HHS Public Access

Author manuscript
Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
Author Manuscript

Published in final edited form as:

Am J Phys Med Rehabil. 2023 January 01; 102(1): 11–18. doi:10.1097/PHM.0000000000002007.

Rehabilitation for Total Hip Arthroplasty: A Systematic Review

Kristin J. Konnyu, BSc MSc BEd PhDa, Dan Pinto, PT, DPT, PhDb, Wangnan Cao, PhDa,
Roy K. Aaron, MDc, Orestis A. Panagiotou, MD, PhDa, Monika Reddy Bhuma, BDS, MPHa,
Gaelen P. Adam, MLIS, MPHa, Ethan M. Balk, MD, MPHa, Louise M. Thoma, PT, DPT, PhDd
aCenterfor Evidence Synthesis in Health, Department of Health Services, Policy, and Practice,
Brown University School of Public Health, Providence, Rhode Island, USA
Author Manuscript

bDepartment of Physical Therapy, Marquette University, Milwaukee, Wisconsin

cDepartment of Orthopaedic Surgery, Warren Albert Medical School of Brown University,
Providence, Rhode Island; Orthopedic Program in Clinical/Translational Research, Warren Albert
Medical School of Brown University, Providence, Rhode Island; Miriam Hospital Total Joint
Replacement Center, Providence, Rhode Island.
dDivision
of Physical Therapy, Department of Allied Health Sciences, University of North Carolina
at Chapel Hill, Chapel Hill, North Carolina

Abstract
We sought to determine the comparative benefits and harms of rehabilitation interventions for
patients who have undergone elective, unilateral THA surgery for the treatment of primary
Author Manuscript

osteoarthritis. We searched PubMed, Embase, The Cochrane Register of Clinical Trials, CINAHL,
PsycINFO, Scopus, and ClinicalTrials.gov from January 1, 2005 through May 3, 2021. We
included randomized controlled trials and adequately-adjusted nonrandomized comparative
studies of rehabilitation programs reporting performance-based, patient-reported, or healthcare
utilization outcomes. Three researchers extracted study data and assessed risk of bias, verified
by an independent researcher. Experts in rehabilitation content and complex interventions
independently coded rehabilitation interventions. The team assessed strength of evidence (SoE).
Large heterogeneity across evaluated rehabilitation programs limited conclusions. Evidence from
15 studies suggests diverse rehabilitation programs may not differ in terms of risk of harm or
outcomes of pain, strength, activities of daily living, or quality of life (all low SoE). Evidence is
insufficient for other outcomes. In conclusion, no differences in outcomes were found between
different rehabilitation programs following THA. Further evidence is needed to inform decisions
Author Manuscript

on what attributes of rehabilitation programs are most effective for various outcomes.

CORRESPONDING AUTHOR: Kristin J. Konnyu, BSc MSc BEd PhD, 121 S. Main Street, Box G-121-8, Providence, RI 02903,
USA, Center for Evidence Synthesis in Health, Department of Health Services, Policy, and Practice, Brown University School of
Public Health, Providence, Rhode Island, USA, kristin_konnyu@brown.edu.
CONFLICTS OF INTEREST
The authors have no conflicts of interest to disclose.
FINANCIAL DISCLOSURES
The authors have no relevant financial relationships to disclose.
PROTOCOL REGISTRATION
This systematic review was prospectively registered through PROSPERO (Registration Number: CRD42020199102)
 Konnyu et al. Page 2

Keywords
Author Manuscript

rehabilitation; total hip arthroplasty; total hip replacement; osteoarthritis; systematic review;
complex intervention

Background
Standard care for adults who have undergone total hip arthroplasty (THA) for end-
stage osteoarthritis is to offer rehabilitation after surgery, with the goal of optimizing
postoperative function, reducing pain, and returning to normal activities of daily living
(ADL). Rehabilitation programs are complex interventions that incorporate multiple specific
components (i.e., resistance exercises, gait training, etc.) that may be initiated at different
times after surgery, performed at different frequencies and intensities, delivered by different
Author Manuscript

personnel in different settings, and may be personalized to individual, social and financial
circumstances and responses to surgery and rehabilitation. Thus, while rehabilitation
following THA in general has been shown to be effective1 it is unclear which specific
rehabilitation intervention(s), or components within interventions are most effective and
should be replicated in practice to achieve the best clinical outcomes and reduce avoidable
complications or joint failures.2

We conducted a systematic review (SR)3 under the Agency for Healthcare Research
and Quality (AHRQ) Evidence-based Practice Center (EPC) Program to understand the
effectiveness and comparative effectiveness of prehabilitation and rehabilitation for total
knee arthroplasty and THA (see questions for the full review in Appendix A). In this
paper, we address the comparative effects and harms of rehabilitation for patients who have
undergone THA on patient-reported outcomes, performance-based outcomes, and healthcare
Author Manuscript

utilization. In companion articles, we address prehabilitation and rehabilitation after total

knee arthroplasty.4, 5

Methods
The Brown Evidence-Based Practice Center (EPC) used established SR methodologies as
outlined in AHRQ Methods Guide.6 The SR was registered in PROSPERO (registration
number CRD42020199102) and is reported in accordance with the Preferred Items for
Reporting in Systematic Reviews and Meta-Analyses7 (Appendix B). We developed the
protocol with input from a broad panel of stakeholders (https://effectivehealthcare.ahrq.gov/
products/major-joint-replacement/protocol). Detailed descriptions of the SR and methods
can be found in the full AHRQ report3 and Appendix C.
Author Manuscript

Data Sources and Searches

For the full SR, we conducted literature searches in Medline (via PubMed), PsycINFO,
Embase, The Cochrane Register of Clinical Trials, CINAHL, and Scopus, restricted to
January 1, 2005 through May 3, 2021. Studies published prior to 2005 were excluded
to account for changes in THA management, particularly postoperative protocols (e.g.,
enhanced recovery after surgery protocol and rapid hospital discharge) more commonly

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 3

employed since the 2000s.8 No language restrictions were employed. See Appendix C for
Author Manuscript

search strategies.

