Archives of Physical Medicine and Rehabilitation

journal homepage: www.archives-pmr.org

Archives of Physical Medicine and Rehabilitation 2018;-:-------

REVIEW ARTICLE (META-ANALYSIS)

Effectiveness of Supervised Home-Based Exercise

Therapy Compared to a Control Intervention on
Functions, Activities, and Participation in Older
Patients After Hip Fracture: A Systematic Review and
Meta-analysis
Isolde A.R. Kuijlaars, MSc,a Lieke Sweerts, MSc,b
Maria W.G. Nijhuis-van der Sanden, PhD,b,c Romke van Balen, MD, PhD,d
J. Bart Staal, PhD,b,e Nico L.U. van Meeteren, PhD,f,g Thomas J. Hoogeboom, PhDb
From the aCenter for Physical Therapy Ramaekers, Weert; bRadboud Institute for Health Sciences, IQ Healthcare, Radboud University Medical
Center, Nijmegen; cDepartment of Rehabilitation, Radboud University Medical Center, Nijmegen; dDepartment of Public Health and Primary
Care, Leiden University Medical Centre, Leiden; eResearch Group Musculoskeletal Rehabilitation, HAN University of Applied Sciences, Nijmegen;
f
Department of Epidemiology, CAPHRI, Maastricht University, Maastricht; and gHealthwHolland, Top Sector Life Sciences and Health, the
Hague, the Netherlands.

Abstract
Objective: The aim of this review was to investigate whether supervised home-based exercise therapy after hospitalization is more effective on
improving functions, activities, and participation in older patients after hip fracture than a control intervention (including usual care). Further-
more, we aimed to account the body of evidence for therapeutic validity.
Data Sources: Systematic searches of Medline, Embase, and CINAHL databases up to June 30, 2016.
Study Selection: Randomized controlled trials studying supervised home-based exercise therapy after hospitalization in older patients (
65y)
after hip fracture.
Data Extraction: Two reviewers assessed methodological quality (Physiotherapy Evidence Database) and therapeutic validity (Consensus on
Therapeutic Exercise Training). Data were primary analyzed using a best evidence synthesis on methodological quality and meta-analyses.
Data Synthesis: A total of 9 articles were included (6 trials; 602 patients). Methodological quality was high in 4 of 6 studies. One study had high
therapeutic validity. We found limited evidence in favor of home-based exercise therapy for short- (4mo) and long-term (>4mo) performance-
based activities of daily living (ADL) and effects at long-term for gait (fast) and endurance. Evidence of no effectiveness was found for short- and
long-term effects on gait and self-reported (instrumental) ADL and short-term effects on balance, endurance, and mobility. Conflicting evidence
was found for strength, long-term balance, short-term gait (comfortable), long-term self-reported ADL, and long-term mobility.
Conclusions: Research findings show no evidence in favor of home-based exercise therapy after hip fracture for most outcomes of functions,
activities, and participation. However, trials in this field have low therapeutic validity (absence of rationale for content and intensity and reporting
of adherence), which results in interventions that do not fit patients’ limitations and goals.
Archives of Physical Medicine and Rehabilitation 2018;-:-------
ª 2018 by the American Congress of Rehabilitation Medicine

Therapeutic exercise training after hip fracture is recommended in

clinical guidelines, as part of a multidisciplinary program.1-3
However, many patients with hip fractures never regain their
Disclosures: none. prefracture levels of functioning in various domains of functioning

0003-9993/18/$36 - see front matter ª 2018 by the American Congress of Rehabilitation Medicine
https://doi.org/10.1016/j.apmr.2018.05.006
2 I.A.R. Kuijlaars et al
according to the International Classification of Functioning,
Disability and Health (ICF).4-6 Over the last years, home reha-
bilitation has been encouraged in the Netherlands.7,8 Besides a
cost argument, it is conceivable that patients recover better at
home on functional and quality-of-life outcomes, because they can
incorporate what they learn directly into their own context.9
Furthermore, in an observational study, approximately half of
the patients recovering from a hip fracture chose to undertake
home rehabilitation. These patients had a higher rate of recovery
compared to patients who chose institution-based rehabilitation.
However, in this observational nonrandomized design, prefracture
functional status in activities of daily living (ADL) was better and
less people lived alone before fracture in this home-based reha-
bilitation group.10 Interestingly, home-based exercise in-
terventions yield smaller effects than institution-based programs
as suggested in 2 systematic reviews with searches up to 2012.
One review focused on supervised and nonsupervised in-
terventions,11 and 1 review reported only physical function and
quality of life as outcome measures.12 Several new randomized
controlled trials (RCTs) have been published since 2012 which
might shed another light on the topic.
Moreover, previous reviews on the effects of home-based
rehabilitation programs for patients after bone fractures have not
yet accounted for the interventions’ therapeutic validity.13,14
Herbert and Bo (2005) recommended to assess the quality of
complex exercise interventions,13 known as therapeutic validity.
Therapeutic validity is defined as “the potential effectiveness of a
specific intervention given the potential target group of
patients”15(p.2) that focuses on patient selection, therapist and
setting selection, rationale and content of the intervention, and
adherence. Recently, a number of systematic reviews have been
published using this new approach that aids reviewers to account
for the validity of studied interventions to pool the best in-
terventions and their effects rather than just provide evidence.15-17
It appears that interventions rarely meet the requirements for
therapeutic validity.15-17 By including studies with limited validity
or nonadequate patient selection, the overall (beneficial) effects
may dilute. Therefore, we aim to introduce this methodology
within the literature of exercise therapy in patients after hip
fracture to explore therapeutic validity in this field. The aim of the
present study was to systematically review RCTs on the short-
(4mo) and long-term (>4mo) effectiveness after hospitalization
on body functions, activities, and participation (conform the ICF)
of supervised home-based exercise therapy in older patients
(
65y) after hip fracture compared with a control intervention
(including usual care). We hypothesized that home-based in-
terventions were superior to control interventions. Furthermore,
we aimed to account the body of evidence for therapeutic validity.
We hypothesized that studies with higher methodological quality
and therapeutic validity would demonstrate greater effectiveness.
List of abbreviations:
ADL activities of daily living
CI confidence interval
CONTENT Consensus on Therapeutic Exercise Training
IADL instrumental activities of daily living
ICF International Classification of Functioning,
Disability and Health
MD mean difference
RCT randomized controlled trial
SMD standardized mean difference
Methods
This systematic review was performed according to Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
Statement.18 The research proposal is registered in PROSPERO
(CRD42015029535).
Literature search
We searched the electronic databases Medline, Embase, and
CINAHL for eligible articles up to June 30, 2016. We used the
high-sensitivity strategy for the search of RCTs.19 The search did
not have language restrictions. The search strategy can be found in
supplemental table S1 (available online only at http://www.
archives-pmr.org/). In addition, reference tracking was per-
formed in all eligible studies.
Two reviewers (I.K. and L.S.) conducted inclusion of articles.
Initially, articles were screened for eligibility on title and abstract.
If title and abstract suggested that an article was potentially
eligible for inclusion, a full paper copy of the report was obtained.
See fig 1 for the flow chart of the included studies.
Articles were eligible for inclusion if (1) the design was an
RCT; (2) older patients were included (
65y) after hip fracture
(box 1); (3) the authors compared structured (ie, preplanned
treatment), supervised (by an exercise trainer or physical thera-
pist), home-based exercise therapy (as primary intervention) with
an alternative intervention (including usual care and traditional
rehabilitation) after hospitalization; and (4) the authors reported
performance-based or self-reported outcome measures of body
functions (ICF domain neuromuscular and movement-related
functions) or activities and participation (ICF domains mobility,
self-care, domestic life).6 Articles were excluded if they studied a
multidisciplinary program.
Data extraction
For all included studies, we assessed whether a design paper was
available to obtain more information regarding the intervention,
which we then used as an additional source in our data extraction.
If no design paper was available, we contacted the corresponding
authors of the included studies for additional information. One
reviewer (I.K.) extracted data from each study using a self-
administered standardized extraction form. Another reviewer
(L.S.) checked a random portion of 33% of the data extraction. If
any error occurred, a second (and so on) 17% of the data was
checked. Data extracted were author names, publication year,
participants (sample size, age, sex, residence), recruitment rate,
time from fracture to start of the study, prestudy and within study
description and exercise parameters, control group description,
number of participants lost to follow-up, outcome measures, and
study results. Only point estimates were reported in this review for
between-group comparisons. Continuous outcomes were
expressed as mean differences (MD) with corresponding 95%
confidence interval (95% CI). Dichotomous outcomes were
expressed as relative risk with corresponding 95% CI. If this in-
formation was not presented in the articles, we calculated these
measures if enough data were available.19 Authors of the RCTs
were contacted if these data were not reported in the published
article. The outcomes of the different studies were divided in
domains within the components of functions or activities and
participation to interpret the results. Results were classified into
short-term (4mo) and long-term results (>4mo) after baseline.20
www.archives-pmr.org
Home-based exercise after hip fracture
Fig 1 Flow chart of the included studies.
For strength outcomes of lower extremity muscle strength by
dynamometers, strap with a spring gauge or a leg extension power
rig was included. Step-up tests were not presented in this review.
Cadence measures were not presented for gait.
Methodological quality and therapeutic validity
Two reviewers (I.K. and L.S.) independently assessed methodo-
logical quality (table 1) and therapeutic validity (table 2) of the
included articles. Methodological quality is the likelihood of the
www.archives-pmr.org
3
trial design to generate unbiased results that are sufficiently pre-
cise and allow replication in clinical practice.21 The methodo-
logical quality was scored using the Physiotherapy Evidence
Database (PEDro) scale.22 The PEDro scale is a reliable measure
for rating quality of RCTs.21 Scores range from 0 to 10, scores
6
indicate moderate-to-high methodological quality.22 The thera-
peutic validity was scored using the Consensus on Therapeutic
Exercise Training (CONTENT) scale, developed by Hoogeboom
et al (2012)15 for assessing therapeutic validity of therapeutic
exercise programmes. Scores range from 0 to 9, scores
6
4 I.A.R. Kuijlaars et al

