al., fectiveness of Nurse Practitioners (Health Tech- “Delegation of Office Visits in Primary Care to nology Case Study #16), prepared for the Office PAs and NPs: The Physicians’ View, ” Physician of Technology Assessment, U.S. Congress, OTA- Assistant 9(1):159-169, January 1985. HCS-16 (Washington, DC: U.S. Government 126. Kane, R. L., Geriatrics in the United States (Lex- Printing Office, August 1980). ington, MA: D.C. Heath & Co., 1981). 139. Levine, D. M., Morlock, L. L., Mushlin, A. I., et 127. Kane, R. L., Gardner, J., Wright, D. D., et al., al., “The Role of New Health Practitioners in a “Differences in the Outcomes of Acute Episodes Prepaid Group Practice: Provider Differences in of Care Provided by Various Types of Family Process and Outcomes of Medical Care, ” Med. Practitioners, ” J. F’am. Prac, 6(1):133-138, June Care 14(4):326-347, April 1976. 1978. 140. Levy, B. S., Wilkinson, F.S, and Marine, W. M., 128. Kane, R. L., Jorgensen, L. A., Teteberg, B,, et al., “Reducing Neonatal Mortality Rates With Nurse “Is Good Nursing Home Care Feasible?” J. A.M.A. Midwives, ” Am. J. Obstet. Gynecol. 109(1):50- 235(5):516-519, Feb. 2, 1976. 58, Jan. 1, 1971. 129. Kane, R. L., Olsen, D. M., and Castle, C. H., “Me- 141. Lewis, C. E., and Resnick, B. A., “Nurse Clinics dex and Their Physician Preceptors: Quality of and Progressive Ambulatory Patient Care, ” N. Care, ” J. A.M.A. 236(22):2509-2512, Nov. 28, Engl. J. Med. 277(23):1236-1241, Dec. 7, 1967, 1976. 142. Lewis, C. E., Resnick, B. A., Schmidt, G., et al., 130 Komaroff, A. L., Sawayer, K., Flatley, M., et al., “Activities, Events and Outcomes in Ambulatory “Nurse Practitioner Management of Common Patient Care,” N. EngZ. J. Med. 280(12):645-649, Respiratory and Genitourinary Infections, Using Mar. 20, 1969. Protocols, ” Nurs. I/es. 25(2):84-89, March/April 143. Light, D. W., “Is Competition Bad?” N. Engl. J. 1976. Med. 309(21):1315-1318, Nov. 24, 1983. .—

144 Light, D. W., “The Impact of Competition on for the Quality of Life in a Rural Count y,” Bulle- Physician Surplus: GMENAC Revisited, ” LDI tin of Nurse-Midwifery I7(4):103-II3, Novem- policy discussion paper No. 4, Leonard Davis In- ber 1972. stitute of Health Economics, University of Penn- 159. Mendenhall, R., and Repicky, P. A., Collection sylvania, Philadelphia, PA, 1984. and Processing of Baseline Data for the Physician 145. Linn, L. S., “Patient Acceptance of the Family Extender Reimbursement Study, Pub. No. (HRA) Nurse Practitioner, ” A4ed. Care 14(4):357-364, 76-43 (Washington, DC: U.S. Government Print- April 1976. ing Office, 1978). 146, Lohr, K, N., and Brook, R, H., “Quality Assur- 160. Mendenhall, R. C., Repicky, P. A., and Neville, ance in Medicine, ” Am. Behav. Scientist 27(5): R. E., “Assessing the Utilization and Productivity 583-607, May/June 1984. of Nurse Practitioners and Physician’s Assistants: 147. Lohrenz, F., Payne, R., Intress, R., et al., “Place- Methodology and Findings on Productivity, ” ment of Primary Care Physicians Assistants in Med. Care 18(6):609-623, June 1980. Small Rural Communities, ” Wis. Med. J. 75:593, 161. Michels, K. A., Assistant Director, Congressional 1976. and Agency Relations, American Nurses’ Asso- 148. Lubic, R. W., “Evaluation of an Out-of-Hospital ciation, Washington, DC, personal communica- Maternity Center for Low Risk Patients, ” in tion, June 17, 1986. Health Policy and Nursing Practice, L.H. Aiken 162. Miller, F. N., “Nurse Providers: A Resource for (cd. ) (New York, NY: McGraw-Hill Book Co., Growing Population Needs, ” Business Health 1980). 