All identified citations were independently double-screened using Abstrackr (http://

abstrackr.cebm.brown.edu) by a team of six researchers. Conflicts were resolved by group
discussion. All potentially relevant studies were rescreened in full text in duplicate.

Study Selection
We included studies of adults who had undergone elective non-revision, unilateral THA
for primary osteoarthritis and who received a rehabilitation program within 6 months
after surgery. The 6-month timepoint was selected with stakeholder input to restrict to
evidence of programs that were more recent to the THA surgery. We excluded studies
of patients with non-elective (e.g., emergency), bilateral, or revision THA, and THA for
conditions other than primary osteoarthritis. We defined rehabilitation as active, structured
Author Manuscript

physical activity or activities designed to attain measurable goals of reducing impairments

and improving movement-related function as defined by the International Classification of
Function, Disability, and Health (2018).9 The rehabilitation intervention had to be delivered,
supervised, and/or monitored by a healthcare professional or other trained individual (e.g.,
physical therapist, nurse trained in rehabilitation, health educator with training in exercise
delivery or rehabilitation). Rehabilitation could be combined with adjunctive modalities (ice,
heat, etc.), although we excluded studies designed to evaluate solely adjunctive modalities.

Rehabilitation could be compared with other rehabilitation (of different content), similar
rehabilitation with an additional adjunctive modality, or rehabilitation of similar content that
varied in terms of intensity or delivery personnel or delivery setting. We did not evaluate
pharmaceutical or over-the counter interventions as co-interventions with the rehabilitation
Author Manuscript

interventions.

We included performance-based measures, patient-reported outcomes, healthcare utilization,

and harms (specifically from rehabilitation). We included randomized controlled trials
(RCT) and nonrandomized comparative studies (NRCS) that used analytic methods to
minimize selection bias (e.g., multivariable adjustment, propensity score analysis). We
required at least 20 participants per intervention group.

Data Extraction and Quality Assessment

From each eligible study, we extracted data on study design, population characteristics,
outcomes, results, and quality assessment into the Systematic Review Data Repository
(SRDR) (https://srdrplus.ahrq.gov/public_data?id=2965&type=project). Two independent
Author Manuscript

reviewers (one with expertise in rehabilitation content, the other with expertise in complex
intervention taxonomies) separately extracted intervention details. Based on stakeholder
input, we restricted outcome to 6 months or later.

We first categorized the content of the rehabilitation interventions according to a

categorization scheme defined by Oatis and Franklin.10, 11 The taxonomy comprehensively
lists 147 specific rehabilitation content components that are hierarchically linked to larger
rehabilitation goals. The larger component goals include: strengthening exercise, aerobic

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 4

exercise, flexibility exercise, balance-motor/learning-agility exercise, task specific training,

Author Manuscript

patient education, and adjunctive modalities. We extracted information on the goals of the
exercise(s) and the specific exercises used to achieve these goals; information on whether the
rehabilitation was progressed (i.e., advanced over time) and, if so, whether it was appropriate
(i.e., according to patient-specific parameters assessed by the therapist and not uniform for
all participants); and who delivered the intervention (personnel), how (mode of delivery),
and where (setting). We did not assess dose, intensity, and duration aside from minimal
criteria needed to meet our rehabilitation definition or in cases where rehabilitation content
was the same but varied only in terms of dose, intensity or duration. More details about how
we operationalized the taxonomy and our extractions of the rehabilitation interventions can
be found in Appendix C. We categorized studies by whether rehabilitation occurred during
the acute postoperative period (<2 weeks) or post-acute (≥2 weeks).

To assess study methodological quality, we used the Cochrane risk of bias tool for RCTs12
Author Manuscript

and for NRCSs, elements from the ROBINS-I Tool13 related to confounding and selection
bias. For RCTs, we also used items from the National Heart, Lung, and Blood Institute
(NHLBI) tool on the adequacy of descriptions of study eligibility criteria, interventions,
and outcomes.14 Each study was extracted and assessed for methodological quality by one
methodologist and reviewed and confirmed by at least one other experienced methodologist.
Disagreements were resolved by discussion among the team.

Data Synthesis and Analysis

We compared interventions with their comparators for their effects, using mean difference
(MD) between groups in continuous outcome data. For categorical outcomes, we evaluated
odds ratios. Meta-analysis was not feasible due to intervention and outcome heterogeneity.
Author Manuscript

For all analyses, we graded the strength of evidence (SoE) as per the AHRQ Methods
Guide.15, 16 For each SoE assessment, we considered the number of studies, their designs,
limitations/risk of bias, the directness of the comparisons to the research question,
consistency of study results, precision of estimates of effect, likelihood of reporting bias,
and other limitations. Based on these assessments, we assigned an SoE rating as being either
high, moderate, or low, or there was insufficient evidence to allow a conclusion.

Role of the Funding Source, Stakeholders, and Reviewers

The AHRQ Learning Health System (LHS) panel nominated this topic for SR to AHRQ.
A member of the LHS panel joined stakeholder panels who contributed to discussions that
helped refine the key questions and protocol. AHRQ program officers, the LHS member,
and other reviewers (both invited and public) provided comments on draft versions of the
Author Manuscript

protocol and full evidence report. The LHS partner and AHRQ did not participate in the
literature search, determination of study eligibility criteria, data analysis, study evaluation,
or interpretation of findings. After completion of the AHRQ report,3 we discussed draft
versions of the manuscript with the LHS partner to ensure it met the decision-making needs
of LHSs and properly conveyed our conclusions. The LHS partner did not draft any portion
of the manuscript or attempt to alter conclusions.