In case of clinical heterogeneity of the studies, decided by

Box 1 Definition for older patients after hip fracture according to
20
Crotty et al.
We used the definition of Crotty et al20 to ensure only studies
on older patients after hip fracture were included, namely:
“The main study population was older people with any type
of fracture of the proximal femur. Most participants were
aged 65 years or over and had undergone surgery for their
hip fracture. Trials that included younger participants were
included if the mean age minus one standard deviation was
greater than 65 years. Participants younger than 65 years
were included as long as the number of these was relatively
small and there was adequate randomization of younger
patients to intervention and control groups. Studies which
focused on younger people with hip fracture were excluded,
as were trials involving people with multiple trauma.”20(p.4)
reflecting high therapeutic validity.15 Disagreements were
resolved in a consensus meeting. The strength of interrater
agreement was measured by Cohen k coefficient (95% CI), with
kZ0.41-0.60 indicating moderate agreement, kZ0.61-0.80 rep-
resenting good agreement, and k
0.81 representing very good
agreement.23,24
Data analysis
Data were pooled when studies were clinically homogenous
concerning participants, intervention, and outcome measures.
Continuous measures of functions, activities, and participation in
the home-based exercise and control groups were transformed to
standardized mean difference (SMD) (Hedges g) to cope with
variety of outcome measures.19 For studies with more than 1
home-based exercise group, we chose to include the intervention
in the meta-analysis that was most alike the other interventions to
avoid a unit-of-analysis error.19 Calculations were performed
using a random-effects model. An SMD of 0.2, 0.5, and 0.8 rep-
resents respectively a small, moderate, and large effect.25 An
alpha value of <0.05 was considered statistically significant. Stata
13a was used to pool the data.
the authors based on the participants residence, content, and
time frame of the interventions (I.K., L.S., T.H.), we performed
a best evidence synthesis based on methodological quality.
Studies were rated as high-quality study when the PEDro score
was
6. Studies were rated as low-quality study when the
PEDro score was <6.22
For the best evidence, synthesis results of the studies were
rated on effectiveness (yes or no). A P value of <.05 was
considered statistically significant. Because some studies reported
more results on 1 outcome with both statistically and non-
statistically significant results, we decided if more outcome
measures in 1 study were not statistically significant than statis-
tically significant, results were rated as no effectiveness. If an
equal number of items were statistically significant as not statis-
tically significant, the long-term effects were reported in the best
evidence synthesis, as we deemed these most meaningful to
the patients.
The levels of evidence were based on Heymans et al (2004).26
Evidence was rated as no evidence (no RCTs), limited evidence
(only 1 RCT [either high or low quality]), conflicting evidence
(inconsistent findings in multiple RCTs), moderate evidence
(generally consistent findings in 1 high-quality RCT plus 1 or
more low-quality RCTs, or by generally consistent findings in
multiple low-quality RCTs), and strong evidence (generally
consistent findings in multiple high-quality RCTs).26 Consistent
findings was defined as similar results in
75% of the studies.27
A sensitivity analysis was performed to assess the impact of
therapeutic validity, by including only studies with high method-
ological quality (
6) and high therapeutic validity (
6) in the
best-evidence synthesis.
Results
Literature search
Our search resulted in 913 potentially relevant articles. After
selection 9 articles met our inclusion criteria,28-36 reporting
data of 6 trials.28-33 According to the Cochrane Handbook,

Table 1 Methodological quality of the included studies

PEDro Score
Source, y 1 2 3 4 5 6 7 8 9 10 11 Total
Latham et al, 201428 þ þ  þ   þ  þ þ þ 6
Salpakoski et al, 201429 þ þ þ þ   þ þ þ þ þ 8
Orwig et al, 201130 þ þ  þ   þ  þ þ þ 6
Mangione et al, 201031 þ þ þ þ   þ  þ þ þ 7
Mangione et al, 200532 þ þ  þ   þ   þ þ 5
Sherrington et al, 199733 þ þ  þ    þ  þ þ 5
NOTE. One point was awarded when a criterion is clearly satisfied (þ: 1 point was awarded, : criterion was not satisfied). Item 1 was not used to
calculate the PEDro score. Scores range from 0 to 10. The criteria are defined as follows: (1) eligibility criteria were specified; (2) participants were
randomly allocated to groups (in a crossover study, participants were randomly allocated an order in which treatments were received); (3) allocation
was concealed; (4) the groups were similar at baseline regarding the most important prognostic indicators; (5) there was blinding of all participants;
(6) there was blinding of all therapists who administered the therapy; (7) there was blinding of all assessors who measured at least 1 key outcome; (8)
measures of at least 1 key outcome were obtained from >85% of the participants initially allocated to groups; (9) all participants for whom outcome
measures were available received the treatment or control condition as allocated or, where this was not the case, data for at least 1 key outcome was
analyzed by intention to treat; (10) the results of between-group statistical comparisons are reported for at least 1 key outcome; and (11) the study
provides both point measures and measures of variability for at least 1 key outcome.

www.archives-pmr.org
Home-based exercise after hip fracture 5

Table 2 Therapeutic validity of the included studies

CONTENT Scale
Source, y A1 A2 B3 C4 C5 D6 D7 D8 E9 Total
Latham et al, 2014 28
þ þ      þ  3
Salpakoski et al, 201429 þ   þ    þ  3
Orwig et al, 201130 þ  þ þ  þ  þ  5
Mangione et al, 201031 þ þ    þ þ   4
Mangione et al, 200532 þ  þ þ þ þ þ  þ 7
Sherrington et al, 199733    þ      1
NOTE. One point was awarded when a criterion is clearly satisfied (þ: 1 point was awarded, : criterion was not satisfied). Scores range from 0 to 9. The
criteria are defined as follows: (1) was the patient selection described?; (2) was the patient selection adequate?; (3) were eligibility criteria for
therapist and setting determined and adequate?; (4) was the therapeutic exercise based on a priori aims and intentions?; (5) was the rationale for the
content and intensity of the therapeutic exercise described and plausible?; (6) was the intensity of the therapeutic exercise described?; (7) was the
therapeutic exercise monitored and adjusted when considered necessary?; (8) was the therapeutic exercise personalized and contextualized to the
individual participants?; (9) was adherence to the therapeutic exercise determined and acceptable?

