Improving Community Ambulation After Hip Fracture: Protocol for a Randomized Controlled

The CAP Study

Improving Community Ambulation after Hip Fracture: Protocol for a randomized controlled trial comparing two physical therapy interventions

Authors: D Orwig1, KK Mangione2, M Baumgarten1, M Terrin1, R Fortinsky3, AM Kenny4, AL Gruber-Baldini1, B Beamer5,6, ANA Tosteson7,8, M Shardell9, L Magder1, E Binder10, K Koval11, B Resnick12, RL Craik2, J Magaziner1

Institutional affiliations: 1Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD; 2Department of Physical Therapy, College of Health Sciences, Arcadia University, Glenside, PA; 3UConn Center on Aging, UConn Health, Farmington, CT; 4Department of Medicine, UConn Health, Farmington, CT; 5Gerontology Research, Education and Clinical Center (GRECC) at Baltimore Veterans Affairs Medical Center, and 6Division of Geriatric Medicine and Gerontology at University of Maryland School of Medicine; 7Department of Medicine and 8The Dartmouth Institute for Health Policy and Clinical Practice, Geisel School of Medicine at Dartmouth, Hanover, NH; 9National Institute on Aging, National Institutes of Health, Baltimore, MD; 10Division of Geriatrics and Gerontology, Washington University School of Medicine, St Louis, MO; 11Department of Orthopaedic Surgery, Orlando Regional Medical Centre, Orlando, FL; 12University of Maryland School of Nursing, Baltimore, MD.

ABSTRACT: 730 words

Introduction: After a hip fracture in older persons, significant disability often remains; dependency in functional activities persists beyond three months post-surgery. Endurance, dynamic balance, quadriceps strength, and function are compromised and contribute to failure of hip fracture patients’ ability to walk independently in the community. In the United States, people 65 years and older are eligible to receive Medicare funding for physical therapy for a limited time after a hip fracture. A goal of outpatient physical therapy is independent, safe household ambulation two to three months after surgery. Current Medicare-reimbursed post-hip fracture rehabilitation fails to return many patients to pre-fracture levels of function. Interventions delivered in the home after usual hip fracture physical therapy care ends could promote higher levels of functional independence in these frail, older adult patients.

Primary Objective: To evaluate the effect of a specific multi-component physical therapy intervention (PUSH), compared to a non-specific multi-component control physical therapy intervention (PULSE), on the ability to ambulate independently in the community 16 weeks post- randomization.

Design: Parallel, two-group randomized multi-center trial of 210 older adults with a hip fracture assessed at baseline and at 16 weeks post-randomization, and at 40 weeks post randomization for a subset of approximately 150 participants.

Protocol Version 10.0 (DSMB approval 23JUN2016) 1 The CAP Study

Participants and Setting: 210 hip fracture patients are being enrolled at three clinical sites and randomized up to 26 weeks post-admission. Study inclusion criteria are: 1) Closed, non- pathologic, minimal trauma hip fracture with surgical fixation; 2) Age ≥60 years at time of randomization; 3) Community-residing at time of fracture and randomization; 4) Ambulating without human assistance two months prior to fracture; and 5) Unable to walk at least 300 m in six minutes at baseline. Participants are ineligible if the interventions are deemed to be unsafe or not feasible, or if the participant has low potential to benefit from the interventions.

Interventions: Participants are randomly assigned to one of two multi-component treatment groups: 1) PUSH or 2) PULSE. PUSH is based on aerobic conditioning, specificity of training, and muscle overload, while PULSE includes transcutaneous electrical nerve stimulation, flexibility activities, and active range of motion exercises. Participants in both groups receive 32 visits in their place of residence from a study physical therapist (PT) (2 visits per week on non- consecutive days for 16 weeks). PTs’ adherence to the treatment protocol, and participants’ receipt of the prescribed activities, are assessed. Participants also receive counseling from a registered dietician and vitamin D, calcium and multivitamin supplements during the 16-week intervention period.

Measurements: The primary outcome (community ambulation) is operationalized as the ability to walk 300 m or more in six minutes, as assessed by the Six-Minute Walk Test, at 16 weeks post-randomization. Other measures at 16 and 40 weeks include cost-effectiveness, endurance, dynamic balance, walking speed, quadriceps strength, lower extremity function, activities of daily living, balance confidence, quality of life, physical activity, depressive symptoms, increase of ≥50 m in distance walked in six minutes, cognitive status, and nutritional status.

Analysis: Analyses for all aims will be performed according to the intention-to-treat paradigm. Except for testing of the primary hypothesis, all statistical tests will be two-sided and not adjusted for multiple comparisons. The test of the primary hypothesis (comparing groups on the proportion who are community ambulators at 16 weeks post-randomization) will be based on a one-sided 0.025-level hypothesis test using a procedure consisting of four interim analyses and one final analysis with critical values chosen by a Hwang-Shih-Decani alpha-spending function. Analyses will be performed to test group differences on other outcome measures and to examine the differential impact of PUSH relative to PULSE in subgroups defined by pre-

Protocol Version 10.0 (DSMB approval 23JUN2016) 2 The CAP Study selected participant characteristics. Generalized estimating equations will be used to explore possible delayed or sustained effects in a subset of participants by comparing the difference between PUSH and PULSE in the proportion of community ambulators at 16 weeks with the difference at 40 weeks.

Discussion: This multi-center randomized study will be the first to test whether a home-based multi-component physical therapy intervention targeting specific precursors of community ambulation (PUSH) is more likely to lead to community ambulation than a home-based non- specific multi-component PT intervention (PULSE) in older adults after hip fracture. The study will also estimate the potential economic value of the interventions.

Trial registration: clinicaltrials.gov Registration number: NCT01783704 Was this trial prospectively registered? Yes Date of trial registration: 01/31/2013 Funded by: National Institute on Aging Funder approval number: R01 AG035009 Anticipated completion date: April 2018

Corresponding author contact details: Denise Orwig, Ph.D. Associate Professor Director, Baltimore Hip Studies University of Maryland School of Medicine Department of Epidemiology & Public Health 660 West Redwood Street, Suite 200 Baltimore, MD 21201 410-706-2406 410-706-4433 (fax)

Improving Community Ambulation after Hip Fracture: Protocol for a randomized controlled trial comparing two physical therapy interventions

Human research approval committees: University of Maryland, Baltimore Institutional Review Board (Study Oversight): HP-00046956 Arcadia University Institutional Review Board: 12-08-01 University of Maryland, Baltimore Institutional Review Board (University of Maryland, Baltimore Clinical Site): HP-00052790 University of Connecticut Health Center Institutional Review Board: IE-13-041HSF-3

INTRODUCTION

In the US, standard Medicare-reimbursed rehabilitation therapy fails to restore community ambulation to older persons who have had a hip fracture. Residual mobility disability similar to that reported after stroke occurs in the majority of persons who “recover” from hip fracture.1,2

Even though there are over 325,000 hip fractures per year in the US3,4 with a predicted increase to over 650,000 per year by 2040,5 there is a paucity of intervention trials in this patient population. The estimated cost to hip fracture patients, their families, and the health care system is between $14 and $20 billion annually.6-11

The goal of current rehabilitation practice is independent, safe household ambulation two to three months after surgery.12 Hip fracture-acquired dependency in functional activities of daily living persists well beyond three months post-surgery. Twenty percent of patients have been reported to need help putting on pants, 50% need assistance to walk, and 90% need assistance to climb stairs one year post-fracture.13 A Cochrane Collaboration review on exercise interventions post-hip fracture (13 clinical trials including 1,065 participants) concluded that there is insufficient evidence to determine if the evaluated interventions substantially reduce residual disability and enhance community ambulation.14