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 5

Results
Author Manuscript

The literature searches yielded 22,361 citations (for all topics addressed in the full report).
We found 1,016 citations to retrieve for full-text screening (Appendix D) of which 15 studies
(14 RCTs and 1 NRCS) in 20 articles addressed rehabilitation after THA.17–36

Details on study designs, population characteristics, and detailed coding of interventions

evaluated can be found in Appendices E and F. Of note, as the review was focused on the
comparative effectiveness of different rehabilitation programs after THA, and thus we coded
the rehabilitation content, where reported, of all arms. The average ages of participants
ranged from 54 to 71 years across studies. Between 21% and 70% of study participants
were women (median 60%). Only one study27 reported the proportion of patients who had
undergone prior contralateral THA (25%). Studies enrolled between 54 and 280 participants
each (median=97). Most studies were conducted in Europe (n=10), three in Australia, and
Author Manuscript

two in the U.S. Studies were mostly at moderate risk of bias due to lack of blinding. Four
RCTs were at high risk of bias due to additional threats to the randomization process.
The one included NRCS was adjusted for important confounders as this was part of our
eligibility criteria and was deemed moderate risk of bias (Appendices G and H).

We did not find a single rehabilitation intervention that was evaluated in more than one
study; all included RCTs evaluated a different rehabilitation program comprised of different
components delivered in varying settings, by diverse personnel, and at various points during
the rehabilitation period (Figure 1, Appendix F). For this reason, we did not perform
meta-analyses of outcomes as we decided with our stakeholders that an “average effect” of
a single concept of “rehabilitation” was not accurate given the notably diverse interventions
identified. Meta-regression (an alternative evidence synthesis approach to explore the impact
Author Manuscript

of specific features of interventions on study outcomes) was not feasible due to the
heterogeneity of outcomes.

We, thus, summarized studies based on post-surgical timing (acute and post-acute).
Four RCTs delivered rehabilitation during the acute phase (<2 weeks) and 11 studies
(10 RCTs and 1 NRCS) during the post-acute phase (≥2 weeks). Of the four acute
rehabilitation studies, two evaluated more intensive rehabilitation programs versus less
intensive rehabilitation programs and two evaluated similar programs with varying timing
and settings of delivery (e.g., early vs. late aquatic therapy, early supervised vs. home
unsupervised). Of the 11 post-acute rehabilitation studies, six evaluated novel rehabilitation
programs versus standard care (variously defined) or alternative rehabilitation programs, and
five studies evaluated similar programs different settings or personnel.
Author Manuscript

The content of the specific rehabilitation programs varied across interventions. Most
interventions included exercises to address the goal components of strength (14/15 studies;
21/31 arms) and flexibility (11/15 studies; 15/31arms). Fewer studies included components
to address task-specific training (8/15 studies; 13/31 arms), patient education (7/15
studies; 11/31 arms), and balance-motor-learning-agility (5/15 studies; 7/31 arms). Few
studies included exercises targeted at aerobic endurance (3/15 studies; 3/31 arms). Three
studies included an adjunctive modality, including massage and mindfulness/stress reducing

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 6

activities (in 4 arms). Seven of the 15 studies reported some form of progression, of which
Author Manuscript

we assessed five to be appropriate progression (i.e., patients were not progressed based on
patient-specific parameters).

Rehabilitation interventions were delivered by physical therapists in 11 studies. The

remaining four studies did not report on personnel, explicitly.18, 20, 26, 29 No studies
reported the use of non-physical therapist personnel to deliver interventions. In fourteen
studies, the rehabilitation intervention was delivered to patients in-person for at least one of
the study arms. Five studies had self-guided components in which patients performed self-
directed exercise and one study evaluated rehabilitation delivered remotely via telephone.
Interventions were delivered in varying settings (and often, a combination of settings). Six
studies evaluated one or both rehabilitation arms exclusively in acute inpatients 33, 34 or
outpatient patients.18, 20–22, 24–26, 28, 35, 36 Three studies compared rehabilitation programs
delivered in outpatient and home settings versus home alone.17, 19, 30–32 One study evaluated
Author Manuscript

a rehabilitation program initiated in a fitness center followed by home exercise compared

to home exercises alone27 and another compared inpatient versus no inpatient care but did
not report the specifics of the type of setting for each arm (acute vs. some other inpatient
facility).29

Specific codes for intervention (and control arm, where present) goals and exercises, use of
progression (and assessment of appropriateness), and details on personnel, mode of delivery,
and setting are detailed in Figures 2 and 3 and Appendix F.

Comparative Effectiveness of Rehabilitation Interventions

Evidence from the included fifteen studies suggests diverse rehabilitation programs
(delivered in the acute and post-acute phase) may not differ in terms of pain, strength,
Author Manuscript

activities of daily living, quality of life (QoL) or risk of harms (all low SoE) (Table
1). Evidence is insufficient on the impact of rehabilitation on hip range of motion and
satisfaction with care. Interventions and outcomes were too heterogeneous to explore
whether outcomes varied by the presence of specific factors (i.e., intervention content,
personnel, setting, progression). Complete data on all outcomes by outcome domain (e.g.,
pain, strength, QoL) can be found in Appendix I.

Body Structure and Function Outcomes

Twelve studies reported on body structure and function outcomes (symptoms, pain, range
of motion, muscle strength, energy and vigor, and emotional functioning).17–20, 22–25, 27–32
Of the nine studies18, 20, 22–25, 27, 28, 30–32 that reported data on symptoms (various scales
including the Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC]
Author Manuscript

and Hip Disability and Osteoarthritis Outcome Score [HOOS]), seven18, 20, 22, 23, 25, 27, 30–32
observed no differences between groups at follow-up ranging from six months to five years
post-THA. The two studies that reported differences found reduced stiffness with ergometer
cycling and “standard physical therapy” compared with “standard physical therapy” alone
with (MD −5.2; P=0.047)24 and with task-oriented exercises compared with open chain
kinetic exercises (MD −9.0, P=0.037).28

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 7

Eleven studies reported pain data using six different measurement instruments (the pain
Author Manuscript

components of HOOS and WOMAC, the bodily pain component of the 36-Item Short
Form Survey [SF-36], the visual analogue scale [VAS] of the EuroQol-5 dimensions, VAS,
and the Latinen scale measuring pain intensity).18, 20–32 As with symptoms, most studies
(n=8) found no difference in pain between comparison groups. Three studies18, 20, 28
reported statistically significant reduced pain in their respective intervention groups (sport
therapy; hydrotherapy; task-oriented exercises) compared with controls (no sports therapy;
no hydrotherapy; open-chain kinetic exercises; respectively). Effective programs in these
three studies varied in terms of their component goals but all included strength and
flexibility.