information from multiple reports of 1 study was collated.19 Two
publications reported the same data of a 6-month home exercise
program for hip fracture recovery,28,34 from here cited as Latham
et al.28 Three publications reported about the same study of the
promotion mobility rehabilitation program,29,35,36 from here
cited as Salpakoski et al (2014).29 A flow diagram of search and
selection is shown in fig 1. Design papers were available for the
study of Latham et al37 and Orwig et al.38 None of the contacted
authors provided additional information regarding their
interventions.
Study characteristics
A second reviewer extracted 67% of the data. Agreement between
the reviewers indicated no further extraction from a second
reviewer was necessary.
Included studies had a total of 602 patients (range mean age
77-83y; 81% women). Table 3 shows the characteristics of
included studies. Three studies included only community-dwelling
participants29,31,32; the other studies also included participants
living in supervised caregiver settings.28,30,33 Five studies
excluded people with cognitive impairment (Mini Mental Status
Examination score <18 or <20).28-32 The interventions started at
different times. One study started as soon as possible after
discharge to home,29 3 studies started after finishing physical
therapy,30-32 1 trial started after finishing usual care,28 and 1 study
initiated therapy within 9 months after fracture.33 Mean time from
surgery to the start of the trials ranges from 9.2 weeks29 up to 9.5
months.28 Intervention duration and follow-up time points varied
from 1 month33 till 1 year.29,30 For a visible depiction, see fig 2. In
5 studies, physical therapists provided the exercise pro-
gram.28,29,31-33 In 1 study, exercise trainers were trained to provide
the intervention.30
In 2 studies, control interventions consisted of usual care.29,30
In one of these studies, all participants received standard care after
discharge home; 70% of the participants received written home
exercise program without follow-up.29 In the other study, the
control group received physician-prescribed postfracture standard
care, which included relatively short hospital stays and approxi-
mately 2-4 weeks of rehabilitation.30 In 3 other studies, the control
group received another prescribed treatment, that is, nutrition
education for cardiovascular health,28 conventional transcutaneous
electrical nerve stimulation,31 and biweekly mailings of the Na-
tional Institutes of Health Age Pages on a variety of nonexercise
topics.32 In 1 study, no description of the intervention in the
control group was given.33
The percentage of participants lost to follow-up after baseline
varied from 4.9%29 to 30.8%.31 The number of participants who
dropped out was equal in the home-based exercise and control
groups. Only in 1 study a higher dropout rate was reported in the
resistance group (6 dropouts of 17 participants) in comparison
with the aerobic (1 dropout of 12 participants) and control group
(1 dropout of 11 participants), because of hospitalization not
related to the training.32
Methodological quality
Table 1 shows the assessment of methodological quality in indi-
vidual studies. The initial agreement of the reviewers on the
assessment of methodological quality was 93.9% (62 of 66 items).
Cohen k (95% CI) was 0.87 (0.72-0.97). Two studies were
assessed as having low methodological quality32,33 and 4 as
having moderate-to-high methodological quality.28-31 Two studies
met the criteria of concealed allocation29,31 and adequate follow-
up.29,33 All studies scored positively for items about eligibility
criteria, random allocation, similar groups at baseline, between-
group comparisons, and providing point measures and measures
of variability.28-33
Therapeutic validity
Table 2 shows the assessment of therapeutic validity in individual
studies. The initial agreement of the reviewers on assessment of
therapeutic validity was 90.7% (49 of 54 items). Cohen k (95%
CI) was 0.81 (0.63-0.96). Five studies showed low therapeutic
validity28-31,33 and 1 study reflected high therapeutic validity.32
One study scored yes on items about the rationale for the con-
tent and intensity of the therapeutic exercise (C5) and adherence
(E9).32 Items about adequate patient selection (A2)28,31 and
monitoring and adjustment of the intervention (D7)31,32 were 2
times scored positively. Items about description of the patient
selection (A1) and a priori aims and intentions for the therapeutic
exercise (C4) were scored respectively 528-32 and 429,30,32,33 times
as positive.

www.archives-pmr.org
 6
Table 3 Characteristics of the included studies
Participants Home-Based Exercise Intervention
Mean Age  SD Total Score
Source, y N (I/C) (y) Women n (%) Description of Intervention Exercise Parameters Control Group PEDro/CONTENT
Latham et al, 232 (120/112) I: 77.210.2 I: 83 (69.2) A home-based program with functionally Prestudy intervention: traditional Nutrition 6/3
201428 C: 78.99.4 C: 77 (68.8) oriented exercises. The program included rehabilitation. All participants received education for
repeating simple functional tasks using Thera- some physical therapy after their cardiovascular
Bands for resistance and standing exercises fracture (92.2% home care therapy health based
using steps. The therapists also used cognitive service, 42.4% outpatient therapy). on the Dietary
and behavioral strategies to positively Intervention duration: 6 mo Guidelines for
enhance the attitudes and beliefs of each Weekly frequency: 3 Americans
study participant related to exercise. Session duration:
 Supervised sessions: 1 h
 Unsupervised sessions: unknown
Weekly duration: unknown
Intensity: using Thera-Bands, steps of
varying height and weighted vests
Supervised sessions: 3 or 4 sessions and
monthly telephone call
Salpakoski 81 (40/41) I: 80.97.7 I: 31 (78) Standard care and the ProMo. A multicomponent Prestudy intervention: Inpatient Standard care 8/3
et al, C: 79.16.4 C: 32 (78) home-based rehabilitation, included rehabilitation
201429 evaluation and modification of environmental Intervention duration: 1 y
hazards, guidance for safe walking, individual Weekly frequency:
nonpharmacologic pain management  Strengthening and stretching exer-
evaluation, individual progressive home cises: 3
exercise program (strengthening exercises for  Balance and functional exercises: 2-3
lower limb muscles, balance training, Session duration: 30 min
stretching and functional exercises), and Weekly duration: 150-180 min
physical activity counseling. Intensity:
 Strengthening exercises: using resis-
tance bands of 3 different strengths
 Balance training: reducing hand sup-
port and the base of support
Supervised sessions: 5-6 and 2 telephone
calls
(continued on next page)
www.archives-pmr.org

I.A.R. Kuijlaars et al
www.archives-pmr.org

Home-based exercise after hip fracture

Table 3 (continued )
Participants Home-Based Exercise Intervention
Mean Age  SD Total Score
Source, y N (I/C) (y) Women n (%) Description of Intervention Exercise Parameters Control Group PEDro/CONTENT
Orwig et al, 180 (91/89) I: 82.57.1 Only women Usual care and the home-based exercise plus Prestudy intervention: physician Usual care 6/5
201130 C: 82.36.9 were included. program consisted of an exercise component prescribed postacute physical therapy
and a self-efficacyebased motivational Intervention duration: 1 y
component. The exercise sessions combined Weekly frequency:
strength training for the upper and lower  Strength training: 2
extremities, aerobic stepping exercises, and  Aerobic exercise: 3
stretching exercises. The plus or motivational Session duration:
component included a self-efficacy-based  Warm-up and cool-down: 20-30 min
motivational component, education and  Strength training: 30 min
encouragement, physiologic feedback, cueing,  Aerobic exercise: 20-30 min
and self-modeling. Weekly duration*: 120-150 min
Intensity:
 Strength training: The duration of each
exercise was increased until 3 sets of
10 repetitions on both sides; intensity
was increased using the resistance of
exercise bands and ankle and wrist
weights.
 Aerobic training: Once participants
could perform 20 min, light ankle
weights were added.
Supervised sessions: 56 and telephone
calls
Mangione 26 (14/12) I: 79.65.9 I: 12 (86) A high-intensity leg strengthening home Prestudy intervention: physical therapy Conventional 7/4
et al, C: 82.06.0 C: 9 (75) exercise program. A portable progressive Intervention duration: 10 wk TENS
201031 resistive exercise machine was used for hip Weekly frequency: 2
and knee muscles. Session duration: 30-40 min
Weekly duration: 60-80 min
Intensity: 8-RM, reevaluated every 2 wk, 3
sets of 8 repetitions
Supervised sessions: 20
(continued on next page)

7
 8
Table 3 (continued )
Participants Home-Based Exercise Intervention
Mean Age  SD Total Score
Source, y N (I/C) (y) Women n (%) Description of Intervention Exercise Parameters Control Group PEDro/CONTENT
Mangione 41 A: 79.85.6y A: 9 (75)y A: Prestudy intervention: physical therapy Biweekly 5/7
et al, A:13 R: 77.97.9y R: 7 (64)y A moderate-intensity aerobic home exercise Intervention duration: 12 wk mailings of
200532 R:17 C: 77.87.3y C: 8 (80)y program with walking on level surfaces and on Weekly frequency: 2 (first 2mo) to 1 (third the National
C:11 stairs for 20 min. If the patient was unable, he month) Institutes of
or she had to perform additional exercises. Session duration: 30-40 min Health Age
R: Weekly duration: 60-80 min (first 2mo) to Pages on a
A high-intensity leg strengthening home 30-40 min (third month) variety of
exercise program. A portable progressive Intensity: 65%-75% of predicted nonexercise
resistive exercise machine was used for hip maximum heart rate topics
and knee muscles. Supervised sessions: 20
Intervention duration: 12 wk
Weekly frequency: 2 (first 2mo) to 1 (third
month)
Session duration: 30-40 min
Weekly duration: 60-80 min (first 2mo) to
30-40 min (third month)
Intensity: 8-RM, reevaluated every 2 wk, 3
sets of 8 repetitions
Supervised sessions: 20
Sherrington 42 (21/21) I: 80.08.1 I: 13 (61.9) Home-based weight-bearing exercise on a Prestudy intervention: unknown No description of 5/1
et al, C: 77.18.2 C: 20 (95.2) stepping block Intervention duration: 1 mo the control
199733 Weekly frequency: 7 intervention
Session duration: unknown
Weekly duration: unknown
Intensity: using a stepping block of 1 or 2
telephone books. Maximum number of
repetitions at assessment with slowly
increase the number of repetitions at
least once a day.
Supervised sessions: 2
Abbreviations: A, moderate-intensity aerobic training group; C, control group; I: intervention group; ProMo, promotion mobility rehabilitation program; R: high-intensity supervised resistance group; RM:
repetition maximum; TENS, transcutaneous electrical nerve stimulation.
* Warm-up and cool-down were not included in the weekly duration.
www.archives-pmr.org