Community ambulation is associated with five physical precursors: endurance, dynamic balance, quadriceps strength, walking speed, and lower extremity function.15,16 In two studies, one of older adults with hip fracture and one among frail community-residing older adults without

Protocol Version 10.0 (DSMB approval 23JUN2016) 4 The CAP Study hip fracture, the most successful interventions in terms of improving strength, balance, and gait speed were completed over a 6-month time period in an exercise center.17,18 Interventions included various combinations of strengthening exercise, balance training, functional training, and active range of motion (AROM) and flexibility. Results show positive outcomes in terms of function, gait speed, balance, strength, and endurance.17-24 However, there is no evidence regarding the effectiveness of delivering a similarly intense multi-component intervention in the home. Mangione reported dramatic improvements in physical precursors to community ambulation using an intervention that provided function, strength, balance, and endurance training in a single patient three months post-hip fracture and as soon as usual home care was complete.25

The Community Ambulation Project (CAP) was designed to determine whether a specific multi-component 16-week intervention based on specificity of training and overload (the PUSH intervention) delivered in the home and initiated up to 26 weeks following hospital admission for hip fracture, will be successful in producing community ambulation (ability to walk a distance of

300 m or more in 6 minutes) at 16 weeks after randomization. We will compare PUSH to a non- specific home-based multi-component intervention of sensory level electrical stimulation, flexibility and active range of motion (AROM) exercises (the PULSE intervention). We are testing the hypothesis that a greater proportion of PUSH participants than PULSE participants will achieve community ambulation 16 weeks after randomization. In addition, the cost- effectiveness of the interventions as well as effects on other outcomes will be examined.

METHODS

Study Organization

This multi-center randomized controlled trial (RCT) has a Clinical Coordinating Center, a

Data Coordinating Center, an Economic Evaluation Core, and three clinical sites. In addition to these units, the project has a Steering Committee and three Scientific Oversight Committees

Protocol Version 10.0 (DSMB approval 23JUN2016) 5 The CAP Study that provide expertise needed to conduct this investigation. Members of the Data and Safety

Monitoring Board (DSMB) were appointed by and report to the National Institute on Aging (NIA) which is the funding agency. The DSMB monitors participant safety; evaluates the progress of the study; reviews procedures for maintaining confidentiality, quality of data collection, statistical analysis plans, and results of interim analyses in accordance with pre-specified stopping guidelines.

Study Design

This is a randomized, parallel, two-group multi-center RCT of 210 older adults who have experienced a hip fracture. Half of the participants receives a specific multi-component intervention (PUSH) and the other half receives a non-specific multi-component intervention

(PULSE). Patients age 60 years and older who have had surgical repair for hip fracture are identified at three clinical sites (Arcadia University-AU, University of Maryland, Baltimore- UMB, and University of Connecticut Health Center- UCHC). Following consent to participate after hospital discharge, eligible participants undergo a comprehensive baseline assessment and are randomized to one of the two treatment groups. Participants are assessed again 16 weeks post- randomization. Figure 1 shows the sequence of participant contacts. For participants randomized prior to the current version of the protocol, follow-up assessment visits also occurred at 40 weeks from the date of randomization and telephone interviews were conducted every four weeks during the full 40-week study period.

Within a week of randomization, a physical therapist (PT) initiates the intervention in the participant’s place of residence and they receive counseling with a registered dietician.

Participants receive calcium, vitamin D, and multivitamin supplements during the 16-week intervention period. A $20 honorarium is given to the participant at the completion of the each of the three assessment visits.

Figure 1. Participant timeline for screening, randomization, and follow-up

Recruitment Procedures, Setting, and Participants

Most potential participants are identified by CAP staff at study hospitals, rehabilitation centers, and other agencies that care for older adults. A smaller number of participants are recruited through the use of flyers, advertisements, social media, or referral from an orthopedic surgeon or physical therapist.

Eligibility Screening

Participants are evaluated for eligibility in three phases completed no later than 26 weeks after admission to the hospital for hip fracture. Patients identified in acute care hospitals, and who are provisionally eligible based on medical chart review (phase 1), are approached in the hospital (or soon after discharge) and told about the study. Written permission from the patient (or the legally authorized representative) is obtained to allow future contact to collect additional eligibility information. Patients who are identified through means other than screening in study hospitals are asked to provide written authorization for release of medical records to review for eligibility (phase 1) criteria.

The participant is assessed for medical, safety, and feasibility criteria during phase 2.

Written informed consent is obtained during the phase 2 assessment. Blood is collected no more than 4 weeks prior to randomization and is tested for hemoglobin, serum creatinine and serum albumin.

Prior to the phase 3 visit, a medical provider familiar with the patient since the hip fracture (e.g., primary care physician, orthopedic surgeon, etc.) reviews the potential participant’s medical history and confirms safety for participation. Phase 3 eligibility screening is performed during the baseline visit which is conducted at the clinical site. Screening for final eligibility includes a review for all disqualifying medical conditions, assessment of cognitive

Protocol Version 10.0 (DSMB approval 23JUN2016) 7 The CAP Study status with the Modified Mini-Mental State examination (3MS), and evaluation of ability to walk

300 m or more in six minutes (using the SMWT).

Inclusion and Exclusion Criteria

Participants are included if they had a minimal trauma, closed, non-pathologic hip fracture with surgical repair, are age 60 years or older at time of randomization, were living in the community at time of fracture and randomization, were ambulating without human assistance two months prior to fracture, and are unable to walk 300 m or more in six minutes at baseline.

Patients who are not able to provide their own informed consent to participate in the study are not eligible. Patients are excluded if they have end stage renal disease, recent myocardial infarction, chest pain or shortness of breath, uncompensated congestive heart failure, severely diminished lower extremity sensation or ulceration, symptoms of angina, or uncontrolled hypertension or if they are not fully weight bearing on either leg. Patients are ineligible if they develop chest pain, intolerable dyspnea, or severe pain during the SMWT, if it takes >40 seconds to walk the first four meters of the SMWT, or if serum albumin is less than

2.5 g/dL, hemoglobin is less than 9 g/dL, or calculated creatinine clearance is less than 15 ml/min. Furthermore, patients are not eligible if they are receiving physical therapy in a hospital or inpatient rehabilitation facility at the time of randomization, are cognitively impaired (Modified

Mini-Mental State Exam (3MS) score <73),25,26 are not English speaking, do not live within a reasonable distance of the clinical site, are participating in another clinical trial, or plan to move out of the area during the 16-week intervention period. Patients are not eligible if medical clearance is denied by the patient’s medical provider or if the clinical site clinician thinks the patient is not a good candidate for the trial due to having medical conditions that would prohibit safe participation in the trail (e.g., advanced progressive neuromuscular diseases such as severe Parkinson’s disease or multiple sclerosis, uncontrolled psychoses, or unstable

Protocol Version 10.0 (DSMB approval 23JUN2016) 8 The CAP Study cardiovascular disease) or if the patient is not likely to survive study period. Finally, patients are not eligible if they are community ambulators (walk ≥300 m in six minutes) at baseline or if randomization has not occurred by the end of the 26th week post-admission.

Informed Consent

Written informed consent and appropriate HIPAA authorizations and/or waivers are obtained in compliance with procedures reviewed and approved by each clinical site’s

Institutional Review Board (IRB) prior to data collection. Potential participants who choose to enroll are assessed for their ability to provide informed consent using a local IRB-approved method. Individuals who do not understand the study purpose, the voluntary nature of the study, what will be asked of them during the study, and study risks are not eligible for participation.