Five studies or fewer reported data on strength (n=5)18, 21, 22, 26, 27, 30, 31, emotional
functioning (n=4) 17, 19, 28, 30, 31, range of motion (n=3)21, 22, 26, 35, 36, and energy and vigor
(n=1)28 and generally observed comparable effects between intervention groups at follow-up
Author Manuscript

ranging from 6 months to 5 years after THA.

Activity and Participation Outcomes

Thirteen studies reported on activity and participation outcomes (physical function and
activities of daily living, transfers, mobility, and timed up and go test).17–25, 27, 28, 30–36

Twelve studies17–25, 27, 28, 30–34 reported data on patient-reported physical function and
ADLs using various scales (function component of WOMAC, ADL and sports and
recreation scales of the HOOS, physical component scale of the SF-36) measured at follow-
ups between 6 months and 5 years after THA surgery. Eight studies found no difference
between groups in terms of patient-reported function and ADL17–19, 25, 27, 30–34; four
studies20–24, 28 reported significant differences between groups. Interventions associated
Author Manuscript

with significant improvements in at least one measure of function and ADL included
a hydrotherapy-based program,20 a walking skills program, 21–23 ergometer cycling
program,24 an early supervised and task-oriented exercise program.28

Five studies reported data on mobility19, 21, 22, 27, 30, 31, 35, 36 and two studies reported data
on transfers 22, 23, 27 using various outcome measures (e.g., 6-meter walk test, 40-meter
fast paced walk test, stair climb tests, step test, figure eight test, speed in a 5-meter
walkway, timed up-and-do test, and step length, 20-second time chair stand test). Three
studies reported no differences between groups. Heiberg 2012 found patients randomized
to the walking skill training program performed significantly better than “standard care”
patients on the 6-meter walk test (mean difference [MD] 52 meters, P<0.001) at 12 months21
although the effect was no longer significant at 5 years.23 Nelson et al. reported patients
Author Manuscript

randomized to telerehabilitation performed the timed-up-and go test significantly quicker

than patients randomized to in-person usual care (MD −1.5 seconds, 95% CI −2.4 to
−0.6).30, 31

Other Patient-Reported Outcomes

Twelve studies reported on other patient-reported outcomes (health-related quality of life,
patient satisfaction with care, patient global assessments).17–19, 21–25, 27–32, 35, 36 Four
studies reported data on QoL (using the QoL component of HOOS)21–23, 27, 30–32 and two

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 8

studies reported data on satisfaction with care;24, 25; no differences were observed in these
outcomes between groups. Eleven studies17–19, 21–25, 28, 29, 32–36 reported data on patients’
Author Manuscript

self-reported global assessment of their health using ten different measurement instruments.
The studies mostly observed no differences between groups at follow-up times between 6
months and 5 years after THA. Beck 2019 found a significant difference in favor of sports
therapy compared with control (not defined) on the EQ-5D scale at 6-month follow-up, but
this difference was not sustained at the 12-month follow-up.

Healthcare Utilization Outcomes

Few studies reported on healthcare-utilization outcomes following rehabilitation compared
to various controls. Two studies reported on the number of participants who needed to be
readmitted to hospital due to various adverse events within three months of THA and found
rates were low (4 to 10%, including post-surgical adverse events) and comparable between
Author Manuscript

groups.24, 25

Harms From Rehabilitation

Only four studies reported the occurrence of adverse events associated with the
rehabilitation programs (e.g., pain, falls).24, 25, 27, 28 Of these, the proportion of patients
experiencing harms due to participation in the rehabilitation interventions were low (usually
10% or lower within intervention and comparison arms) and comparable in number between
intervention and control groups.

Heterogeneity
Only two studies formally analyzed whether effects from rehabilitation versus its various
comparators differed among subgroups and focused on factors associated with the implant
Author Manuscript

itself. Lyp 201626 found that the type of prosthesis used (cemented or uncemented) did not
impact the relative effects of rehabilitation on range of motion or strength. Smith 200933, 34
found outcomes from the Iowa Level of Assistance scale and SF-12 did not vary among
groups by the type of prosthesis used (cemented or uncemented) or surgical approach
(posterior or anterolateral). No study reported heterogeneity of intervention effects due to
patients’ baseline risk (e.g., patients with higher vs. lower measures of strength, flexibility,
function, at baseline).

Discussion
Based on evidence from 15 studies included in our review, it remains unclear which
rehabilitation program should be chosen for patients who have undergone THA, or what
specific attributes of programs are effective in achieving optimal patient outcomes. Each
Author Manuscript

study evaluated unique rehabilitation interventions, which differed in content and delivery
method (e.g., timing, progression, personnel, and setting) and reported the effects of these
interventions on disparate outcomes measures. Risk of bias was moderate at best largely
due to the challenge of blinding of rehabilitation interventions. Given this, we found low
SoE of comparable effects among diverse rehabilitation programs in terms of outcomes of
pain, strength, ADLs, QoL, or risk of harms. There was insufficient evidence on outcomes
of hip range of motion or satisfaction with care, precluding conclusions on these outcomes.

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 9

And for all outcomes there was insufficient evidence to explore the effects of any specific
Author Manuscript

attributes of rehabilitation programs.