2(3):38-42, January/February 1985. 149. Lubic, R, W., “Childbirthing Centers: Delivering 163. Miller, G., Vice President, Practice Development, More for Less, ” Am. J. Nurs. 83(7):1053-1056, Humana, Inc., Louisville, KY, personal commu- July 1983. nication, May 16, 1986. 150. Lubic. R. W., Director, Maternity Center Asso- 164. Milstead, D., Health Care Financing Administra- ciation, New York, NY, personaI communication, U.S. Department of Health and Human tion, June 9, 1986. Services, Baltimore, MD, personal communica- 151. Lynaugh, J., Primary Care Family Nurse Clini- tion, March 1985. cal Program, University of Pennsylvania, Phila- 165. Mitchell, J., President, Health Economic Re- delphia, PA, personal communication, January search, Inc., Medicare Reimbursement for Phy- 1986. sician Services, Hearings Before the Subcommit- 152. MagiII, G., and Hartze, L., Immediate Care tee on Health, House Committee on Ways and Centers: Fast Medicine for the ‘80s, DHHS Pub. Means, U.S. Congress, Apr. 14, 1986. No. HRP-0905987 (Washington, DC: U.S. Gov- 166. Mitchell, J. B., and Hewes, H., Medicare Access ernment Printing Office, November 1984). to Physician Services in Nursing Homes, final re- 153. Major, E. E., “A Comment on ‘A Re-Appraisal of port prepared for the Health Care Financing Ad- the Productivity Potential and Economic Bene- ministration, U.S. Department of Health and Hu- fits of Physician’s Assistants, ’ “ Med. Care man Services, Baltimore, MD, Nov. 22, 1982. 18(6):686-687, June 1980. 167. Mitchell, K., “NAPNAP’s Scope of Practice Sur- 154. Manber, M. M., “NPs, MDs, and PAs: Meshing vey: Results, Revisions, and Issues, ” Peal. Nurs. Their Changing Roles, ” ,’Wed. World News 9(3):199-203, May/June 1983. 26(18):53-71, Sept. 23, 1985. 168. Morgan, W. A., and Sullivan, N. D., “Nurse Prac- 155. Master, R. J., Feltin, M., Jainchill, J., et al., “A titioner and Physician’s Assistant Clinics in Ru- Continuum of Care for the Inner City: Assess- ral California, Part II: A Survey, ” West. J. Med. ment of Its Benefits for Boston’s Elderly and High- 132(3):259-264, March 1980. Risk Populations, “N. Eng/. J. Med. 302(26):1434- 169. Morris, S. B., and Smith, D. B., “The Distribution 1440, June 26, 1980. of Physician Extenders, “ Med. Care 15(12):1045- 156, McKaig, C., Hindi-Alexander, M., Myers, T. R., 1057, December 1977. et al., “Implementation of the School Nurse Prac- 170. Morrisey, M. A., Allen, J., and Shorten, S., titioner Role: Barriers and Facilitators,’r J. of Occup. “Medical Staff Size, Hospital Privileges and Com- Safety & Hea]th 54(1):21-23, January 1984. pensation Arrangements, ” For Profit Enterprises 157. McDermott, S. M., Mountain States Health Corp., in HeaJth Care, Institute of Medicine, National Boise, ID, personal communication, Apr. 25, Academy of Sciences (Washington, DC: National 1986. Academy Press, June 1986). 158. Meglan, M. C., “A Prototype of Health Services 171. Morrisey, M. A., and Brooks, D. C., “The Ex- 85