I.A.R. Kuijlaars et al
y
Mangione et al (2005): Age and sex were only reported for patients who completed the study.
Home-based exercise after hip fracture
Fig 2 Time plot of the included RCTs. )Orwig et al (2011) reported no time between fracture and baseline for the control group.
Effectiveness of home-based exercise therapy
Results are shown in supplemental table S2 (available online only
at http://www.archives-pmr.org/) for body functions, activities,
and participation. No point measures were available for the study
of Orwig et al (2011). Table 4 shows the best evidence synthesis
for all outcomes based on methodological quality.
Data of Mangione et al (2010)31 and Mangione et al (2005)32
(resistance group) were pooled for summed lower extremity
strength, gait (comfortable gait speed), and endurance. Other data
were not pooled because of heterogeneity of the studies.
Body functions
All studies reported several outcome measures for body func-
tions.28,29,31-33 Functions were divided into strength, balance, gait,
and endurance. All outcomes were performance based.
Body functions: strength
Two high-quality studies28,31 and 2 low-quality studies32,33 re-
ported outcome measures for muscle strength. Two studies re-
ported baseline muscle strength for the fractured and nonfractured
leg, demonstrating an actual strength deficit in the frac-
tured leg.28,33
Short-term effects on lower extremity strength were reported in
3 studies.31-33 Two studies had an intervention group with an
exercise program specially focused on strength training.31,32 No
www.archives-pmr.org
9
differences were found after the interventions.31,32 After pooling
the lower extremity strength, a nonsignificant SMD with a cor-
responding 95% CI was found of 0.04 (0.66 to 0.58). In the
other study, patients in the home-based exercise group had more
strength than patients in the control group, in both the affected leg
(MD [95% CI]: 3.1 [0.32-5.88]) and the nonaffected leg (MD
[95% CI]: 3.5 [0.01-6.99]).33
Long-term effects on lower extremity strength were reported in
2 studies.28,31 Strength did not differ at 6 months in 1 study.28 At 9
months, strength was better in the fractured leg (MD [95% CI]: 4.3
[0.30-8.30]) and nonfractured leg (MD [95% CI]: 5.0 [0.97-9.03])
in the home-based exercise group.28 The other study found no
difference in lower extremity strength.31
There is conflicting evidence for effectiveness on strength at
short- and long-term follow-up.
Body functions: balance
Two high-quality studies28,29 and 1 low-quality study33 reported
outcome measures of balance. Short-term effects on balance were
reported in 2 studies.29,33 Both studies found no significant results
in balance.29,33
Long-term effects on balance were reported in 2 studies.28,29
One study reported significant improvement at 6 (MD [95% CI]:
3.3 [0.28-6.32]) and 9 months (MD [95% CI]: 5.2 [2.04-8.36])
after baseline28; the other study did not find a significant inter-
vention effect.29
10 I.A.R. Kuijlaars et al

Table 4 Best evidence synthesis based on methodological quality

Was There an Effect?
Outcome Yes No Best Evidence Synthesis
Strength
Short-term Low (33) Low (32) Conflicting evidence
High (31)
Long-term High (28) High (31) Conflicting evidence
Balance
Short-term Low (33) Moderate evidence of no effectiveness
High (29)
Long-term High (28) High (29) Conflicting evidence
Gait (comfortable gait speed)*
Short-term High (31) Low (32,33) Conflicting evidence
Long-term High (31) Limited evidence of no effectiveness
Gait (fast gait speed)*
Short-term High (31) Limited evidence of no effectiveness
Long-term High (31) Limited evidence of effectiveness
Endurance
Short-term Low (32) Moderate evidence of no effectiveness
High (31)
Long-term High (31) Limited evidence of effectiveness
ADLdperformance-based
Short-term High (31) Limited evidence of effectiveness
Long-term High (31) Limited evidence of effectiveness
ADLdself-reported
Short-term Low (32) Strong evidence of no effectiveness
High (29,31)
Long-term High (31) High (28,29) Conflicting evidence
IADLdself-reported
Short-term High (29) Limited evidence of no effectiveness
Long-term High (29) Limited evidence of no effectiveness
Mobilitydperformance-based
Short-term High (29) Limited evidence of no effectiveness
Long-term High (28) High (29) Conflicting evidence
Mobilitydself-reported
Short-term High (29) Limited evidence of no effectiveness
Long-term High (28) High (29) Conflicting evidence
NOTE. Studies were rated as high-quality studies when the PEDro score was
6. Studies were rated as low-quality studies when the PEDro score was <6.
* Results of gait were divided in comfortable and fast gait speed.

There is moderate evidence for no effectiveness on balance at
short-term follow-up and conflicting evidence for effectiveness at
long-term follow-up.
Body functions: gait
One high-quality study31 and 2 low-quality studies32,33 reported
outcome measures of gait. Short-term effects on gait were re-
ported in 3 studies.31-33 The studies reported no significant short-
term effect on gait,31-33 in exception of usual gait speed 10 weeks
after baseline in 1 study (MD [95% CI]: 0.11 [0.05-0.27]).31
Short-term data of comfortable gait speed were pooled for both
studies of Mangione31,32 (SMD [95% CI]: 0.40 [0.18 to 0.98]).
Long-term effects on gait were reported in 1 study, which re-
ported significant results for the home-based exercise group for
fast gait speed (MD [95% CI]: 0.22 [0.03-0.41]).31 Usual gait
speed did not differ between groups.31
There is conflicting evidence for effectiveness on gait
(comfortable gait speed) at short-term follow-up, limited evidence
for no effectiveness on gait (comfortable gait speed) at long-term
follow-up and gait (fast gait speed) at short-term follow-up, and
limited evidence for effectiveness on gait (fast gait speed) at long-
term follow-up.
Body functions: endurance
One high-quality study31 and 1 low-quality study32 reported
outcome measures on endurance. Short-term effects on endurance
were reported in 2 studies.31,32 No significant short-term effects
were found.31,32 Data were pooled for endurance (SMD [95% CI]:
0.40 [0.19 to 0.98]) with no statistical significant differences
between both groups.31,32
The long-term effect on endurance was reported in 1 study,31
which reported a significant improvement for the home-based