Demographic and Surgical Characteristics

For descriptive purposes, information on the following participant characteristics is collected either during screening or at the baseline assessment: age, gender, race, living situation, marital status, educational level, fracture type, surgery type, height and weight. Other characteristics include the Functional Comorbidity Index which is a clinically based measure of

18 comorbidities developed by Groll et al.,27 the Life Orientation Test-Revised (LOT-R) to assess self-reported optimism and pessimism,28 and the Brief Resilience Scale to assess self- reported ability to bounce back after a stressful event.29

Randomization Procedures

The randomization schedule for each clinical site ensures that treatments are randomly assigned within blocks of 2, 4, 6, or 8, with equal numbers of participants assigned to each treatment within each block. Block sizes are randomly selected with the probability of each block size specified by the data coordinating center.

Randomization is performed by unblinded study staff, the clinical site principal investigator or the clinical site coordinator, using a secure Web-based randomization system.

The study staff member accesses the system by logging in to a secure Web site and receives the treatment assignment after responding to prompts confirming that the participant meets all inclusion and exclusion criteria and has given informed consent for enrollment.

Interventions

Participants are assigned to one of two treatment groups: PUSH or PULSE. Both groups receive 32 visits of approximately 60 minutes duration from a study PT in the participant’s place of residence. Participants receive two visits per week, on non-consecutive days, for 16 weeks.

Missed visits in a given week can be replaced by makeup visits on non-consecutive days in subsequent weeks. At the start of the study, PTs were randomly assigned by the DCC to deliver one of the two interventions within strata defined by geographic area. When there is staff turnover, newly hired PTs are assigned to the intervention group and geographic area of the PT whom they are replacing.

The PTs are trained in intervention-specific face-to-face sessions with the Intervention

Monitor (IM). The training includes one half day of didactic instruction, skills practice with the equipment, and live videotaped performance of the procedures with older adult volunteers. The intervention manual of procedures is reviewed throughout the session. PTs are certified after they demonstrate competence (90% passing score) on a written exam and video observation.

PUSH Intervention

The multi-component PUSH intervention focuses on the precursors for community ambulation. Endurance is addressed with continuous upright exercise for 20 minutes. Function is addressed by improving fast walking (needed to navigate streets outdoors, standing from a chair, and stair negotiation). Muscle performance is addressed through exercises to enhance lower extremity strength in functionally relevant muscles moving through locomotion-appropriate movements and ranges. Finally, balance is addressed by performing unilateral activities and activities with decreased base of support. The exercise components are woven together into

Protocol Version 10.0 (DSMB approval 23JUN2016) 10 The CAP Study one program to reduce participant burden.19,30

The strength components are addressed using a portable progressive resistive exercise device (Shuttle® MiniPress, Contemporary Design Company, P.O. Box 5089, Glacier, WA

98244). The device has six latex bands each with a starting load equal to approximately seven pounds. At full excursion one band can provide approximately 15 pounds of force.31

Muscle performance focuses on bilateral hip extensors, hip abductors, knee extensors, and plantar flexors because of their role in function, specifically gait and transfer activities.32-34

The intensity of strength training is the amount of resistance the participant can push against so that s/he can complete a maximum of eight repetitions (eight repetitions maximum (RM) or 8-

RM). For the first visit, participants are tested to find the load associated with an intensity of 8-

RM. During the second session, participants perform two sets of eight repetitions at the 8-RM intensity. From the third session through the remainder of the program, participants perform three sets of eight repetitions at an intensity of 8-RM. The PT provides strong verbal encouragement during each repetition for the participant to move “as fast and as hard as possible” during the concentric phase and to move slowly and in a controlled way in the eccentric phase. Intensity is re-assessed every two weeks.

The participants are supine for two exercises: the combination of hip and knee extension exercise (leg press) and the hip abduction exercise (Figures 2a and 2b). Balance and strength are addressed with two additional exercises performed while standing. The first exercise that combines balance training and strength training is standing hip extension (Figure 2c). Upright balance is challenged as the participant moves the carriage of the exercise device with one leg as the other leg maintains stability. The second exercise that combines balance and strength training is standing plantar flexion. Initially, participants are asked to decrease their standing base of support by rising onto the balls of their feet. This exercise also strengthens the plantar flexors. Balance and strength are progressively challenged by advancing the activity to unilateral heel raises. For hip extension and plantar flexion, the person may hold lightly onto an

Protocol Version 10.0 (DSMB approval 23JUN2016) 11 The CAP Study assistive device for balance or support. The PT encourages the participant to use less external support for balance during each session.

The endurance intervention begins with two to three minutes of continuous upper and lower extremity AROM with the participant sitting. The participant then walks on level surfaces or, if steps are available, up and down one or more steps to keep the heart rate (HR) within the training zone for 20 minutes. The HR training zone is calculated based on the heart rate reserve

(HRR) method [(HR max-HR rest) multiplied by 50% and then added to HR rest, where HR max is maximal HR and HR rest is resting heart rate]. HR max is calculated as 220 minus the participant’s age. If the person is taking medication that controls heart rate (e.g., beta-blockers),

Borg’s Rating of Perceived Exertion (RPE) scale is used to assess the individual’s rating of exercise intensity.35 The training intensity using the RPE scale is “moderate” work and consistent with a range of 3-5 on the 0-10 scale. The PT can also engage the participant in additional exercises such as upper and lower extremity AROM exercises to keep the HR elevated. The target HR is 30% of HRR in the first week of the program, 40% of HRR in the second week, and 50% of HRR in weeks 3 through 16.

PULSE Intervention

The PULSE intervention group receives flexibility exercises, AROM for the upper and lower extremities, breathing exercises, and transcutaneous electrical nerve stimulation (TENS).

During the AROM exercises, participants work to increase flexibility and range of motion in order to increase the motion the participant produces. The exercises involve the neck, shoulders, arms, hands, trunk, hips, knees, and ankles. During all of the AROM exercises, the participant focuses on deep breathing techniques. Progression is gradual beginning with three repetitions of each exercise, with emphasis on proper technique and holding at end range and slowly progressing by 2-3 repetitions each week to 20 repetitions by week 8. This portion of the session lasts approximately 20-30 minutes.

The second part of the session uses sensory level electrical stimulation to the lower extremity muscle groups. Conventional TENS uses low-level electrical current to stimulate superficial cutaneous nerve fibers through the skin. The amount of current for sensory level stimulation is the level which the participant detects as a "tingling" sensation and is not high enough to produce a visible muscle contraction (below motor threshold).36 The goal of the TENS portion of the intervention is two-fold. First, TENS is being used to promote somatosensory input. The goal of this input is to induce changes in the excitability of the motor neurons and to the somatosensory motor cortex to assist in activation of the motor neurons37 used for mobility.

Second, TENS is also intended to decrease pain.38,39 The Select 3 TENS Kit (TENS Pros, P.O.

Box 410504, Saint Louis, MO 63141) is being used. We are using a frequency of 80 pulses per second, pulse duration 50 µsec, and sufficient amplitude to produce a comfortable paresthesia

(tingling or 'pins and needles' sensation). Flexible carbonized, disposable electrodes coated with a self-adhesive conductive polymer are applied to the skin. The electrodes are placed bilaterally near the motor points on muscle bellies (the gluteal complex, the quadriceps, and the gastrocnemius) for seven minutes per muscle group.

Treatment Fidelity

Treatment fidelity data provide information on the adherence of the PTs to delivery of the interventions, and of the participants’ receipt of the prescribed activities. PTs’ knowledge of procedures is tested by written examination and by psychomotor skills observation via video recording and scoring using a treatment fidelity checklist. To ensure ongoing treatment fidelity, there is direct observation of skills by the clinical site coordinators, monthly telephone calls with the IM and all active PTs, periodic review of PT log books, and regular meetings of the

Intervention Committee.

Fidelity to delivery and receipt of the interventions is monitored by calculating the proportion of PT visits during which there is adherence to various components of the intervention. In the PUSH group, adherence to the strength component and to the endurance component is monitored as well as the proportion of visits during which supine exercises were performed on the floor. In the PULSE group, adherence to the AROM exercises and to the

TENS protocol is monitored.