The findings of this review are consistent with two earlier systematic reviews published
in 2009 and 2013 which noted that the evidence “neither supported nor denied” 1 the
effectiveness of different rehabilitation protocols after THA and there is insufficient
evidence “to build up a detailed evidence-based exercise protocol”, respectively.2 The
persistence of systematic review findings over 12 years despite the accumulation of more
evidence should give stakeholders pause. It means that patient-care and healthcare services
for nearly half a million adults who undergo THA annually (estimated to increase to 635,000
by 2030),37 is lacking an evidence-base to guide practice and new studies are not improving
on the limitations of earlier ones.

A significant challenge in advancing the science of rehabilitation following THA is that

Author Manuscript

it is a very complex intervention. For example, the detailed coding of interventions

included in this review confirm concretely that “rehabilitation” is not a singularly conceived
intervention, but the composite of multiple different intervention components that can be
combined in different ways across episodes of care to achieve different rehabilitation goals.
In addition to the “active ingredients” of the rehabilitation content, the various rehabilitation
interventions can be delivered different ways – in different settings by different personnel
at different periods (and intensities) following surgery. Such complex interventions imply
a complex “response surface”38 of multiple interrelated variables, which depending on
their combination and potential interactions, would be expected to lead to varying effects.
Primary studies alone cannot answer the multifaceted questions stakeholders want to know
(what specific program parts work and under what conditions that are captured in this
“response surface”). Rather, a coordinated and systematic program of primary research
Author Manuscript

paired with evidence synthesis (including meta-regression models) are needed to isolate the
interependent effects of various factors on key outcomes of interest to the field. To work
however, rehabilitation studies moving forward would need to address two main limitations
of the evidence to date: 1) lack of standardized intervention reporting and 2) lack of
standardized outcomes.

Intervention description and reporting

A major limitation to the synthesis of studies of rehabilitation following THA was the
poor description of rehabilitation interventions evaluated and lack of standardization of
terminology for its various components. How can we know what works if we don’t know
‘what’ (i.e., rehabilitation) is or have a shared method of describing the core elements?
Moreover, how can we understand the impact of methods of delivery (e.g., settings
Author Manuscript

and providers) without first understanding what rehabilitation is being delivered? (i.e.,
if differences exist in the outcomes, how do we know if these are due to differences
in the setting, the rehabilitation program, or both?). Unfortunately, studies to date
have employed different terminologies to define their rehabilitation programs and these
descriptions have varied extensively in the level of detail provided. Some studies use
few words like “inpatient rehabilitation” while other studies provide comprehensive multi-
page supplemental protocols to detail the contents of the rehabilitation. Standardization of

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 10

terminology is also lacking. Consistent with other complex scientific problems, the science
Author Manuscript

of rehabilitation can only advance if the field as a collective can agree upon and use a shared
taxonomy to define intervention content and consistently report content in a complete and
transparent manner. As a starting point, we would recommend trialists consider using the
taxonomy defined by Oatis and Franklin used in this review to define intervention content,10
and the Template for Intervention Description and Replication (TIDieR) checklist to ensure
comprehensive reporting of all intervention elements.39

Standardized outcomes
Adding to the complexity of the diverse interventions, studies reported highly diverse
outcomes. These not only included different outcome domains, but also different scales
and metrics within shared outcome domains, at various points of follow-up (6 months
to 5 years). Learning across studies was difficult since we could not adequately evaluate
Author Manuscript

comparable outcomes across studies. Further, several outcomes prioritized by stakeholders,

as well as other experts in the field, 40, 41 (e.g., strength, range of motion, satisfaction with
care, QoL, healthcare utilization) were reported infrequently. Additional work is needed to
define what core outcomes are essential to evaluation of rehabilitation after THA and ensure
that such outcomes become expected standards for funding studies and publishing their
results.

Other opportunities for improvement

Many of the studies included in this review had moderate to high risk of bias, in part due to
lack of blinding of clinicians, patients, and outcome assessors. While it is difficult to blind
clinicians and patients in rehabilitation studies, threats to bias may be minimized by blinding
outcome assessors (for performance-based outcomes) and analysts (for all outcomes). In this
Author Manuscript

review, only five studies were rated as low risk of bias for the domain of blinding outcome
assessors; five studies were rated as unclear risk of bias and three were rated as high risk
of bias. Thus, in addition to considering what core outcomes should be included in studies,
stakeholders should consider standards for how such core outcomes should be collected and
analyzed to minimize threats to bias.

Finally, studies had relatively small sample sizes, which may have resulted in underpowered
studies and, thus, misleadingly statistically nonsignificant findings. This is especially
important for comparison of two or more active interventions like two rehabilitation
programs as large samples are needed to demonstrate an advantage of a likely smaller
relative effect size. Jette et al. make similar calls for action in rehabilitation research for
TKA and noted that the challenge of sample size may necessitate the need for coordinated
multisite programs of research to facilitate larger sample sizes.42
Author Manuscript

Strengths and limitations of the review

We used a comprehensive strategy to ensure our literature search was complete and did
not systematically miss studies. While we restricted the evidence base to the past 15 years,
this decision was made with input from our diverse stakeholders and based on changing
orthopedic surgery practices and rehabilitation protocols in the early 2000s. An important
strength of this review was the extensive independent coding of the rehabilitation content

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 11

and delivery factors using a novel and comprehensive taxonomy. These codes are a crucial
Author Manuscript

first step in articulating the challenge of understanding rehabilitation evidence to date and
setting a benchmark for intervention design and reporting to come. While the taxonomy was
not able to be used to its full advantage in meta-regression models to isolate the unique
effect of specific components, such analyses could be feasible in the future if sufficient
number of studies (or real world data) were available on specific outcomes. A limitation
of our synthesis is that we did not capture timing, intensity, or dosage, in any form (e.g.,
challenge of the exercise, duration of the session in minutes, duration of the program in
days/weeks, number of sessions). Finally, in terms of applicability there was wide variation
in the sex ratios (27 to 100 percent of participants being female) and mean age (54 to
71 years) across studies and most studies did not report whether patients had undergone
previous contralateral replacement surgery (although of those that that did, proportions
were less than 25%). As such, the conclusions in this SR are likely most applicable to
Author Manuscript

middle-to-older-aged adults in high-income countries who are receiving their first total THA
for osteoarthritis.