panding Medical Staff: Nonphysician Providers, ” “Nurse Practitioner and Physician Care Com- Hospitals 59(15):58-59, Aug. 1, 1985. pared for Nonurgent Emergency Room Patients, ” 172. National Association for Ambulatory Care, “An- Nurse Pract. 9(2):39-52, February 1984. nual Survey: Compensation Practices and Em- 187. Prescott, P. A., and Driscoll, L., “Evaluating ployee Benefits in Ambulatory Care Centers, ” Nurse Practitioner Performance, ” Nurse Pract. prepared by Commerce International, Inc., 5(4):28-32, July/August 1980. Washington, DC, September 1985. 188. Purdy, C. J., Merenstein, G. B., and Walls, M. J. I., 173. Nelson, E. C., Jacobs, A. R., and Johnson, K. G., “Retrospective Audit of the Pediatric Nurse Prac- “Patients’ Acceptance of Physician’s Assistants, ” titioner in the Normal Newborn Nursery, ” Peal. J. A.M.A. 228(1):63-67, Apr. 1, 1974. Nurs. 5(6):55-57, November/December 1979. 174. Newhouse, J. P., Williams, A. P., Bennett, B. W., 189. Ramsay, J. A., McKenzie, J. K., and Fish, D. G., et al., “Where Have All the Doctors Gone?” “Physicians and Nurse Practitioners: Do They J. A.M.A. 247(17):2392-2396, May 7, 1982. Provide Equivalent Health Care?” Am. J. Public 175. Ollivier, S., Lesser, C., and Bell, K. B., “Provid- Health 72(1):55-57, January 1982. ing Infertility Care, ” J. of Geriatric Nurs. (Sup- 190. Record, J. C., and Cohen, H. R., “The Introduc- plement) 13(2) :85s-90s, March/April 1984. tion of Midwifery in a Prepaid Group Practice, ” 176. Owens, A., “Doctor Surplus: Where Things Am. J. Public Health 62(3):354-360, March 1972. Stand Now, ” Soc. Sci. Med. /c] 57(20):63-80, 191. Record, J. C., and Greenlick, M., “New Health Sept. 29, 1980. Professionals and the Physician Role: An Hy- 177. Penner, R. G., Director, Congressional Budget pothesis From Kaiser Experience, ” Public Health Office, U.S. Congress, Washington, DC, per- Reports 90(3):241-246, June 1975. sonal communication [o The Honorable Ron 192. Record, J. C., McCally, M., Schweitzer, S.0., et Wyden, Nov. 5, 1984. al., “New Health Professions After a Decade and 178. Perrin, E. C., and Goodman, H. C., “Telephone a Half: Delegation, Productivity and Costs in Pri- Management of Acute Pediatric Illnesses, ” N. mary Care, ” J. Health Polit. Policy Law 5(3):470- Engl. J. Med. 298(3):130-135, Jan. 19, 1978. 497, fall 1980. 179. Perry, H. B., “An Analysis of the Professional 193. Reid, M. L., and Morris, J. B., “Perinatal Care and Performance of Physician’s Assistants,” J. Med. Cost-Effectiveness: Changes in Health Expendi- Educ. 52(8):639-647, August 1977. tures and Birth Outcome Following the Establish- 180. Perry, H. B., “An Analysis of the Specialty and ment of a Nurse-Midwife Program, ” Med. Care Geographic Location of Physician Assistants in 17(5):491-500, May 1979. the United States, ” Am. J. Public Health 194. Renn, S. C., Schramm, C. J., Watt, J. M., et al., 68(10):1019-1021, October 1978. “The Effects of Ownership and System Affilia- 181. Perry, H. B., “Role of the Nurse-Midwife in Con- tion on the Economic Performance of Hospitals, ” temporary Maternity Care, ” Psychosomatic Ob- Inquiry 22(3~:219-236, fall 1985. stetrics and Gynecology, D.D. Youngs and A.A. 195. Richards, S. J., and DeCastro, F. J., “Communi- Ernhardt (eds. ) (New York: Appleton-Century- cation With Patients: A Parameter in Evaluating Crofts, 1980). Nurse Practitioners, ” Mo. Med. 70(10):719-720, 182. Perry, H. B., and Breitner, B., Physician Assis- October 1973. tants: Their Contribution to Health Care (New 196. Riessman, C. K., “The Use of Health Services by York: Human Sciences Press, 1982). the Poor: Are There Any Promising Models?” So- 183. Perry, H. B., and Redmond, E. L., Deployment cial Policy 14(4):30-40, spring 1984. and Career Trends of Physician Assistants, final 197. Robert Wood Johnson Foundation, Updated Re- report to the National Center for Health Services port on Access to Health Care for the American Research (Springfield, VA: National Technical In- People: Special Report (Princeton, NJ: 1983). formation Service, 1980). 