Another measure of treatment fidelity is the participants’ physiologic response (HR) during the intervention sessions in both groups measured using Polar heart rate monitors (Polar

Electro Inc.,1111 Marcus Avenue, Suite M15, Lake Success, NY 11042-1034). The Polar monitor allows the PT to indicate the start and end of each portion of the multicomponent intervention. The Polar software generates curves of HR by time that are reviewed both qualitatively and quantitatively by the IM. The duration and pattern of activity (height of curve, duration of the rise in HR, number of times HR rises) are evaluated according to pre-specified criteria established for each intervention.

Nutritional Intervention

All participants receive 2000 IU of vitamin D3, 600 mg of calcium, and a multivitamin daily during the 16-week intervention period as well as nutritional counseling to ensure weight

Protocol Version 10.0 (DSMB approval 23JUN2016) 14 The CAP Study stability and adequate nutrient intake of 1 g protein/kg body weight inclusive of a healthy diet.

Participants are screened for nutritional risk at the time of randomization using the Mini

Nutritional Assessment-Short Form (MNA®-SF).40 Those who score ≤7 (malnourished) at baseline and participants with serum albumin 2.5-3.5 g/dl (regardless of MNA®-SF score) receive a visit from a registered dietician (RD) within seven days of randomization. The RD evaluates and counsels participants on dietary modifications and follows up by telephone one week after the visit. Participants who score 8-11 (at risk of malnutrition) at baseline and have serum albumin level >3.5 g/dl receive a telephone dietary consultation from RD. Those with a score ≥12 on the MNA®-SF and who have serum albumin level >3.5 g/dl at baseline receive brief telephone contact from the RD to discuss the importance of calorie and protein intake.

Weight is monitored during home PT visits every four weeks and those who lose 2% or more of body weight between two measurements receive another RD telephone consultation.

Blinding of Study Staff

Given the nature of the interventions being tested in this study, it is not possible to blind the participants or the PTs providing the interventions to the participant’s treatment group. Also, it is impossible to ensure that the PTs have not read about the study on ClinicalTrials.gov or in scientific publications. However, to minimize the chances of “contamination” of the interventions, each PT provides only one of the interventions and receives a “blinded” copy of the study protocol that does not include details on the other intervention, study outcomes, or analyses.

PTs do not have contact with participants in the other group or with PTs providing the intervention to the other group. The PTs receive education on the rationale for blinding as part of their training. They are told that specific and non-specific interventions are being tested for their ability to increase mobility following hip fracture, but that specific hypotheses and outcomes will not be shared with them to maintain the integrity of the study. Clinical site PIs and coordinators who are responsible for assigning work or assessing treatment fidelity are not

Protocol Version 10.0 (DSMB approval 23JUN2016) 15 The CAP Study blinded to study outcomes. Study staff involved in randomization, who are not blind to treatment assignment, do not perform follow-up data collection. Staff performing follow-up assessments or telephone interviews are blind to treatment assignment.

Primary Outcome Measure: Community Ambulation

Although community ambulation is complex, covering a minimal distance within a specified period of time is a critical feature.41,42 The SMWT is a standardized test that can be used to examine both the distance and time components of community ambulation,43 and the

SMWT is a useful integrated measure of mobility function taking into account limitations imposed by major body systems.44 The SMWT distance is highly correlated with workloads, heart rate, oxygen saturation, and dyspnea responses when compared to bicycle ergometry and treadmill exercise tests in older persons.43,45,46

The study definition of community ambulation is walking a distance of 300 m or more in six minutes.43 This distance corresponds to a gait speed of 0.83 m/s.43,44 A threshold of 300 m was selected because: 1) at six months post-fracture or post-stroke, average group performance is < 300 m;19,20,47-49 2) achieving < 300 m on the SMWT is a predictor of mortality;50 and 3) the average distance from a parking space in a supermarket parking lot to task completion in the supermarket in both urban and rural communities is 301 m.41 A gait speed of

0.8 m/s(equivalent to approximately 300 m in six minutes) has been suggested as being indicative of a community walker.51,52 One year post-hip fracture, usual gait speeds range from

0.44 to 0.97 m/s,19,21,47,53-58 the average of which is below the community ambulation threshold.

Some authors have reported that a speed threshold of 0.8 m/s is a useful and discriminative primary endpoint in clinical trials of exercise rehabilitation.58

Adjudication of the Primary Outcome

All cases for which, despite intensive efforts, administration of the SMWT is not possible after baseline testing are submitted for adjudication. The purpose of the adjudication is to

determine, as much as possible, outcomes for participants who would otherwise be missing

outcomes by identifying treatment failures (inability to walk at least 300 m in six minutes) when

the SMWT could not be administered. There are three adjudicators, all experts in mobility

assessment and the clinical management of older persons.

Based on the adjudication, participants who meet one or more of the following criteria at

the follow-up assessment are classified as treatment failures in the primary analysis for the

given follow-up time:

1) Participant died

2) Participant was too sick, based on adjudicators’ review of source documentation, to walk 300 m

or more in six minutes

3) Gait speed based on the Short Physical Performance Battery (SPPB) four-meter walk was less

than 0.6 m/s (a very slow speed incompatible with ability to walk 300 m in six minutes)

4) Participant or proxy reports participant limitation in (a) walking several blocks or (b) walking

across a small room without help from another person

Adjudicated participants who do not meet any of the above criteria are classified as

“indeterminate” with respect to the primary outcome. For the primary analysis, when calculating

the proportion of community ambulators, the denominator includes those who are community

ambulators (based on the results of the SMWT), those who are not community ambulators

(based on the results of the SMWT), and those who are treatment failures based on

adjudication.

Secondary Outcomes

Five variables are hypothesized to be precursors to community ambulation. To gain a better understanding of the interventions’ mechanisms, the two groups will be compared at 16 weeks post-randomization on each of the following outcomes and at 40 weeks in the subset of participants randomized prior to current version of the protocol.

Endurance: Endurance is operationalized as a continuous measure of total distance walked in six minutes using the SMWT.46

Dynamic balance: We are using an enhanced balance measure that includes the balance subscale of the SPPB and two additional single leg stands (eyes open and eyes closed), as used in the National Health and Aging Trends Study.59 For the SPPB, participants are asked to maintain balance in three positions with progressive narrowing of the base of support (side-by-side, semi-tandem, tandem position) for a maximum of 10 seconds in each position. The single leg stands are performed on the side of the fracture for up to 30 seconds.

The number of seconds is summed across the five items to obtain the measure of balance.

These tests are hierarchical such that when a participant is unable to perform an item, the harder ones are not administered and the participant receives a score of 0 for the items not done.

Quadriceps strength: Isometric force for bilateral knee extensors is measured with a portable, hand-held dynamometer (Microfet2 Manual Muscle Tester; Hoggan Scientific, LLC,

3653 W 1987 S #7, Salt Lake City, UT 84104). Participants are seated on a special strength testing chair, with hip flexion 90° and knee flexed to 70°, stabilization straps on the pelvis and thigh, and resistance applied just proximal to the ankle on the anterior surface of the leg.60

Participants are asked to push as hard and as fast as possible for five seconds. Three maximal effort trials, with a one-minute rest between trials, are performed. The peak force is recorded for each of the three trials and the highest value is used.

Lower extremity function: A modified version of the Physical Performance Test (mPPT)61 is being used. The mPPT includes nine timed standardized tasks (e.g., picking up a penny from the floor, standing up five times from a 16-inch chair). The tasks are performed twice and the times from the two trials are averaged. The score for each item ranges from 0 to 4, with 36 representing a perfect score.

Fast walking speed: Within the mPPT, participants are asked to walk a distance of 50 feet walking quickly but safely. The time required to walk 50 feet is the measure of fast walking speed.