In conclusion, many questions remain inadequately answered regarding the best

rehabilitation management of patients with osteoarthritis after THA. The lack of clarity
regarding comparative effectiveness and harms is due to the diversity of interventions, often
inadequate intervention description, and wide variety of outcomes reported across studies.
To advance rehabilitation for adults after THA, there is an urgent need for well-conducted
studies that address both effectiveness (and harms) of interventions and heterogeneity of
treatment effects using standardized intervention terminology and core outcome sets.

Supplementary Material
Author Manuscript

Refer to Web version on PubMed Central for supplementary material.

ACKNOWLEDGEMENTS
The authors gratefully acknowledge the contribution of David W. Niebuhr, M.D., M.P.H., M.Sc., our AHRQ
Task Order Officer; members of the Key Informant and Technical Expert Panels, reviewers of our overall review
(all listed in the AHRQ full report); Jennifer Racine-Avila, M.B.A. for early feedback and development of the
protocol; Carol A. Oatis, P.T., Ph.D., for sharing her taxonomy and providing important conceptual feedback in its
application; Shivani Mehta B.A., a research associate who played a role in conduct of the overall review

FUNDING SOURCE
This project was funded under Contract No. HHSA 75Q80120D00001 from the Agency for Healthcare Research
and Quality (AHRQ), U.S. Department of Health and Human Services (DHHS). The authors of this paper are
responsible for its content. Statements in the report should not be construed as endorsement by AHRQ or DHHS.
AHRQ retains a license to display, reproduce, and distribute the data and the report from which this manuscript was
Author Manuscript

derived under the terms of the agency’s contract with the author.

References
1. Di Monaco M, Vallero F, Tappero R, et al. Rehabilitation after total hip arthroplasty: a systematic
review of controlled trials on physical exercise programs. Eur J Phys Rehabil Med. 2009
Sep;45(3):303–17. [PubMed: 19238130]
2. Di Monaco M, Castiglioni C. Which type of exercise therapy is effective after hip arthroplasty?
A systematic review of randomized controlled trials. Eur J Phys Rehabil Med. 2013 Dec;49(6):893–
907, quiz 21–3. [PubMed: 24172644]

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 12

3. Konnyu KJ, Thoma LM, Bhuma MR, Cao W, Adam GP, Mehta S, Aaron RK, Racine-Avila
J, Panagiotou OA, Pinto D, Balk EM. Prehabilitation and Rehabilitation for Major Joint
Author Manuscript

Replacement. Comparative Effectiveness Review No. 248. (Prepared by the Brown Evidence-based
Practice Center under Contract No. 75Q80120D00001.) AHRQ Publication No. 21(22)-EHC033.
Rockville, MD: Agency for Healthcare Research and Quality; November 2021. DOI: 10.23970/
AHRQEPCCER248. Posted final reports are located on the Effective Health Care Program search
page.
4. Konnyu KJ, Thoma LM, Cao W, Aaron RK, Panagiotou OA, Bhuma MR, Adam GP, Pinto D, Balk
EM. Prehabilitation for total knee and total hip replacement: A systematic review.
5. Konnyu KJ, Thoma LM, Cao W, Aaron RK, Panagiotou OA, Bhuma MR, Adam GP, Balk EM,
Pinto D. Rehabilitation for total knee replacement: A systematic review.
6. Quality AfHRa. Methods Guide for Effectiveness and Comparative Effectiveness Reviews. January
2014. Chapters available at: www.effectivehealthcare.ahrq.gov.
7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews
and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.
Bmj. 2009 Jul 21;339:b2700. doi: 10.1136/bmj.b2700. [PubMed: 19622552]
Author Manuscript

8. Zhu S, Qian W, Jiang C, et al. Enhanced recovery after surgery for hip and knee arthroplasty:
a systematic review and meta-analysis. Postgrad Med J. 2017 Dec;93(1106):736–42. doi: 10.1136/
postgradmedj-2017-134991. [PubMed: 28751437]
9. International Classification of Functioning, Disability and Health (ICF). Geneva, Switzerland: World
Health Organization; 2018. https://www.who.int/classifications/icf/en/. Accessed on April 30, 2020.
10. Franklin PD. Defining Components of Physical Therapy Achieving Maximum Function after TKR.
Northwestern University at Chicago, Chicago, IL, United States; 2019.
11. Oatis CA, Johnson JK, DeWan T, et al. Characteristics of Usual Physical Therapy Post-Total Knee
Replacement and Their Associations With Functional Outcomes. Arthritis Care Res (Hoboken).
2019 Sep;71(9):1171–7. doi: 10.1002/acr.23761. [PubMed: 30281207]
12. Higgins JP, Altman DG, Gøtzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk
of bias in randomised trials. Bmj. 2011 Oct 18;343:d5928. doi: 10.1136/bmj.d5928. [PubMed:
22008217]
13. Sterne JA, Hernán MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in
non-randomised studies of interventions. Bmj. 2016 Oct 12;355:i4919. doi: 10.1136/bmj.i4919.
Author Manuscript