198. Rodeghero, J. A., “Perceptions of Nurse Practi- 184. Piechnik, S. L., and Corbett, M. A., “Reducing tioners Among the Rural Poor, ” Nurse Pract. Low Birth Weight Among Socioeconomically 3(5):7, September/October 1978. High-Risk Adolescent Pregnancies, ” J. ofNurse- 199. Romm, J., Berkowitz, A., Cahn, M. A., et al,, Midwifery 30(2):88-98, March/April 1985. “The Physician Extender Reimbursement Exper- 185. Pollard, M. R., Director, Office of Policy Anal- iment, ” J. Amb. Care Management 2(2):1-12, ysis, Pharmaceutical Manufacturers Association, May 1979. Washington, DC, personal communication, Apr. 200. Rooks, J. P., “President’s Pen, ” Quickening 2, 1985. 14(4):1-2, July/August 1983. 186. Powers, M. J., Jalowiec, A., and Reichelt, P. A., 201. Rooks, J. P., President, American College of 86

Nurse-Midwives, Washington, DC, personal isons for Physicians and Nurse Practitioners, ” communication, June 5, 1986. Med. Care 20(2):143-153, February 1982. 202. Rooks, J. P., “Cost-Effectiveness of Nurse- 214. Schafft, G., American Academy of Physician As- Midwifery Care, ” Nurse-Midwifery in America, sistants, Arlington, VA, personal communica- J. Rooks and E. Hass (eds. ) (Washington, DC: tion, Aug. 20, 1986. American College of Nurse-Midwives Founda- 215. Schafft, G., Fasser, C. E., and Cyr, A. B., “The tion, 1986). Assessment and Improvement of Knowledge and 203, Rooks, J. P., and Fischman, S. H., “American Skills in Geriatrics for Physician Assistants, ” pre- Nurse-Midwifery in 1976-1977: Reflections of 50 pared under Grant No. (OHDS) 90-AT-0094, Years of Growth and Development, ” Am J. Pub- Washington, DC, June 1985. lic Health 70(9):990-996, September 1980. 216. Scharon, G. M., and Bernacki, E. J., “A Corporate 204. Rosenaur, J., Stanford, D., Morgan, W., et al., Role for Nurse Practitioners, ” Business and “Prescribing Behaviors of Primary Care Nurse Health 1(9):26-27, September 1984. Practitioners,” Am. J. Public HeaZth 74 (1):10-13, 217. Scheffler, R. M., “Physician Assistants: Is There January 1984. a Return to Training?” Industrial Relations 205. Rosenberg, C., “Nurse/Physician Relations: A 14(1):78-89, February 1975. Perspective From Medicine, ” Bull. N. Y. Acad. 218. Scheffler, R. M., et al., “Physicians and New Med. 60(8):807-810, October 1984. Health Professions: Issues for the 1980’ s,” Inquiry 206. Rossiter, L. F., “Prospects for Medical Group 16(3):195-229, fall 1979. Practice Under Competition, ” Med. Care 22(l): 219. Schroeter, J., Information Specialist, American 84-92, January 1984. Academy of Physician Assistants, Arlington, VA, 207. Ruby, G., “Consumer Acceptance of Nurse Prac- personal communication, May 20, 1985. titioners and Physician Assistants, ” staff paper 220. Schultz, P. R., and McGlone, F. B., “Primary prepared for the report, “A Manpower Policy for Health Care Provided to the Elderly by a Nurse Primary Care, ” Institute of Medicine, National Practitioner/Physician Team: Analysis of Cost- Academy of Sciences, Washington, DC, Jan. 31, Effectiveness, “J. Am. Geriat. Soc. 25(10):443- 1977. 446, October 1977. 