Cost-effectiveness of the interventions

The economic value of the interventions will be determined by assessing the impact on

Quality-Adjusted Life Years (QALYs), cost, and cost per QALY gained over the 40 weeks following randomization. Resource utilization is tracked by telephone interview every four weeks and unit costs will be assigned to each resource for which data are obtained. Resource use data are collected as health care use, both formal and informal care, according to intervals between telephone interviews. In addition to caregiver costs, the questionnaire documents other resource use including admission to the hospital and other care facilities, health care visits, and diagnostic tests. A health care diary is provided to each participant to assist in accurately recording health care encounters and use of health services.

Delayed and sustained effects

For the subset of participants who were followed for 40 weeks, a secondary objective is to determine whether the proportion of community ambulators differs between the PUSH and

PULSE interventions at 40 weeks post-randomization, and whether the difference in proportions at 40 weeks changed from the difference in proportions at 16 weeks.

Tertiary outcomes

Activities of daily living are measured using a modified version of the Pepper

Assessment Tool for Disability (PAT-D).62,63 Two items (walking a quarter mile and walking across a small room) have been added to address perceived gaps in the original PAT-D scale and two items (walking several blocks and lifting heavy objects) were deleted to avoid duplication with other items in the scale. Other tertiary outcomes include: balance confidence assessed with the Activities-specific Balance Confidence scale;64 quality of life using the interviewer-administered version of the 36-Item Short Form Health Survey (SF-36);65 physical activity using the Yale Physical Activity Survey, an interviewer-administered questionnaire that includes five categories of common activities related to work, exercise, and recreation performed during a typical week in the past month;66 the SPPB to evaluate lower extremity performance based on timed short distance walk, repeated chair stands, and a set of balance tests;67-69 depressive symptoms using the 20-item Center for Epidemiologic Studies Depression

(CES-D) scale;70 increase of ≥ 50 m in distance walked in six minutes; cognitive status at follow- up using the 3MS26; and nutritional risk status using a version of the MNA®-SF that was adapted for use as an interviewer-administered instrument.40,71

Adherence

Receipt of PT intervention visits is operationalized as the total number of PT visits conducted divided by the total number of expected visits per participant.

Adherence to the interventions is monitored by calculating the proportion of PT visits during which there is adherence to specific components of the intervention as described above as part of treatment fidelity assessment.

Adherence with study vitamins and supplements is monitored by pill counts every four weeks during the intervention period and by self-report during the four-week telephone calls.

Adverse Events

Reportable adverse events (RAEs), which include serious adverse events, unexpected adverse events, or injury that occurs under supervision by study staff, are obtained throughout the study period starting when the participant signs the consent form. Information about RAEs is solicited every four weeks during the telephone interviews using standardized questions. During these interviews, participants (or their proxies) are asked about life-threatening or significant medical events and the outcomes of these events. Information about RAEs is also solicited by study staff prior to each PT visit and clinical site follow-up assessment. RAEs may be spontaneously reported to any study staff member at any time.

Expected adverse events are assessed during the telephone interview every four weeks.

Participants (or their proxies) are asked a series of standardized questions related to pain (feet, hip, back, knees); breathing problems or chest pain; skin irritations, rash, or skin tears; numbness or tingling; and falls (with and without injury).

Statistical Analyses & Sample Size Calculations

Analyses for all aims will be performed according to the intention-to-treat paradigm. With the exception of the statistical test of the primary hypothesis, all statistical tests will be two-sided and will not be adjusted for multiple comparisons.

Primary Aim

The primary aim is to determine if the PUSH intervention is more successful in producing community ambulation at 16 weeks post-randomization compared to the PULSE intervention.

The one-sided null hypothesis that the PUSH intervention does not result in a higher proportion of community ambulators 16 weeks post-randomization will be tested at four interim analyses

(after 20%, 40%, 60% and 80% of the data are available) and one final analysis based on a Z- statistic for the difference in proportions. A one-sided test was chosen because PUSH is the intervention of scientific interest. PULSE is a control intervention and, therefore, we are only

interested in knowing whether PUSH is superior to PULSE. It is not necessary to distinguish

between equivalence and inferiority because, in either case, PULSE would not be promoted as

a clinically meaningful intervention.

At each interim analysis, there are three possible decisions that can be made: 1) stop

the trial based on evidence for efficacy of PUSH, 2) stop the trial based on absence of evidence

for efficacy of PUSH, or 3) continue the trial. At the final analysis, we will determine whether

there is sufficient evidence for the efficacy of PUSH. The critical values for each interim analysis

and the final analysis were chosen based on a Hwang-Shih-Decani alpha-spending function to

preserve an overall type-1 error rate of 0.025 as implemented in the R package “gsDesign”.72 If

the exploratory analyses reveal data sparseness (expected frequency less than 5 for at least

one combination of treatment group and community ambulation status), Fisher’s exact test will

be performed. The difference in proportions with a 95% confidence interval will also be reported.

In addition, a series of analyses will be performed to examine the differential impact of

the PUSH intervention relative to the PULSE intervention in subgroups defined by participant

characteristics. To do this, a variable-by-intervention interaction term will be tested for each of

the following variables:

1. Sex

2. Age at baseline (≥85 years versus 60-84 years)

3. Baseline depressive symptoms (CES-D score ≥ 16 versus CES-D score < 16)

4. Baseline balance confidence (with median Activities-Specific Balance Confidence scale score

as the cutpoint to define the subgroups)

5. Baseline nutritional status (MNA®-SF score <8 versus MNA®-SF score ≥8)

6. Baseline cognitive status (3MS score <91 versus 3MS score ≥91)

If the variable-by-interaction term for any of these subgroup variables is significant at the 0.05

level, results will be presented separately in strata of the subgroup variable.

Secondary and Tertiary Outcomes

For the subset of participants who were followed for 40 weeks, longitudinal regression models fit by generalized estimating equations (GEEs)73 will be used to examine whether the proportion of community ambulators differs between the PUSH and PULSE interventions at 40 weeks post-randomization and whether the difference in proportions at 40 weeks changed from the difference in proportions at 16 weeks. Additionally longitudinal regression models fit by GEE will be used to examine the difference between the treatments with respect to the secondary and tertiary outcomes described earlier.

Cost-Effectiveness of Interventions

To assess the cost-effectiveness of study interventions, we will conduct analyses of within-trial comparisons for the economic endpoints (resource utilization/costs and a specially derived reduced version of the SF-36 (SF-6D)74/QALYs) over the 40-week post-randomization period. We will also analyze the economic value of the trial interventions versus usual care using a model-based analysis that combines trial data with data from the Baltimore Hip Studies with similar hip fracture patient groups. Statistical analyses of SF-6D will produce an estimate of the incremental QALYs associated with the PUSH intervention at follow-up. The estimated difference in QALYs attributable to PUSH will be estimated by taking a time-weighted average of the time-specific intervention effects. Statistical analyses of cost data, adjusted to a constant dollar year (e.g., 2012 US dollars), will produce an estimate of the incremental costs associated with PUSH vs. PULSE.

The incremental cost-effectiveness ratio (ICER), which is defined as the net change in cost divided by the net change in effectiveness (QALYs) when interventions are ranked in order of increasing cost, is the focus of the economic analysis. When estimated as added cost per

QALY gained, the ICER allows the value of interventions in hip fracture to be compared with interventions in other diseases.