[PubMed: 27733354]
14. National Heart, Lung, and Blood Institute. Study Quality Assessment Tools.; 2019. https://
www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools. Accessed on January 23, 2020.
15. Berkman ND, Lohr KN, Ansari M, et al. AHRQ Methods for Effective Health Care Grading
the Strength of a Body of Evidence When Assessing Health Care Interventions for the Effective
Health Care Program of the Agency for Healthcare Research and Quality: An Update. Methods
Guide for Effectiveness and Comparative Effectiveness Reviews. Rockville (MD): Agency for
Healthcare Research and Quality (US); 2008.
16. Berkman ND, Lohr KN, Ansari MT, et al. Grading the strength of a body of evidence when
assessing health care interventions: an EPC update. J Clin Epidemiol 2015 Nov;68(11):1312–24.
doi: 10.1016/j.jclinepi.2014.11.023. [PubMed: 25721570]
17. Austin MS, Urbani BT, Fleischman AN, et al. Formal Physical Therapy After Total Hip
Arthroplasty Is Not Required: A Randomized Controlled Trial. J Bone Joint Surg Am. 2017 Apr
19;99(8):648–55. doi: 10.2106/jbjs.16.00674. . [PubMed: 28419032]
Author Manuscript

18. Beck H, Beyer F, Gering F, et al. Sports Therapy Interventions Following Total Hip Replacement:
A Randomized Controlled Trial. Deutsches Aerzteblatt International. 2019;116(1/2):1–8. doi:
10.3238/arztebl.2019.0001. Language: English. Entry Date: In Process. Revision Date: 20190223.
Publication Type: Article. Journal Subset: Biomedical.
19. Coulter C, Perriman DM, Neeman TM, et al. Supervised or Unsupervised Rehabilitation After
Total Hip Replacement Provides Similar Improvements for Patients: A Randomized Controlled
Trial. Arch Phys Med Rehabil. 2017 Nov;98(11):2253–64. doi: 10.1016/j.apmr.2017.03.032.
[PubMed: 28506775]

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 13

20. Giaquinto S, Ciotola E, Dall’armi V, et al. Hydrotherapy after total hip arthroplasty: a follow-
up study. Arch Gerontol Geriatr 2010 Jan–Feb;50(1):92–5. doi: 10.1016/j.archger.2009.02.005.
Author Manuscript

[PubMed: 19282040]
21. Heiberg KE, Bruun-Olsen V, Ekeland A, et al. Effect of a walking skill training program in patients
who have undergone total hip arthroplasty: Followup one year after surgery. Arthritis Care Res
(Hoboken). 2012 Mar;64(3):415–23. doi: 10.1002/acr.20681. [PubMed: 22170790]
22. Heiberg KE, Figved W. Exercise, recovery of physical functioning, and prediction of physical
activity after total hip arthroplasty. 5-year follow-up of a rct. Annals of the rheumatic diseases.
2015;74:1318–9. doi: 10.1136/annrheumdis-2015-eular.3078.
23. Heiberg KE, Figved W. Physical Functioning and Prediction of Physical Activity After Total
Hip Arthroplasty: Five-Year Followup of a Randomized Controlled Trial. Arthritis Care Res
(Hoboken). 2016 Apr;68(4):454–62. doi: 10.1002/acr.22679. [PubMed: 26239078]
24. Liebs TR, Herzberg W, Rüther W, et al. Ergometer cycling after hip or knee replacement surgery:
a randomized controlled trial. J Bone Joint Surg Am. 2010 Apr;92(4):814–22. doi: 10.2106/
jbjs.H.01359. [PubMed: 20360503]
25. Liebs TR, Herzberg W, Rüther W, et al. Multicenter randomized controlled trial comparing early
Author Manuscript

versus late aquatic therapy after total hip or knee arthroplasty. Arch Phys Med Rehabil. 2012
Feb;93(2):192–9. doi: 10.1016/j.apmr.2011.09.011. [PubMed: 22196125]
26. Łyp M, Kaczor R, Cabak A, et al. A Water Rehabilitation Program in Patients with Hip
Osteoarthritis Before and After Total Hip Replacement. Med Sci Monit. 2016 Jul 25;22:2635–42.
doi: 10.12659/msm.896203. [PubMed: 27455419]
27. Mikkelsen LR, Mechlenburg I, Søballe K, et al. Effect of early supervised progressive
resistance training compared to unsupervised home-based exercise after fast-track total hip
replacement applied to patients with preoperative functional limitations. A single-blinded
randomised controlled trial. Osteoarthritis Cartilage. 2014 Dec;22(12):2051–8. doi: 10.1016/
j.joca.2014.09.025. [PubMed: 25305374]
28. Monticone M, Ambrosini E, Rocca B, et al. Task-oriented exercises and early full weight-bearing
contribute to improving disability after total hip replacement: a randomized controlled trial. Clin
Rehabil 2014 Jul;28(7):658–68. doi: 10.1177/0269215513519342. [PubMed: 24459172]
29. Naylor JM, Hart A, Mittal R, et al. The effectiveness of inpatient rehabilitation after uncomplicated
total hip arthroplasty: a propensity score matched cohort. BMC Musculoskelet Disord. 2018 Jul
Author Manuscript

18;19(1):236. doi: 10.1186/s12891-018-2134-3. [PubMed: 30021552]

30. Nelson M, Bourke M, Crossley K, et al. Telerehabilitation Versus Traditional Care Following
Total Hip Replacement: A Randomized Controlled Trial Protocol. JMIR Res Protoc. 2017 Mar
2;6(3):e34. doi: 10.2196/resprot.7083. [PubMed: 28254734]
31. Nelson M, Bourke M, Crossley K, et al. Telerehabilitation is non-inferior to usual care
following total hip replacement - a randomized controlled non-inferiority trial. Physiotherapy.
2020 Jun;107:19–27. doi: 10.1016/j.physio.2019.06.006. [PubMed: 32026820]
32. Rao BM, Cieslewicz TJ, Sochacki KR, et al. Worse Preoperative Pain and Higher Activity Levels
Predict Patient Choice of Formal Physical Therapy After Primary Anterior Total Hip Arthroplasty.
J Arthroplasty. 2021 Mar 16. doi: 10.1016/j.arth.2021.03.027.
33. Smith TO, Mann CJ, Clark A, et al. Bed exercises following total hip replacement: 1 year
follow-up of a single-blinded randomised controlled trial. Hip Int 2009 Jul–Sep;19(3):268–73. doi:
10.1177/112070000901900314. [PubMed: 19876883]
34. Smith TO, Mann CJV, Clark A, et al. Bed exercises following total hip replacement: a randomised
Author Manuscript

controlled trial. Physiotherapy. 2008;94(4):286–91. Language: English. Entry Date: 20090227.