208. Ruby, G., “Physician Acceptance of Nurse Prac- 221. Schwartz, J. L., “Economic Feasibility and Patient titioners and Physician Assistants, ” staff paper Diagnostic Mix of Family Nurse Practitioners, ” prepared for the report, “A Manpower Policy for PubZic Health Rep. 94(2):148-155, March/April Primary Care, ” Institute of Medicine, National 1979. Academy of Sciences, Washington, DC, June 20, 222. Schwartz, W. B., Newhouse, J. P., Bennett, B. W., 1977. et al., “The Changing Geographic Distribution 209. Ruby, G., “New Health Professionals in Child of Board-Certified Physicians, ” N. Engl. J. Med. and Maternal Health, ” Better ~ealt~ for Our 303(18):1032-1038, oct. 30, 1980. Children: A National Strategy, Report of the Se- 223. Sharpe, T. R., and Banahan, B. F., “Evaluation of lect Panel for the Promotion of Child Health, the Use of Rural Health Clinics: Attitudes and Be- Volume W, U.S. Department of Health and Hu- haviors of Primary Care Physicians in Service man Services, Public Health Service Pub. No. 79- Areas of Nurse Practitioner Clinics, ” PubZic 55071 (Washington, DC: 1981). Health Rep. 97(6):566-571, November/December 210. Runnerstrom, L., “The Effectiveness of Nurse- 1982. Midwifery in a Supervised Hospital Environ- 224. Silver, H. K., Murphy, M. A., and Gitterman, merit, ” Bulletin of the American College of Nurse B. A., “The Hospital Nurse Practitioner in Pedi- Midwives 14(2):40-52, May 1969. atric Practice, ” AJDC 138:237-239, March 1984. 211. Runyon, J. W., “The Memphis Chronic Disease 225. Simborg, D. W., Starfield,B.H., and Horn, S. D., Program: Comparisons in Outcome and the “Physicians and Non-Physician Health Practi- Nurse’s Extended Roles,” J. A.M.A. 231(3):264- tioners: The Characteristics of Their Practices and 270, Jan. 20, 1975. Their Relationships, ” Am. J. Public Health 212. Sackett, D. L., “The Burlington Randomized Trial 68(1):44-48, January 1978. of the Nurse Practitioner: Health Outcomes of 226. Slome, C., Wetherbee, H., Daly, M., et al., “Ef- Patients,” Ann. Intern. Med. 80(2):137-142, Feb- fectiveness of Certified Nurse-Midwives: A Pro- ruary 1974. spective Evaluation Study, ” Am. J. Obstet. Gy- 213. Salkever, D. S., Skinner, E. A., Steinwachs, necol. 124(2):177-182, Jan. 15, 1976. D. M., et al., “Episode-Based Efficiency Compar- 227, Smith, C. W., “Patient Attitudes Toward Physi- 87

cians’ Assistants, ” J. Farz-t. Pract. 13(2):201-204, Practice of Medicine,” N. Engl. J. Med. 308(20): August 1981. 1235-1244, May 19, 1983. 228. Sobolewski, S. D., “Cost-Effective School Nurse 241. Tresnowski, B. R., “HMOS Are Rolling—and We Practitioner Services, ” J. Sch. Health 51(9):585- Can All Benefit,” Inquiry 21(3):203-204, fall 1984. 588, November 1981. 242. U.S. Congress, Congressional Budget Office, 229, Solomon, S., Director, National League for Nurs- Physician Extenders: Their Current and Future ing, New York, NY, personal communication, Role in Medical Care DeZivery (Washington, DC: Aug. 25, 1986. U.S. Government Printing Office, April 1979). 230. Sex, H. C., Jr., “Quality of Patient Care by Nurse 243 U.S. Congress, Library of Congress, Congres- Practitioners and Physician’s Assistants: A Ten- sional Research Service, ‘The Civilian Health and Year Perspective, “ Arm. Intern. Med. 91(3):459- Medical Services (CHAMPUS) and the Civilian 68, September 1979. Health and Medical Program of the Veterans 231. Spitzer, W. O., Sackett, D. L., Sibley, J. C., et al., Administration (CHAMPVA): An Overview, “The Burlington Randomized Trial of the Nurse UI-1460, 76-143 ED, Washington, DC, July 27, Practitioner, “ N. Engl. J. Med. 290(5):251-256, 1976. Jan. 31, 1974. 