Sample Size Adequacy: Primary Aim

Using a group sequential design with four interim analyses, and assuming an overall loss to follow-up of 6.8%, a total sample of 210 enrolled patients (105 patients per group) is sufficient to achieve 80% power for the detection of a 20 percentage point difference between the groups, with 16% achieving the 300 m threshold of the SMWT in the PULSE group vs 36% in the PUSH group, using a one-sided 0.025 significance level. The power specified is for the detection of a time-specific difference at the 16-week follow-up. This calculation accounts for the sample size inflation of 2.4% entailed in the group sequential design.75,76 In addition, this accounts for an anticipated 12.2% noncompliance to the intervention under the assumption that participants assigned to the PUSH group who do not adhere to the PUSH intervention have a proportion achieving community ambulation at 16 weeks equal to that anticipated among participants assigned to the PULSE group.77,78

Sample Size Adequacy: Secondary Aims

Secondary and Tertiary Outcomes. Using a two-sided test with alpha=0.05, a sample of

210 patients with 6.8% loss to follow-up and 12.2% treatment noncompliance will provide 80% power to detect an effect size of 0.46 of a standard deviation at 16 weeks, which is considered a moderate effect size, under the assumption that non-compliers assigned to the PUSH group and participants assigned to the PULSE group have equal mean and standard deviation.

Economic Evaluation. We estimate that there will be approximately 150 participants for whom the full 40-week follow-up will be available. Power considerations were based on the minimum detectable incremental net monetary benefit defined as *QALY-Cost where  is the willingness to pay threshold of $100,000 per QALY gained. We estimated that with 75 per group that we could detect an incremental net monetary benefit of $7,234 or greater with 80% power and using a two-sided 0.05-level test.

Delayed and Sustained Effects. As with the economic evaluation, follow-up ends at 16 weeks post-randomization for those randomized under current version of the protocol. We estimate that there will be approximately 150 participants (not the full 210) for whom the 40- week measure of community ambulation will be available. In addition, based on our experience to date, we estimate that the community ambulation outcome will be indeterminate for 10% of the participants at the 40-week assessment. If we assume a sample size of 150 then there is

64% power for detecting a difference of 20 percentage points at alpha = 0.05 using a two-tailed test.

Discussion

Older adult who experience a hip fracture continue to suffer from significant residual disability long after conventional rehabilitation ends. Endurance, dynamic balance, quadriceps strength, and function are compromised and contribute to failure of hip fracture patients’ ability to walk independently in the community. Although the incidence of hip fracture appears to be declining, the absolute number of hip fractures is expected to increase over the next 25 years because of the growing number of people surviving to older ages. Hence, the number of hip fracture patients with residual deficits and the inability to ambulate and carry out tasks of daily living in the community will also increase. One of the challenges is to find ways to deliver rehabilitative services to frail older adults over a longer period of time; a home-based program is one such strategy that can enhance adherence and promote independence after completion of the program. However, practice guideline changes will not occur without compelling evidence that interventions performed in the home are safe, result in a reduction in the level of post- fracture disability, increase community participation, and are cost-effective. The need for a rigorous trial to examine home-based multi-component interventions is critical.

This will be the first study to test whether the multi-component intervention based on specificity of training and overload (PUSH) has a greater probability of leading to community

Protocol Version 10.0 (DSMB approval 23JUN2016) 25 The CAP Study ambulation in older adults following hip fracture compared to a multi-component intervention that is non-specific (PULSE). The PUSH intervention is unique in that it addresses precursors of ambulation in the community. As a comparison, the PULSE intervention is non-specific and targets the same muscle groups as PUSH with the addition of a modality for pain relief. Several key aspects of the interventions are comparable (delivered in the home by a licensed PT with the same frequency and duration of visits) allowing for valid comparisons between the groups which is not typical of many prior rehabilitation trials after hip fracture.18,23,79 Another strength is the attention to treatment fidelity to ensure adherence to the delivery and receipt of the interventions. This is a rigorously designed and executed study that conforms to all

Consolidated Standards of Reporting Trials (CONSORT)80 criteria for the conduct and reporting of RCTs, including blinding to extent possible in a behavioral intervention of this type (i.e., PTs have no knowledge about the other treatment; staff measuring outcomes are blinded to treatment group; participants are told not to discuss specifics of treatment with evaluators), standardization of measures, and control of data quality.

If there is evidence that the PUSH intervention is effective, this RCT will also provide critical information on the cost-effectiveness of the intervention. Although a number of studies have addressed the cost-effectiveness of hip fracture prevention through pharmacological and hip-protector interventions,51,52,56-58,81-84 the potential economic value of exercise interventions to reduce post-fracture disability has not been adequately evaluated.

The mobility disability that remains post-hip fracture is remarkable and our primary outcome (community ambulation) is a measure that is relevant to individuals and their families.

Community ambulation confers the ability to independently leave the home and perform certain activities of daily living Compared to older adults who are active in the community, persons with a mobility disability do not travel alone, take fewer trips and perform fewer activities per trip, walk shorter distances, cross the street less often, carry fewer objects, have fewer postural

Protocol Version 10.0 (DSMB approval 23JUN2016) 26 The CAP Study transitions (turning the head, extending their reach, or changing direction),85 and need more assistance from others. We believe that walking 300 m or more in six minutes in a clinical testing situation captures this ability to ambulate independently in the community.

The protocol has undergone four major revisions. In the first major revision (April 2014), we modified the inclusion and exclusion criteria making them less restrictive and allowing a larger pool of hip fracture patients to participate. We also widened the randomization window from 12 weeks post-admission to 20 weeks post-admission to allow participation of people who receive standard physical therapy for more than 12 weeks. In the second major revision

(November 2014), we decreased the total number of intervention visits from 40 to 32 over 16 weeks. There is evidence suggesting that interventions provided three times per week are not superior to interventions that are given two times per week in terms of improving strength or function.86,87 It was anticipated that reducing the frequency of intervention visits would reduce participant burden and might result in increased recruitment. The interval between hospital admission and randomization was further extended to 26 weeks. In the third major revision

(December 2015), the target sample size was reduced to 210 participants (105 per group) from the originally planned 300 (150 per group). Calculations based on dropout and non-compliance rates experienced in our study showed that we still had 80% power to assess our primary hypothesis. Most recently, the fourth revision (June 2016) eliminated follow-up after 16-weeks.

Therefore, a subset of participants (approximately 150) randomized prior to this protocol version will have a 40-week assessment and monthly telephone calls in months 17-40 post randomization.

The first participant was randomized on September 16, 2013. There are currently 150 participants who have been randomized. The most common reasons for ineligibility are not living in the community at time of fracture, not ambulating without human assistance two months prior to fracture, cognitive impairment, and not having surgical repair of the hip fracture. Overall, approximately 35% of screened patients 60 and older are eligible for the study. We anticipate

Protocol Version 10.0 (DSMB approval 23JUN2016) 27 The CAP Study completing randomization in July 2017 and ending assessments for collecting the primary outcome on the last participant in November 2017.

Conflict of interest declaration: The authors have no conflicts of interest to declare.

Acknowledgements: This work was supported by grants from the National Institute of Child Health and Human Development and the National Institute on Aging at the National Institutes of Health (R21 HD043269, R37 AG09901 MERIT Award, R01 AG029315, T32 AG00262, and P30 AG028747). This material is the result of work supported with resources and the use of facilities at the VA Maryland Health Care System and Baltimore VA Medical Center.