www.archives-pmr.org
Home-based exercise after hip fracture
exercise group at 26 weeks after baseline (MD [95% CI]: 80.1
[19.48-140.72]).31
There is moderate evidence for no effectiveness on endurance
at short-term follow-up and limited evidence for effectiveness at
long-term follow-up.
Activities and participation
Four studies reported outcome measures related to activities and
participation.28,29,31,32 Activities and participation were divided
into ADL, instrumental activities of daily living (IADL), and
mobility. Mobility is seen as the ability to move from one place to
another, maintaining posture, walking, and moving objects.6 If
more complex tasks were measured, the outcomes were presented
by ADL. Outcomes were divided into performance-based and self-
reported outcomes.
Activities and participation: activities of daily living
Three high-quality studies28,29,31 and 1 low-quality study32 re-
ported outcome measures for ADL. Short-term performance-based
effects on ADL were reported in 1 study.31 Patients in the home-
based exercise group performed better than the control group at 10
weeks after baseline (MD [95% CI]: 6.1 [2.38-9.82]).31
Short-term self-reported effects on ADL were reported in 3
studies.29,31,32 The studies found no significant results
on ADL.29,31,32
One study reported long-term performance-based effects on
ADL.31 Patients in the home-based exercise group performed
better than the control group 26 weeks after baseline (MD [95%
CI]: 7.3 [3.55-11.05]).31
Long-term self-reported effects on ADL were reported in 3
studies.28,29,31 No significant differences were found in 2
studies.28,29 In the other study, self-reported ADL was signifi-
cantly higher in the home-based exercise group than in the control
group (MD [95% CI]: 18.5 [2.73-34.27]).31
There is limited evidence for effectiveness on performance-
based ADL at short- and long-term follow-up. In addition, there is
strong evidence for no effectiveness on self-reported ADL at
short-term follow-up and conflicting evidence for effectiveness on
self-reported ADL at long-term follow-up.
Activities and participation: instrumental activities of daily
living
One high-quality study reported outcome measures for IADL.29
This study found no significant differences between groups at 3,
6, and 12 months after baseline.29 There is limited evidence for no
effectiveness on self-reported IADL at short- and long-term
follow-up.
Activities and participation: mobility
Two high-quality studies evaluated mobility.28,29 Short-term ef-
fects on mobility were reported in 1 study.29 No differences were
found between the groups on performance-based and self-reported
outcomes.29 Long-term performance-based effects on mobility
were reported in 2 studies.28,29 One study found better perfor-
mances in the home-based exercise group than in the control
group at 6 (MD [95% CI]: 1.0 [0.15-1.85]) and 9 months (MD
[95% CI]: 1.3 [0.44-2.16]) after baseline.28 The other study found
no differences between the groups.29
The same studies reported long-term self-reported outcomes on
mobility.28,29 One study found no difference between groups at 6
months. At 9 months after baseline, patients in the home-based
exercise group scored significantly higher than patients in the
www.archives-pmr.org
11
control group (MD [95% CI]: 2.8 [0.28-5.32]).28 In the other
study, more participants were able to negotiate stairs without
difficulties at 12 months (relative risk [95% CI]: 1.54 [1.06-2.24]).
At 6 months, no significant difference was found.29 Perceived
entrance-related barriers and perceived outdoor barriers were not
different between groups in the same trial.29
There is limited evidence for no effectiveness on performance-
based and self-reported mobility at short-term follow-up. In addi-
tion, there is conflicting evidence for effectiveness on performance-
based and self-reported mobility at long-term follow-up.
Sensitivity analysis
None of the eligible studies had high methodological quality and
therapeutic validity, so the sensitivity analysis yielded no studies.
Discussion
This systematic review did not confirm our hypothesis that home-
based exercise therapy was superior to usual care or a nonexercise
control intervention in older patients after hip fracture. The results
demonstrates that there is only limited evidence in favor of home-
based exercise for short- and long-term performance-based ADL
and long-term gait (fast gait speed) and endurance, based on only 1
study investigating a 10-week high-intensity leg strengthening
program.31 The highest level of evidence was found for no effec-
tiveness of home-based exercise therapy in comparison with an
alternative intervention (including usual care) for short-term self-
reported ADL.29,31,32 There was limited and moderate evidence of
no effectiveness for short- and long-term self-reported IADL,29
short-term balance,29,33 gait (fast gait speed),31 endurance,31,32
self-reported and performance-based mobility,29 and long-term
gait (comfortable gait speed).31 Conflicting evidence was found
for strength, long-term balance, short-term gait (comfortable), long-
term self-reported ADL, and long-term mobility. Pooled data
showed nonsignificant small SMDs in favor of the home-based
exercise group for short-term endurance and comfortable gait
speed.31,32 We were unable to confirm our hypothesis that studies
with high methodological quality and therapeutic validity had
greater effects, because none of the eligible studies met these
criteria. The latter suggests that interventions are ill adapted to the
needs and health problems of the individual patients after hip
fracture. In addition, included RCTs showed wide clinical hetero-
geneity concerning the residence of participants, start of the study,
and content and duration of the intervention. Frequently it was
unclear which interventions were provided between the hip fracture
surgery and start of the studied rehabilitation intervention. More-
over, most interventions in the RCTs (4/6; 67%) included started at a
relatively late stage (4-9mo after hip fracture), thereby limiting the
comparability with daily practice, which typically starts earlier after
surgery. There is risk on selection bias, which is higher in the latest
started studies, because patients with appropriate recovery will not
participate in the studies.
Recently, Diong et al (2015)39 published a meta-analysis about
the effect of structured exercise in all settings on overall mobility
in people after hip fracture. The authors concluded that structured
exercise produced small improvements on overall mobility.
Furthermore, interventions in other settings than hospitals were
more effective.39 Two other reviews did not find enough evidence
to confirm the effectiveness of home-based exercise therapy above
exercise therapy in other settings.11,12 The somewhat positive re-
sults in favor of home-based exercise therapy of our review are in
12
line with earlier reviews of RCTs. Furthermore, previous reviews
on home-based exercise therapy did not account their effects for
therapeutic validity of the included interventions. In other sys-
tematic reviews that assessed therapeutic validity, CONTENT
scores varied from 0-515 to 1-716 in other populations, with 0%-
38% of the studies rated as having high therapeutic validity, in
comparison with CONTENT scores of 1-7 and 17% of the RCTs
rated as high therapeutic validity in this systematic review. To
date, the association between therapeutic validity and treatment
effectiveness has not been confirmed in previous systematic re-
views.15-17 In our systematic review, we were also unable to
confirm an association, which have been due to the limited
number of therapeutically valid studies or perhaps due to an
overall lack of effectiveness of exercise therapy in this group.
In the present systematic review, methodological quality of the
included studies was moderate to high in 4 of 6 studies.28-31
Critical items of methodological quality in the included studies
were concealed allocation and adequate follow-up. No study
blinded the participants and therapists who administered the
therapy, because this is not possible in exercise interventions. The
limited number of studies that met the criterion of concealed
allocation29,31 increased the risk of selection bias and could affect
the generalizability of the results. Furthermore, only 2 studies had
an adequate follow-up of at least 1 key outcome29,33 which could
have biased the results.
The assessment of therapeutic validity is seen as an added
value to earlier published reviews on exercise therapy after hip
fracture.11,12 Only in 1 trial the authors argued the content and
intensity of the intervention which are necessary to achieve ef-
fects.32 Despite within-group effects, no significant between-group
effects on the outcome measures were found in this study because
of the small sample size. In addition, this was the only study with
acceptable adherence to the intervention,32 which included a
description of the number of sessions and achievement of target
intensity. For other studies adherence was not described suffi-
ciently and the real training intensity was unclear which could
have resulted in lower training effect. In only 2 trials, the goals of
the therapeutic exercise matched the participants’ problems28,31
and the treatment was monitored and adjusted for an optimal
exercise intensity.31,32 Furthermore, only 3 studies provided an
individual personalized and contextualized exercise program.28-30
This seems to be quintessential for successful physical therapy
interventions in the frail elderly population,40,41 which is in line
with the value of task-specific training in some other patients (eg,
poststroke).42 In summary, the therapeutic validity of included
studies was low and no RCT could be included in our sensitivity
analysis. This indicated that in most of the studies a home-based
exercise program was provided that may have had limited op-
portunity to a priori be effective (regardless of the premise
whether or not exercise is effective in this population). In addition,
in most studies patients were selected on their injury and not on
the basis of their limitations, even though limitations underlie the
request for care. Structured interventions, like the included RCTs,
lack a patient-centered approach which enables a customized se-
lection of exercises and goal-directed training for the individual
goals of the patient on predefined decision rules.40,41
Study limitations
A limiting factor of this review was that the CONTENT scale only
reviews what is written down in articles. By doing so, possibly
I.A.R. Kuijlaars et al
items might have been scored false negative. However, in 2
studies, design papers were available and only 1 design paper gave
extra information to tackle false negative scores.30 In both studies,
the CONTENT scores were still low (3 and 5 points).28,30
Furthermore, the reliability of this new scale has not been deter-
mined but there was good agreement between the assessors
(Cohen k [95% CI]: 0.81 [0.63-0.96]), comparable to the agree-