244 U.S. Congress, Office of Technology Assessment, 232. Stallmeyer, J., American Nurses’ Association, Medical Technology and Costs of the Medicare “Direct Reimbursement for Non-Physician Health Program, OTA-H-227 (Washington, DC: U.S. Professionals Under FEHB, ” Hearings Before the Government Printing Office, July 1984). Subcommittee on Compensation and Employee 245< U.S. Congress, Office of Technology Assessment, Benefits, House Committee on Post Office and Technology and Aging in America, OTA-BA-264 Civil Service, U.S. Congress, Washington, DC, (Washington, DC: U.S. Government Printing Of- Apr. 15, 1986. fice, June 1985). 233. Stein, J., “Nursing Can Gain From Prospective 246. U.S. Congress, Office of Technology Assessment, Payment, ” Business and Health 1(5):36, May Payment for Physician Services: Strategies for 1984. Medicare, OTA-FI-294 (Washington, DC: U.S. 234. Stein, R. E., and Jessop, D, J., “Does Pediatric Government Printing Office, February 1986). Home Care Make a Difference for Children With 247. U.S. Department of Health, Education, and Wel- Chronic Illness? Findings From the Pediatric Am- fare, “The Effect of Task Delegation on the Re- bulatory Care Treatment Study, ” Pediatrics quirements for Selected Health Manpower Cate- 73(6):845, June 1984. gories in 1980, 1985, and 1990, ” final report 235. Streimer, R. A., Deputy Director, Bureau of Eligi- contract No. NIH 72-44006, Washington, DC, bility, Reimbursement, and Coverage, Health 1974. Care Financing Administration, U.S. Department 248. U.S. Department of Health and Human Services, of Health and Human Services, Baltimore, MD, Health Care Financing Administration, Medi- personal communication, 1985. care’s Carrier Manual (Baltimore, MD: n.d. ). 236. Sultz, H. A., “Nurse Practitioners: An Overview 249. U.S. Department of Health and Human Services, of Nurses in the Expanded Role, ” The New Health Public Health Service, Health Research Services Professionals: Nurse Practitioners and Physicians Administration, Registered Nurse Population and Assistants, A.A. Bliss and E.D. Cohen (eds. ) Overview, DHHS Pub. No. HRS-P-OD-83-l (Germantown, MD: Aspen Systems Corp., 1977’), (Washington, DC: U.S. Government Printing Of- 237. Sultz, H. A., Henry, O. M., Bullough, B., et al., fice, November 1982). “Nurse Practitioners: A Decade of Change—Part 250. U.S. Department of Health and Human Services, IV, ” Nurs. Outlook 32(3):158-163, May/June Public Health Service, Health Resources and 1984. Services Administration, Diffusion and the 238. Sultz, H. A., Zielezny, M., Gentry, J. M., et al., Changing Geographic Distribution of Primary Longitudinal Study of Nurse Practitioners, Phase Care Physicians, DHHS Pub. No. HRS-P-OD- 111, DHEW Pub. No. HRA-80-2 (Washington, 84-1, HRP 0904702 (Washington, DC: U.S. Gov- DC: U.S. Government Printing Office, 1980). ernment Printing Office, June 1983; revised No- 239. Swoap, D. B., “Beyond DRGs: Shifting the Risk vember 1983). to Providers, ” Health Affairs 3(4):117-121, win- 251 U.S. Department of Health and Human Services, ter 1984. Public Health Service, Bureau of Health Profes- 240. Tarlov, A. R., “Shattuck Lecture—The Increas- sions, Report to the President and Congress on ing Supply of Physicians, the Changing Structure the Status of Health Personnel in the United of the Health Services System, and the Future States, Vols. I and 11, DHHS Pub. No. HRS-P- 88