References

1. Shumway-Cook A, Ciol MA, Gruber W, Robinson C. Incidence of and risk factors for falls following hip fracture in community-dwelling older adults. PhysTher 2005;85:648-55. 2. Shumway-Cook A, Ciol MA, Yorkston KM, Hoffman JM, Chan L. Mobility limitations in the Medicare population: prevalence and sociodemographic and clinical correlates. J Am Geriatr Soc 2005;53:1217-21. 3. Stevens JA, Rudd RA. The impact of decreasing U.S. hip fracture rates on future hip fracture estimates. Osteoporos Int 2013;24:2725-8. 4. Kozak LJ, Hall MJ, Owings MF. National Hospital Discharge Survey: 2000 annual summary with detailed diagnosis and procedure data. Vital Health Stat 13 2002:1-194. 5. Schneider EL, Guralnik JM. The aging of America -- Impact on health care costs. JAMA 1990;263:2335-40. 6. Tosteson A, Solomon D, King A, Dawson-Hughes B, Burge R, Wong J. Projections of osteoporosis fractures and costs by skeletal site in the USA [Abstract]. J Bone Min Res 2005;20:S21. 7. Praemer A, Furner S, Rice D. Musculoskeletal conditions in the United States. Surgeon General's Report. Rosemont, IL: American Academy of Orthopaedic Surgeons; 1992. 8. Brainsky A, Glick H, Lydick E, et al. The economic cost of hip fractures in community- dwelling older adults: A prospective study. J Am Geriatr Soc 1997;45:281-7. 9. Ray NF, Chan JK, Thamer M, Melton LJ. Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J Bone Miner Res 1997;12:24-35. 10. Office of Technology Assessment. Hip fracture outcomes in people aged 50 and over: Mortality, service use, expenditures, and long-term functional impairment: Congress of the United States; 1993. 11. Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Miner Res 2007;22:465-75. 12. Koval KJ, Cooley MR. Clinical pathway after hip fracture. Disabil Rehabil 2005;27:1053-60. 13. Magaziner J, Hawkes W, Hebel JR, et al. Recovery from hip fracture in eight areas of function. J Gerontol A Biol Sci Med Sci 2000;55:M498-507. 14. Handoll H, Sherrington C. Mobilisation strategies after hip fracture surgery in adults. Cochrane Database Syst Rev 2007;23:CD001704. 15. Lee KB, Lim SH, Ko EH, Kim YS, Lee KS, Hwang BY. Factors related to community ambulation in patients with chronic stroke. Top Stroke Rehabil 2015;22:63-71. 16. Ferreira ML, Sherrington C, Smith K, et al. Physical activity improves strength, balance and endurance in adults aged 40-65 years: a systematic review. J Physiother 2012;58:145-56. 17. Binder EF, Schechtman KB, Ehsani AA, et al. Effects of exercise training on frailty in community-dwelling older adults: results of a randomized, controlled trial. J Am Geriatr Soc 2002;50:1921-8. 18. Binder EF, Brown M, Sinacore DR, Steger-May K, Yarasheski KE, Schechtman KB. Effects of extended outpatient rehabilitation after hip fracture: a randomized controlled trial. JAMA 2004;292:837-46. 19. Mangione KC, RL; Tomlinson, SS; Palombaro, KM. Can elderly patients who have had a hip fracture perform moderate- to high-intensity exercise at home? PhysTher 2005;85:727-39.

20. Mangione KKC, R.L.; Lopopolo, R.B.; Tomlinson, J.D.; Brenneman, S.K. Gait speed predictors in patients after hip fracture. Physiotherapy Canada 2008;60:10-8. 21. Sherrington CL, S R. Home exercise to improve strength and walking velocity after hip fracture: a randomized controlled trial. Arch Phys Med Rehabil 1997;78:208-12. 22. Sherrington CL, S.R.; Herbert,R.D. A randomized controlled trial of weight-bearing versus non-weight-bearing exercise for improving physical ability after usual care for hip fracture. Arch Phys Med Rehabil 2004;85:710-6. 23. Tinetti MEB, D. I.; Gottschalk, M.; Williams, C. S.; Pollack, D.; Garrett, P.; Gill, T. M.; Marottoli, R. A.; Acampora, D. Home-based multicomponent rehabilitation program for older persons after hip fracture: a randomized trial. Arch Phys Med Rehabil 1999;80:916-22. 24. Henderson SAF, O E; Murphy, N; Boreham, C; Mollan, R A; Gilmore, D H; Beringer, T R. Benefits of an exercise class for elderly women following hip surgery. Ulster Med J 1992;61:144-50. 25. Teng EL, Chui HC. The Modified Mini-Mental State (3MS) examination. J Clin Psychiatr 1987;48:314-8. 26. Tombaugh TNM, I.;Kristjansson, B.;Hubley, A.M. Mini-Mental State Examination (MMSE) and the Modified MMSE (3MS): A Psychometric Comparison and Normative Data. Psychological Assessment 1996;8:48-59. 27. Groll DL, To T, Bombardier C, Wright JG. The development of a comorbidity index with physical function as the outcome. J Clin Epidemiol 2005;58:595-602. 28. Scheier MF, Carver CS, Bridges MW. Distinguishing optimism from neuroticism (and trait anxiety, self-mastery, and self-esteem): a reevaluation of the Life Orientation Test. J Pers Soc Psychol 1994;67:1063-78. 29. Smith BW, Dalen J, Wiggins K, Tooley E, Christopher P, Bernard J. The brief resilience scale: assessing the ability to bounce back. Int J Behav Med 2008;15:194-200. 30. Mangione KKP, K. M. Exercise prescription for a patient 3 months after hip fracture. PhysTher 2005;85:676-87. 31. Bolinger RWG, G. Owners Manual: Shuttle Miniclinic. Glacier, Washington: Contemporary Design Company; 1997. 32. Winter DAE, J. J.; Ishac, M. G.; Craik, R. L.; Oatis, C. A. A Review of Kinetic Parameters in Human Walking. Gait Analysis: Theory and Application. St. Louis: Mosby; 1995:252-70. 33. Knutson LMS, G. L.; Craik, R. L.; Oatis, C. A. EMG: Use and Interpretation in Gait. Gait Analysis: Theory and Application. St. Louis: Mosby; 1995:307-25. 34. Chandler JMD, P. W.; Kochersberger, G.; Studenski, S. Is lower extremity strength gain associated with improvement in physical performance and disability in frail, community- dwelling elders? Arch Phys Med Rehabil 1998;79:24-30. 35. Borg GA. Perceived exertion: a note on "history" and methods. Med Sci Sports 1973;5:90-3. 36. Snyder-Mackler L. Electrical stimulation for pain modulation. In: Robinson AS-M, L, ed. Clinical Electrophysiology Electrotherapy and Electrophysiologic Testing. Philadelphia: Lippincott Williams & Wilkins; 1995:279-310. 37. Beekhuizen KS, Field-Fote EC. Sensory stimulation augments the effects of massed practice training in persons with tetraplegia. Arch Phys Med Rehabil 2008;89:602-8. 38. Cheing GL, Hui-Chan CW. Would the addition of TENS to exercise training produce better physical performance outcomes in people with knee osteoarthritis than either intervention alone? Clin Rehabil 2004;18:487-97.