ment between assessors in other systematic reviews.15,17 For the
best evidence synthesis, a cutoff score of
6 was chosen for the
PEDro score and CONTENT scale. This literature-based cutoff
point is dubious because methodological quality and therapeutic
validity are far from optimal with a score of 6. Another factor that
limits our understanding of our findings is that the content of the
comparator is often times not or only partly described. Several
reporting guidelines are available to report the content of in-
terventions; these should not be exclusively used to describe the
experimental intervention under study, but also its counterpart(s).
Strengths
Strengths of this study were the data extraction and assessment of
methodological quality and therapeutic validity by 2 reviewers,
which contributed to the accuracy of the review. Furthermore, the
exploration of the content of the intervention by using the CON-
TENT scale (see table 2), a comprehensive table of patients’
characteristics (see table 3), and the time plot of the RCTs (see fig
2) were a valuable addition to earlier reviews in this field.
Implications
Although multidisciplinary rehabilitation after hip fractures is
recommended,1-3 the content of the components to optimize
effectiveness is still unclear, as well as for exercise therapy and the
best context of its delivery. Because of low therapeutic validity in
the current RCTs, it is still unclear whether a specific manner of
home-based exercise therapy could be an effective training in this
older population. It is important to provide patients an exercise
program that matches the individual patients’ problems and with
adequate intensity. In the future, investigators of trials should be
attentive to therapeutic validity of their research and the reporting
of this, possibly in design articles. Well-designed therapeutic in-
terventions focused on patients’ problems and in the daily context
could be more effective than those studied in the RCTs in
this review.
Conclusion
In this systematic review, we found primarily RCTs with low
therapeutic validity (ie, absence of rationale for exercise content
and intensity and unclear adherence to the exercise program),
which may have resulted in interventions that suboptimally
address patients’ abilities and capacities. The article with suffi-
cient therapeutic validity had low methodological quality and only
within-group effects were found. Furthermore, clinical heteroge-
neity was high for the starting point, content, and duration of the
trials. Perhaps the low therapeutic validity explains why there is
only limited evidence in favor of supervised home-based exercise
therapy in older patients 2-10 months after hip fracture in com-
parison to usual care or a nonexercise control interventions. For
future research, it is essential to improve both the methodological
and therapeutic quality.
www.archives-pmr.org
Home-based exercise after hip fracture
Supplier
a. Stata, version 13; StataCorp.
Keywords
Community health services; Elderly; Exercise; Hip fractures;
Rehabilitation
Corresponding author
Thomas J. Hoogeboom, PhD, Radboud Institute for Health
Sciences, IQ Healthcare, Radboud University Medical Center,
P.O. box 9101 (114), 6500 HB Nijmegen, the Netherlands. E-mail
address: thomas.hoogeboom@radboudumc.nl.
References
1. Beaupre LA, Cinats JG, Jones CA, et al. Does functional recovery in
elderly hip fracture patients differ between patients admitted from
long-term care and the community? J Gerontol A Biol Sci Med Sci
2007;62:1127-33.
2. 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.
3. American Academy of Orthopaedic Surgeons (AAOS). Management
of hip fractures in the elderly. 1st ed. Rosemont, IL: American
Academy of Orthopaedic Surgeons; 2014.
4. Van Vugt AB, Van Balen R, Van der Cammen TJ, et al. Richtlijn
Behandeling van de proximale femurfractuur bij de oudere mens.
2008.
5. National Institute for Health and Care Excellence (NICE). The man-
agement of hip fracture in adults. London: National Clinical Guideline
Centre; 2011.
6. World Health Organization. International Classification of Func-
tioning, Disability and Health (ICF). Available at: http://www.who.int/
classifications/icf/en/. Accessed August 1, 2016.
7. Zorginstituut Nederland. Geriatrische revalidatiezorg. Available at:
https://www.zorginstituutnederland.nl/pakket/zvw-kompas/
geriatrischeþrevalidatiezorg. Accessed August 1, 2016.
8. Rijksoverheid. Langer zelfstandig wonen ouderen. Available at: http://
www.rijksoverheid.nl/onderwerpen/ouderenzorg/ouderen-langer-
zelfstandig-wonen. Accessed August 1, 2016.
9. Stolee P, Lim SN, Wilson L, Glenny C. Inpatient versus home-based
rehabilitation for older adults with musculoskeletal disorders: a sys-
tematic review. Clin Rehabil 2012;26:387-402.
10. Giusti A, Barone A, Oliveri M, et al. An analysis of the feasibility of
home rehabilitation among elderly people with proximal femoral
fractures. Arch Phys Med Rehabil 2006;87:826-31.
11. Auais MA, Eilayyan O, Mayo NE. Extended exercise rehabilitation
after hip fracture improves patients’ physical function: a systematic
review and meta-analysis. Phys Ther 2012;92:1437-51.
12. Mehta SP, Roy JS. Systematic review of home physiotherapy after hip
fracture surgery. J Rehabil Med 2011;43:477-80.
13. Herbert RD, Bo K. Analysis of quality of interventions in systematic
reviews. BMJ 2005;331:507-9.
14. Hoogeboom TJ. Meta-analyses on therapeutic physical exercise: time
for a new direction. Phys Ther Rev 2013;18:219-20.
15. Hoogeboom TJ, Oosting E, Vriezekolk JE, et al. Therapeutic validity
and effectiveness of preoperative exercise on functional recovery after
www.archives-pmr.org
13
joint replacement: a systematic review and meta-analysis. PLoS One
2012;7:e38031.
16. Vooijs M, Siemonsma PC, Heus I, Sont JK, Rövekamp TA, van
Meeteren NL. Therapeutic validity and effectiveness of supervised
physical exercise training on exercise capacity in patients with chronic
obstructive pulmonary disease: a systematic review and meta-analysis.
Clin Rehabil 2016;30:1037-48.
17. Snoek JA, van Berkel S, van Meeteren N, Backx FJ, Daanen HA.
Effect of aerobic training on heart rate recovery in patients with
established heart disease; a systematic review. PLoS One 2013;8:
e83907.
18. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items
for systematic reviews and meta-analyses: the PRISMA statement. J
Clin Epidemiol 2009;62:1006-12.
19. Higgins JP, Green S; The Cochrane Collaboration. Cochrane Hand-
book for Systematic Reviews of Interventions version 5.1.. Available
at: http://handbook.cochrane.org/. Accessed August 1, 2016.
20. Crotty M, Unroe K, Cameron ID, Miller M, Ramirez G, Couzner L.
Rehabilitation interventions for improving physical and psychosocial
functioning after hip fracture in older people. Cochrane Database Syst
Rev 2010;(1):CD007624.
21. Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M.
Reliability of the PEDro scale for rating quality of randomized
controlled trials. Phys Ther 2003;83:713-21.
22. Physiotherapy Evidence Database (PEDro). Centre for Evidence-
Based Physiotherapy. Downloads.. Available at: http://www.pedro.
org.au/english/downloads/. Accessed August 1, 2016.
23. Landis JR, Koch GG. The measurement of observer agreement for
categorical data. Biometrics 1977;33:159-74.
24. van Tulder MW, Suttorp M, Morton S, Bouter LM, Shekelle P.
Empirical evidence of an association between internal validity and
effect size in randomized controlled trials of low-back pain. Spine
(Phila Pa 1976) 2009;34:1685-92.
25. Cohen J. Statistical power analysis in the behavioral sciences. 2nd ed.
Hillsdale: Lawrence Erlbaum Associates, Inc; 1988.
26. Heymans MW, van Tulder MW, Esmail R, Bombardier C, Koes BW.
Back schools for non-specific low-back pain. Cochrane Database Syst
Rev 2004;(4):CD000261.
27. Airaksinen O, Brox JI, Cedraschi C, et al. Chapter 4. European
guidelines for the management of chronic nonspecific low back pain.
Eur Spine J 2006;15:s192-300.
28. Latham NK, Harris BA, Bean JF, et al. Effect of a home-based ex-
ercise program on functional recovery following rehabilitation after
hip fracture: a randomized clinical trial. JAMA 2014;311:700-8.
29. Salpakoski A, Tormakangas T, Edgren J, et al. Effects of a multi-
component home-based physical rehabilitation program on mobility
recovery after hip fracture: a randomized controlled trial. J Am Med
Dir Assoc 2014;15:361-8.
30. Orwig DL, Hochberg M, Yu-Yahiro J, et al. Delivery and outcomes of
a yearlong home exercise program after hip fracture: a randomized
controlled trial. Arch Intern Med 2011;171:323-31.
31. 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.
32. Mangione KK, Craik RL, Tomlinson SS, Palombaro KM. Can elderly
patients who have had a hip fracture perform moderate- to high-
intensity exercise at home? Phys Ther 2005;85:727-39.
33. Sherrington C, Lord SR. Home exercise to improve strength and
walking velocity after hip fracture: a randomized controlled trial. Arch
Phys Med Rehabil 1997;78:208-12.
34. Chang F, Latham NK, Ni P, et al. Does self-efficacy mediate func-
tional change in older adults participating in an exercise program after
hip fracture? A randomized controlled trial. Arch Phys Med Rehabil
2015;96:1014-10.e1.
35. Edgren J, Salpakoski A, Sihvonen SE, et al. Effects of a home-based
physical rehabilitation program on physical disability after hip
14
fracture: a randomized controlled trial. J Am Med Dir Assoc 2015;16:
350.e1-7.
36. Portegijs E, Rantakokko M, Edgren J, et al. Effects of a rehabilitation
program on perceived environmental barriers in older patients recov-
ering from hip fracture: a randomized controlled trial. Biomed Res Int
2013;2013:769645.
37. Sipilä S, Salpakoski A, Edgren J, et al. Promoting mobility
after hip fracture (ProMo): study protocol and selected baseline
results of a year-long randomized controlled trial among
community-dwelling older people. BMC Musculoskelet Disord
2011;12:277.
38. Yu-Yahiro JA, Resnick B, Orwig D, Hicks G, Magaziner J. Imple-
mentation of a home-based exercise program post-hip fracture: the
Baltimore hip studies experience. PM R 2009;1:308-18.
I.A.R. Kuijlaars et al
39. Diong J, Allen N, Sherrington C. Structured exercise improves
mobility after hip fracture: a meta-analysis with meta-regression. Br J
Sports Med 2015;50:346-55.
40. de Vries NM, van Ravensberg CD, Hobbelen JS, et al. The Coach2-
Move approach: development and acceptability of an individually
tailored physical therapy strategy to increase activity levels in older
adults with mobility problems. J Geriatr Phys Ther 2015;38:169-82.
41. de Vries NM, Staal JB, van der Wees PJ, et al. Patient-centred physical
therapy is (cost-) effective in increasing physical activity and reducing
frailty in older adults with mobility problems: a randomized controlled trial
with 6 months follow-up. J Cachexia Sarcopenia Muscle 2015;7:422-35.
42. French B, Thomas LH, Leathley MJ, et al. Repetitive task training for
improving functional ability after stroke. Cochrane Database Syst Rev
2007;4:CD006073.
www.archives-pmr.org
Home-based exercise after hip fracture 14.e1