OD 84-4 (Washington, DC: U.S. Government titioner and Physician Adherence to Standing Printing Office, 1984). Orders Criteria for Consultation or Referral, ” 252. U.S. Department of Health and Human Services, Am. J. Public Health 72(1):22-29, January 1982. Public Health Service, Health Resources and 260. Watt, J. M., Derzon, R. A., Renn, S. C., et al., Services Administration, Bureau of Health “The Comparative Economic Performance of Professions, Division of Nursing, “1984 National Investor-Owned Chain and Not-for-Profit Hos- Sample Survey of Registered Nurses, ” Rockville, pitals,” IV. Eng. J. Med. 314(2):89-96, Jan. 9, MD, 1984. 1986. 253. U.S. Department of Health and Human Services, 261. Weston, J. L., “Distribution of Nurse Practitioners Public Health Service, National Center for Health and Physician Assistants: Implications of Legal Statistics, A4jdwjfe and Out-of-Hospital Deliver- Constraints and Reimbursement, ” Public Health ies: United States, Vital and Health Statistics Ser- Rep. 95(3):253-258, May/June 1980. ies 21, No. 40, DHHS Pub. No. 84-1918 (Wash- 262. Weston, J. L., “NPs and PAs: Changes—Where, ington, DC: U.S. Government Printing Office, Whether, and Why,” National Center for Health February 1984). Services Research, Office of the Assistant Secre- 254. U.S. Department of Health and Human Services, tary for Health, Public Health Service, U.S. De- Public Health Service, National Institute on partment of Health and Human Services, Wash- Aging, Report on Education and Training in Ger- ington, DC, November 1984. iatrics and Gerontology, Administrative Docu- 263. Weston, J. L., Senior Research Manager, National ment, February 1984. Center for Health Services Research, Public 255. U.S. Department of Health and Human Services, Health Service, U.S. Department of Health and Public Health Service, Health Resources and Human Services, Washington, DC, personal Services Administration, Fifth Report to the communication, Apr. 2, 1985. President and the Congress on the Status of 264. Williams, A. P., Schwartz, W. B., Newhouse, Health Personnel in the United States, DHHS J. P., et al., “How Many Miles to the Doctor?” Pub. No. HRS-P-OD-86-1, HRP 0906767 (Wash- N. Engl. J. Med. 309(16):958-963, oct, 20, 1983. ington, DC: U.S. Government Printing Office, 265, Williams, C. A., “Nurse Practitioners: Efficacy, March 1986). Barriers to Practice, and Action Alternatives, ” 256. U.S. Office of Personnel Management, Compen- prepared for a Frank Porter Graham Child De- sation Group, “A Study Relating to Expanding velopment Institute Conference, Chapel Hill, NC, the Class of Health Practitioners Authorized To 1983. Receive Direct Payment or Reimbursement in 266, Wright, D. D., Kane, R. L., Snell, G. F,, et al., Accordance With 5 U.S.C. 8902( k)(l),” Washing- “Costs and Outcomes for Different Primary Care ton, DC, March 1986. Providers, ’’].A.A4.A. 238(1):46-50, July 4, 1977. 257. Urban Medical Group, Inc., “Improving Care 267, Zikmund, W. G,, and Miller, S. J., “A Factor and Reducing Costs for Nursing Home Residents: Analysis of Attitudes of Rural Health Consumers A Fact Sheet, ” Jamaica Plain, MA, Feb. 5, 1985. Toward Nurse Practitioners, ” Res. Nurs. Health 258, Utah Department of Health, Office of Manage- 2(2):85-90, June 1979. ment Planning, Bureau of Rural Health Services, 268, Zimmer, J. G., Groth-Juncker, A., and McCus- “Midlevel Practitioners Manpower Study,” draft, ker, J., “Effects of a Physician-Led Home Care Salt Lake City, UT, May 6, 1985. Team on Terminal Care, ” J. Am. Geriat. Soc. 259< Watkins, L. O., and Wagner, E. H., “Nurse Prac- 32(4):288-292, April 1984.