39. Abou-Setta AM, Beaupre LA, Rashiq S, et al. Comparative effectiveness of pain management interventions for hip fracture: a systematic review. Ann Intern Med 2011;155:234-45. 40. Guigoz Y, Vellas B, Garry PJ. Assessing the nutritional status of the elderly: The Mini Nutritional Assessment as part of the geriatric evaluation. Nutr Rev 1996;54:S59-S65. 41. Robinett CSV, M. A. Functional ambulation velocity and distance requirements in rural and urban communities. A clinical report. PhysTher 1988;68:1371-3. 42. Guralnik JMF, L. P.; Salive, M. E. Disability as a public health outcome in the aging population. Annu Rev Public Health 1996;17:25-46. 43. Guyatt GH, Sullivan MJ, Thompson PJ, et al. The 6-minute walk: a new measure of exercise capacity in patients with chronic heart failure. Can Med Assoc J 1985;132:919-23. 44. Harada NDC, V.; Stewart, A. L. Mobility-related function in older adults: assessment with a 6-minute walk test. Arch Phys Med Rehabil 1999;80:837-41. 45. Peeters PM, T. The 6-minute walk as an appropriate exercise test in elderly patients with chronic heart failure. J Gerontol A Biol Sci Med Sci 1996;51:M147-51. 46. American Thoracic Society. ATS statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med 2002;166:111-7. 47. Mangione KK, Craik RL, Palombaro KM, Tomlinson SS, Hofmann MT. Home-based leg- strengthening exercise improves function 1 year after hip fracture: a randomized controlled study. J Am Geriatr Soc 2010;58:1911-7. 48. Latham NKM, V.; Nguyen, A.M.; Jette, A.M.; Olarsch, S.; Papanicolaou, D.; Chandler, J. Performance-based or self-report measures of physical function: which should be used in clinical trials of hip fracture patients? Arch Phys Med Rehabil 2008;89:2146-55. 49. Duncan PWL, S. M.; Bode, R. K.; Perera, S.; DeRosa, J. Stroke Impact Scale-16: A brief assessment of physical function. Neurology 2003;60:291-6. 50. Rostagno CG, G.; Romano, M.; Chiostri, G.; Gensini, G. F. Prognostic value of 6-minute walk corridor testing in women with mild to moderate heart failure. Ital Heart J 2002;3:109- 13. 51. Newman ABH, C.L.; Kritchevsky, S.B.; Nevitt, M.C.; Simonsick, E.M.; Health ABC Collaborative Research Group,. Walking performance and cardiovascular response: associations with age and morbidity - the health, aging and body composition study. J Gerontol A Biol Sci Med Sci 2003;58A:715-20. 52. Langlois JAK, P.M.; Guralnik, J.M.; Foley, D.J.; Marottoli, R.A.; Wallace, R.B. Characteristic of older pedestrians who have difficulty crossing the street. Am J Public Health 1997;87:393- 7. 53. Sherrington CL, S. R. Increased prevalence of fall risk factors in older people following hip fracture. Gerontology 1998;44:340-4. 54. Hauer KS, N.; Schuler, M.; Bartsch, P.; Oster, P. Intensive physical training in geriatric patients after severe falls and hip surgery. Age Ageing 2002;31:49-57. 55. Sherrington CL, S. R.; Herbert, R. D. A randomised trial of weight-bearing versus non- weight-bearing exercise for improving physical ability in inpatients after hip fracture. Aust J Physiother 2003;49:15-22. 56. Ingemarsson AHF, K.; Mellstrom, D.; Moller, M. Walking ability and activity level after hip fracture in the elderly - a follow-up. J Rehabil Med 2003;35:76-83.

57. Mendelsohn MEL, D. S.; Overend, T. J.; Petrella, R. J. Specificity of functional mobility measures in older adults after hip fracture: a pilot study. Am J Phys Med Rehabil 2003;82:766-74. 58. Lord SEM, K.; McNaughton, H. K.; Rochester, L.; Weatherall, M. Community ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil 2004;85:234-9. 59. Freedman VA, Kasper JD, Cornman JC, et al. Validation of new measures of disability and functioning in the National Health and Aging Trends Study. J Gerontol A Biol Sci Med Sci 2011;66:1013-21. 60. Bohannon R. Test-retest reliability of hand-held dynamometry during a single session of strength assessment. PhysTher 1986;66:206-9. 61. Reuben DBS, A. L. An objective measure of physical function of elderly outpatients. The Physical Performance Test. J Am Geriatr Soc 1990;38:1105-12. 62. Rejeski WJ, Ettinger WH, Jr., Schumaker S, James P, Burns R, Elam JT. Assessing performance-related disability in patients with knee osteoarthritis. Osteoarthritis Cartilage 1995;3:157-67. 63. Rejeski WJ, Ip EH, Marsh AP, Miller ME, Farmer DF. Measuring disability in older adults: the International Classification System of Functioning, Disability and Health (ICF) framework. Geriatr Gerontol Int 2008;8:48-54. 64. Powell LEM, A M. The Activities-specific Balance Confidence (ABC) Scale. J of Gerontol: Med Sci 1995;50A:M28-34. 65. Ware JES, C.D. The MOS 36-Item Short-Form Health Survey (SF-36): Conceptual Framework. Med Care 1992;30:473-83. 66. Dipietro L, Caspersen CJ, Ostfeld AM, Nadel ER. A survey for assessing physical activity among older adults. Med Sci Sports Exerc 1993;25:628-42. 67. Guralnik JMS, E M; Ferrucci, L; Glynn, R J; Berkman, L F; Blazer, D G; Scherr, P A; Wallace, R B. A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol: Med Sci 1994;49:M85-M94. 68. Guralnik JMS, T. E.; Tinetti, M. E.; Nevitt, M. C.; Berkman, L. F. Validation and use of performance measures of functioning in a non- disabled older population: MacArthur studies of successful aging. Aging (Milano) 1994;6:410-9. 69. Guralnik JM, Ferrucci L, Pieper CF, et al. Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the short physical performance battery. J Gerontol A Biol Sci Med Sci 2000;55:M221-31. 70. Radloff LS. The CES-D Scale: A self-report depression scale for research in the general population. Appl Psychol Meas 1977;1:385-401. 71. Kaiser MJ, Bauer JM, Ramsch C, et al. Validation of the Mini Nutritional Assessment short- form (MNA-SF): a practical tool for identification of nutritional status. J Nutr Health Aging 2009;13:782-8. 72. Anderson K. (2016) Package 'gsDesign': Group Sequential Design (version 3.0-1) [Software]. Available from https://cran.r-project.org/web/packages/gsDesign/gsDesign.pdf. 73. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. 74. Brazier JR, J.; Deverill, M. The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271-92.

75. O'Brien PCF, T.R. A multiple testing procedure for clinical trials. Biometrics 1979;35:549-56. 76. Hwang IK, Shih WJ, De Cani JS. Group sequential designs using a family of type I error probability spending functions. Stat Med 1990;9:1439-45. 77. Fay MP, Halloran ME, Follmann DA. Accounting for variability in sample size estimation with applications to nonadherence and estimation of variance and effect size. Biometrics 2007;63:465-74. 78. Shardell MD, El-Kamary SS. Calculating sample size for studies with expected all-or-none nonadherence and selection bias. Biometrics 2009;65:635-9. 79. Latham NK, Harris BA, Bean JF, et al. Effect of a Home-Based Exercise Program on Functional Recovery Following Rehabilitation After Hip Fracture. JAMA: Journal of the American Medical Association 2014;311:700. 80. Schulz KF, Altman DG, Moher D, Group C. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. BMC Med 2010;8:18. 81. Fried LPT, C.M.; Walston, J.; Newman, A.B.; Hirsch, C.; Gottdiener, J.; Seeman, T.; Tracy, R.; Kop, W.J.; Burke, G.; McBurnie, M.A.; Cardiovascular Health Study Collaborative Research Group,. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001;56:M146-M56. 82. Molen HH. Problems on the evaluation of gait (thesis). Amsterdam: Amsterdam Free University; 1973. 83. Drillis R. The influence of aging on kinematics of gait. Washington, DC: NAS-NRC; 1961. Report No.: Pub. No. 919. 84. van Schoor NMdB, M. C.; van der Roer, N.; Lommerse, E.; van Tulder, M. W.; Bouter, L. M.; Lips, P. Cost-effectiveness of hip protectors in frail institutionalized elderly. Osteoporos Int 2004;15:964-9. 85. Patla AS-C, A. Dimensions of mobility: defining the complexity and difficulty associated with community mobility. J Aging Phys Act 1998;7:7-19. 86. Latham N, Anderson C, Bennett D, Stretton C. Progressive resistance strength training for physical disability in older people. Cochrane Database Syst Rev 2003:CD002759. 87. Liu CJ, Latham NK. Progressive resistance strength training for improving physical function in older adults. Cochrane Database Syst Rev 2009:CD002759.

Protocol Version 10.0 (DSMB approval 23JUN2016) 33