Supplemental Table S1 Searchstring

Searchstrings Medline, Embase, CINAHL up to June 2016
Searchstring e Medline 30-06-2016
ID Search Hits
#7 #3 AND #4 AND #5 AND #6 352
#6 ((((((((((ambulatory care[MeSH Terms]) OR ambulatory care[Title/Abstract]) OR community health 696017
services[MeSH Terms]) OR community based[Title/Abstract]) OR community dwelling[Title/Abstract])
OR community program[Title/Abstract]) OR primary health care[MeSH Terms]) OR primary health care
[Title/Abstract]) OR outpatient*[Title/Abstract]) OR out-patient*[Title/Abstract]) OR home[Title/
Abstract]
#5 ((((((((exercise[MeSH Terms]) OR exercise[Title/Abstract]) OR modalities, physical therapy[MeSH Terms]) 2266182
OR physical therapy[Title/Abstract]) OR physiotherapy[Title/Abstract]) OR rehabilitation[MeSH
Terms]) OR rehabilitation[Title/Abstract]) OR therapy[Title/Abstract]) OR training[Title/Abstract]
#4 (hip fractures[MeSH Terms]) OR (((((hip[Title/Abstract]) OR collum[Title/Abstract]) OR ((((femoral 33161
[Title/Abstract]) OR femur[Title/Abstract])) AND ((neck[Title/Abstract]) OR proximal[Title/
Abstract])))) AND fracture[Title/Abstract])
#3 #1 NOT #2 3306959
#2 (animals[MeSH Terms]) NOT humans[MeSH Terms] 4233866
#1 (((((((randomized controlled trial[Publication Type]) OR controlled clinical trial[Publication Type]) OR 3833673
randomized[Title/Abstract]) OR placebo[Title/Abstract]) OR drug therapy[MeSH Subheading]) OR
randomly[Title/Abstract]) OR trial[Title/Abstract]) OR groups[Title/Abstract]
Searchstring e Embase 30-06-2016
ID Search Hits
#9 #3 AND #4 AND #5 AND #8 660
#8 #6 OR #7 724377
#7 outpatient*:ab,ti 191872
#6 ‘ambulatory care’/exp OR ‘ambulatory care’:ab,ti OR ‘community care’/exp OR ‘community based’:ab,ti OR 558778
‘community dwelling’:ab,ti OR ‘community program’:ab,ti OR ‘primary health care’/exp OR ‘primary
health care’:ab,ti OR ‘out-patient’:ab,ti OR ‘home care’/exp OR home:ab,ti
#5 ‘exercise’/exp OR exercise:ab,ti OR ‘physiotherapy’/exp OR physiotherapy:ab,ti OR ‘physical therapy’:ab,ti 2933519
OR ‘kinesiotherapy’/exp OR ‘rehabilitation’/exp OR rehabilitation:ab,ti OR ‘training’/exp OR
training:ab,ti OR therapy:ab,ti
#4 ‘hip fracture’/exp OR ((hip:ab,ti OR collum:ab,ti OR ((neck:ab,ti OR proximal:ab,ti) AND (femoral:ab,ti OR 48850
femur:ab,ti))) AND fracture:ab,ti)
#3 #1 NOT #2 6102310
#2 ‘animal’/exp NOT ‘human’/exp 4686549
#1 ‘randomized controlled trial’/exp OR ‘randomized controlled trial’ OR ‘controlled clinical trial’/exp OR 6906409
‘controlled clinical trial’ OR ‘randomized’ OR ‘placebo’/de OR ‘placebo’ OR ‘drug therapy’/exp OR ‘drug
therapy’ OR ‘randomly’ OR ‘trial’ OR ‘groups’
Searchstring e CINAHL 30-06-2016
ID Search Hits
S7 S3 AND S4 AND S5 AND S6 214
S6 (MH “ambulatory care”) OR (MH “community health servicesþ”) OR (MH “primary health care”) OR (TI 363889
“ambulatory care”) OR (TI “community based”) OR (TI “community dwelling”) OR (TI “community
program”) OR (TI “primary health care”) OR (TI outpatient*) OR (TI out-patient*) OR (TI home) OR (AB
“ambulatory care”) OR (AB “community based”) OR (AB “community dwelling”) OR (AB “community
program”) OR (AB “primary health care”) OR (AB outpatient*) OR (AB out-patient*) OR (AB home)
S5 (MH “exerciseþ”) OR (MH “rehabilitationþ”) OR (TI exercise) OR (TI "physical therapy") OR (TI 428987
physiotherapy) OR (TI rehabilitation) OR (TI therapy) OR (TI training) OR (AB exercise) OR (AB
"physical therapy") OR (AB physiotherapy) OR (AB rehabilitation) OR (AB therapy) OR (AB training)
S4 (MH “hip fractures”) OR (((TI hip) OR (TI collum) OR (((TI femoral) OR (TI femur)) AND ((TI neck) OR (TI 6828
proximal)))) AND (TI fracture)) OR (((AB hip) OR (AB collum) OR (((AB femoral) OR (AB femur)) AND
((AB neck) OR (AB proximal)))) AND (AB fracture))
S3 S1 NOT S2 656585
S2 (MH animals) NOT (MH humans) 30282
S1 (PT “randomized controlled trial”) OR (PT “clinical trials”) OR (TX randomized) OR (TX placebo) OR (MM 662825
“drug therapy”) OR (TX randomly) OR (TX trial) OR (TX groups)

www.archives-pmr.org
14.e2 I.A.R. Kuijlaars et al

Supplemental Table S2 Results of individual studies on functions, activities and participation

Results of Individual Studies with Strength as Outcome Measure
Source, Outcome* Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)y
Short-term
Mangione et al., 2010
Summed mean (SD) isometric force (electromechanical dynamometer and handheld dynamometer), N
10 weeks 949 (207) 887 (274) 62 (-132.90;256.90)
Mangione et al., 2005
Summed mean (SD) isometric force (handheld digital strain-gauge dynamometer), kg, in subgroup aerobic
12 weeks 67.1 (22.3) 67.7 (22.2) -0.6 (-20.48;19.28)
Summed mean (SD) isometric force (handheld digital strain-gauge dynamometer), kg, in subgroup resistance
12 weeks 59.6 (18.2) 67.7 (22.2) -8.1 (-26.57;10.37)
Sherrington et al., 1997
Mean (SD) quadriceps strength (strap with a spring gauge), kg
Affected leg
1 month 10.4 (4.9) 7.3 (3.7) 3.1 (0.32;5.88)
Non-affected leg
1 month 12.9 (5.7) 9.4 (5.2) 3.5 (0.01;6.99)
Long-term
Latham et al., 2014
Mean (SD) isometric strength of knee extensors (strain gauge dynamometer), lb
Fractured leg
6 months 27.7 (12.8) 26.7 (14.3) 1.0 (-2.83;4.83)
9 months 29.9 (12.9) 25.6 (14.2) 4.3 (0.30;8.30)
Non-fractured leg
6 months 29.4 (13.3) 28.8 (13.9) 0.6 (-3.24;4.44)
9 months 30.8 (13.7) 25.8 (13.6) 5.0 (0.97;9.03)
Mangione et al., 2010
Summed mean (SD) isometric force (electromechanical dynamometer and handheld dynamometer), N
26 weeks 1073 (167) 927 (354) 146.0 (-72.68;364.68)
95%-CI: 95% confidence interval; SD: standard deviation.
* Short-term is  four months and long-term is > four months.
y
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

www.archives-pmr.org
Home-based exercise after hip fracture 14.e3

Results of Individual Studies with Balance as Outcome Measure*

y
Source, Outcome Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)z
Short-term
Salpakoski et al., 2014
Mean (SD) BBS
3 months 43.63 (6.22) 45.53 (6.98) -1.89 (-4.87;1.07)
Sherrington et al., 1997
Mean (SD) postural sway on floor, mm
1 month 124 (69) 136 (87) -12 (-62.26;38.26)
Mean (SD) postural sway on foam, mm
1 month 298 (161) 285 (159) 13 (-114.72;140.72)
Mean (SD) functional reach, cm
1 month 15.7 (7.9) 16.9 (7.7) -1.2 (-6.19;3.79)
Long-term
Latham et al., 2014
Mean (SD) BBS
6 months 44.4 (10.7) 41.1 (10.7) 3.3 (0.28;6.32)
9 months 45.6 (10.0) 40.4 (11.4) 5.2 (2.04;8.36)
Salpakoski et al., 2014
Mean (SD) BBS
6 months 44.26 (6.22) 45.72 (8.07) -1.45 (-4.71;1.80)
12 months 44.81 (6.14) 44.15 (9.92) 0.66 (-3.10;4.42)
95%-CI: 95% confidence interval; BBS: Berg Balance Scale; SD: standard deviation.
* Results of Orwig et al. (2011) were not presented because they published only change scores.
y
Short-term is  four months and long-term is > four months.
z
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

Results of Individual Studies with Gait as Outcome Measure*

y
Source, Outcome Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)z
Short-term
Mangione et al., 2010
Mean (SD) gait speed usual, m/s
10 weeks 0.81 (0.17) 0.70 (0.22) 0.11 (0.05;0.27)
Mean (SD) fast gait speed, m/s
10 weeks 1.07 (0.23) 0.91 (0.25) 0.16 (-0.03;0.35)
Mangione et al., 2005
Mean (SD) gait speed free, m/s, in subgroep aerobic
12 weeks 0.79 (0.26) 0.65 (0.23) 0.14 (-0.08;0.36)
Mean (SD) gait speed free, m/s, in subgroep resistance
12 weeks 0.71 (0.28) 0.65 (0.23) 0.06 (-0.18;0.30)
Sherrington et al., 1997
Mean (SD) comfortable gait speed, m/s
1 month 0.51 (0.34) 0.50 (0.35) 0.01 (-0.21;0.23)
Long-term
Mangione et al., 2010
Mean (SD) gait speed usual, m/s
26 weeks 0.81 (0.17) 0.67 (0.21) 0.14 (-0.01;0.29)
Mean (SD) fast gait speed, m/s
26 weeks 1.11 (0.22) 0.89 (0.24) 0.22 (0.03;0.41)
95%-CI: 95% confidence interval; SD: standard deviation.
* Results of Orwig et al. (2011) were not presented because they published only change scores.
y
Short-term is  four months and long-term is > four months.
z
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

www.archives-pmr.org
14.e4 I.A.R. Kuijlaars et al

Results of Individual Studies With Endurance as Outcome Measure*

y
Source, Outcome Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)z
Short-term
Mangione et al., 2010
Mean (SD) 6MWT, m
10 weeks 295.7 (79.8) 242.7 (83.0) 53.0 (-13.0;119.0)
Mangione et al., 2005
Mean (SD) 6MWT, m, in subgroup aerobic
12 weeks 321.1 (101.7) 266.2 (82.4) 54.9 (-28.62;138.42)
Mangione et al., 2005
Mean (SD) 6MWT, m, in subgroup resistance
12 weeks 278.9 (114.6) 266.2 (82.4) 12.7 (-79.34;104.74)
Long-term
Mangione et al., 2010
Mean (SD) 6MWT, m
26 weeks 299.5 (80.0) 219.4 (67.8) 80.1 (19.48;140.72)
6MWT: six minutes walk test; 95%-CI: 95% confidence interval; SD: standard deviation.
* Results of Orwig et al. (2011) were not presented because they published only change scores.
y
Short-term is  four months and long-term is > four months.
z
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

Results of Individual Studies with ADL as Outcome Measure*

y
Source, Outcome Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)z
Short-term, performance-based
Mangione et al., 2010
Mean (SD) mPPT
10 weeks 27.4 (3.2) 21.3 (5.8) 6.1 (2.38;9.82)
Short-term, self-reported
Salpakoski et al., 2014
Mean (SD) Evergreens activities of daily living index ADL
3 months 3.4 (3.2) 3.0 (3.0) 0.4 (-1.03;1.83)
Mangione et al., 2010
Mean (SD) SF-36 physical function score
10 weeks 50.4 (15.6) 40.4 (14.2) 10.0 (-2.16;22,16)
Mangione et al., 2005
Mean (SD) SF-36 physical function score, in subgroup aerobic
12 weeks 57.5 (24.3) 48.0 (18.9) 9.5 (-10.18;29.18)
Mean (SD) SF-36 physical function score, in subgroup resistance
12 weeks 57.7 (21.1) 48.0 (18.9) 9.7 (-8.67;28.07)
Long-term, performance-based
Mangione et al., 2010
Mean (SD) modified Physical Performance Test
26 weeks 27.6 (2.6) 20.3 (6.2) 7.3 (3.55;11.05)
Long-term, self-reported
Latham et al., 2014
Mean (SD) AM-PAC daily activity
6 months 61.3 (15.7) 58.6 (15.3) 2.7 (-1.68;7.08)
9 months 63.0 (15.9) 59.0 (14.9) 4.0 (-0.56;8.56)
Salpakoski et al., 2014
Mean (SD) Evergreens activities of daily living index ADL
6 months 3.0 (3.2) 3.4 (3.7) -0.4 (-1.97;1.17)
12 months 3.6 (4.1) 3.0 (3.7) 0.6 (-1.20;2.40)
(continued on next page)

www.archives-pmr.org
Home-based exercise after hip fracture 14.e5

(continued )
Results of Individual Studies with ADL as Outcome Measure*
y
Source, Outcome Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)z
Mangione et al., 2010
Mean (SD) SF-36 physical function score
26 weeks 56.8 (19.6) 38.3 (19.2) 18.5 (2.73;34.27)
95%-CI: 95% confidence interval; ADL: activities of daily living; AM-PAC: Activity Measure for Post-Acute Care; mPPT: Modified Physical Performance
Test; SD: standard deviation; SF-36: Medical Outcomes Study 36-Item Short Form Health Survey.
* Results of Orwig et al. (2011) were not presented because they published only change scores.
y
Short-term is  four months and long-term is > four months.
z
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

Results of Individual Studies with IADL as Outcome Measure*

y
Source, Outcome Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)z
Short-term, self-reported
Salpakoski et al., 2014
Mean (SD) Evergreens activities of daily living index IADL
3 months 7.1 (7.1) 5.5 (6.0) 1.6 (-1.42;4.62)
Long-term, self-reported
Salpakoski et al., 2014
Mean (SD) Evergreens activities of daily living index IADL
6 months 6.4 (6.6) 7.2 (7.8) -0.8 (-4.08;2.48)
12 months 6.8 (7.7) 6.5 (7.1) 0.3 (-3.11;3.71)
95%-CI: 95% confidence interval; IADL: instrumental activities of daily living; SD: standard deviation.
* Results of Orwig et al. (2011) were not presented because they published only change scores.
y
Short-term is  four months and long-term is > four months.
z
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

www.archives-pmr.org
14.e6 I.A.R. Kuijlaars et al

Results of Individual Studies with Mobility as Outcome Measure

Source, Outcome* Home-Based Exercise Group Control Group Mean Difference or Relative Risk (95%-CI)y
Short-term, performance-based
Salpakoski et al., 2014
Mean (SD) SPPB
3 months 6.69 (2.89) 7.26 (2.37) -0.57 (-1.76;0.61)
Short-term, self-reported
Salpakoski et al., 2014
Number (%) able to negotiate stairs without difficulties
3 months 23 (64) 28 (72) 0.89 (0.65;1.22)
Number (%) no perceived entrance-related barriers
3 months 23 (59) 28 (70) 0.84 (0.61;1.17)
Number (%) no perceived outdoor barriers
3 months 13 (33) 11 (28) 1.21 (0.62;2.37)
Long-term, performance-based
Latham et al., 2014
Mean (SD) SPPB
6 months 7.2 (3.0) 6.2 (3.0) 1.0 (0.15;1.85)
9 months 7.6 (2.9) 6.3 (2.9) 1.3 (0.44;2.16)
Salpakoski et al., 2014
Mean (SD) SPPB
6 months 7.18 (2.70) 7.36 (2.42) -0.18 (-1.33;0.98)
12 months 7.18 (2.93) 7.22 (2.51) -0.04 (-1.28;1.20)
Long-term, self-reported
Latham et al., 2014
Mean (SD) AM-PAC basic mobility
6 months 58.1 (7.9) 56.6 (8.1) 1.5 (-0.76;3.76)
9 months 59.5 (9.3) 56.7 (7.6) 2.8 (0.28;5.32)
Salpakoski et al., 2014
Number (%) able to negotiate stairs without difficulties
6 months 28 (76) 23 (62) 1.22 (0.89;1.66)
12 months 27 (75) 19 (49) 1.54 (1.06;2.24)
Number (%) no perceived entrance-related barriers
6 months 28 (72) 26 (67) 1.08 (0.80;1.45)
12 months 29 (76) 31 (79) 0.96 (0.76;1.22)
Number (%) no perceived outdoor barriers
6 months 19 (49) 12 (31) 1.58 (0.90;2.80)
12 months 18 (47) 16 (41) 1.15 (0.70;1.91)
95%-CI: 95% confidence interval; AM-PAC: Activity Measure for Post-Acute Care; SD: standard deviation; SPPB: Short Physical Performance Battery.
* Short-term is  four months and long-term is > four months.
y
Results were presented in bold if the mean difference or relative risk is statistical significant (p<0.05).

www.archives-pmr.org