Revision Date: 20150819. Publication Type: Journal Article.
35. Winther SB, Foss OA, Husby OS, et al. A randomized controlled trial on maximal strength
training in 60 patients undergoing total hip arthroplasty. Acta Orthop. 2018 Jun;89(3):295–301.
doi: 10.1080/17453674.2018.1441362. [PubMed: 29493347]
36. Winther SB, Foss OA, Klaksvik J, et al. Increased Muscle Strength Limits Postural Sway During
Daily Living Activities in Total Hip Arthroplasty Patients. Am J Phys Med Rehabil. 2020
Jul;99(7):608–12. doi: 10.1097/phm.0000000000001382. [PubMed: 31977324]

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 14

37. Projected volume of primary and revision total joint arthroplasty in the United States, 2030–2060.
American Academy of Orthopaedic Surgeons 2018 Annual Meeting (New Orleans, LA); 2018.
Author Manuscript

38. Rubin DB. Meta-Analysis: Literature synthesis or effect-size surface estimation? J. Educ. Stat.
[electronic article]. 1992;17(4):363–374. (http://www.jstor.org/stable/1165129?origin=crossref).
(Accessed August 21, 2017)
39. Hoffmann TC, Glasziou PP, Boutron I, et al. Better reporting of interventions: template
for intervention description and replication (TIDieR) checklist and guide. Bmj. 2014 Mar
7;348:g1687. doi: 10.1136/bmj.g1687. [PubMed: 24609605]
40. Westby MD, Brittain A, Backman CL. Expert consensus on best practices for post-acute
rehabilitation after total hip and knee arthroplasty: a Canada and United States Delphi study.
Arthritis Care Res (Hoboken). 2014 Mar;66(3):411–23. doi: 10.1002/acr.22164. [PubMed:
24023047]
41. Dobson F, Hinman RS, Roos EM, et al. OARSI recommended performance-based tests to assess
physical function in people diagnosed with hip or knee osteoarthritis. Osteoarthritis Cartilage.
2013 Aug;21(8):1042–52. doi: 10.1016/j.joca.2013.05.002. [PubMed: 23680877]
42. Jette DU, Hunter SJ, Jette AM. Overcoming Research Challenges to Improve Clinical Practice
Author Manuscript

Guideline Development. Phys Ther. 2020 Oct 30;100(11):1889–90. doi: 10.1093/ptj/pzaa153.

[PubMed: 33125083]
Author Manuscript
Author Manuscript

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 15
Author Manuscript
Author Manuscript
Author Manuscript

Figure 1. Overview of studies of acute and post-acute rehabilitation interventions versus various
controls for total hip arthroplasty
Figure presents categorization of studies that evaluated acute and post-acute rehabilitation
programs for THA. The first column lists a novel (more intensive) acute rehabilitation
program compared to a different program (first group hypothesized to be better); the second
column lists studies with comparatively similar acute rehabilitation programs in both arms
that were delivered with different timing or intensity (first group hypothesized to be better);
the third column lists a novel (more intensive) post-acute rehabilitation program compared
to a different program (first group hypothesized to be better); the fourth column lists studies
with comparatively similar rehabilitation programs delivered in different settings or by
Author Manuscript

different personnel
Abbreviations: Adj = adjunctive, A = aerobic exercise, B= balance-motor-learning-agility
exercise, E = patient education, F = flexibility exercise, S = strengthening exercise, T =
task-specific training, THA = total hip arthroplasty.
* Intervention included progression which was deemed appropriate

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 16
Author Manuscript
Author Manuscript

Figure 2. Goal components strength, aerobic, and flexibility and their specific exercise
components for rehabilitation interventions versus various controls for total hip arthroplasty
See Figure 3 for goal components Balance-motor-learning-agility, task specific training,
patient education, and adjunctive modalities. The color is added for visual display and does
not provide unique information.
Group I: kinesiotherapy, low-frequency magnetic field and water exercises; Group II:
undergoing kinesiotherapy and low-frequency magnetic field, without water exercises;
Group V: awaiting rehabilitation
Abbreviations: PRT = progressive resistance training, PT = physical therapy, rehab =
Author Manuscript

rehabilitation; ROM = range of motion, ST = strength training

A Task-oriented exercises; encouragement to abandon walking aids.
B Open kinetic chain exercises
Author Manuscript

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Konnyu et al. Page 17
Author Manuscript
Author Manuscript

Figure 3. Goal components balance-motor-learning-agility, task specific training, patient

education, and adjunctive modalities and their specific exercise components for rehabilitation
interventions versus various controls for total hip arthroplasty
See Figure 2 for goal components strength, aerobic, and flexibility. The color is added for
visual display and does not provide unique information. The color is added for visual display
and does not provide unique information.Group I: kinesiotherapy, low-frequency magnetic
field and water exercises; Group II: undergoing kinesiotherapy and low-frequency magnetic
Author Manuscript

field, without water exercises; Group V: awaiting rehabilitation

Abbreviations: AI = acute inpatient, CAM = complementary and alternative therapies, ed
= education, exercises, I = in-person; G = gym/other community center; H = home; O =
outpatient physiotherapy center; OIF = other inpatient facility; PRT PT = physical therapy,
rehab = rehabilitation; SG = self-guided; ST = strength training.
A Task-oriented exercises; encouragement to abandon walking aids
B Open kinetic chain exercises
C Remote via telephone
Author Manuscript

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Table 1.

Evidence profile: Rehabilitation versus various controls for total hip arthroplasty

Outcome Category Outcome N Studies RoB Consistency Precision Directness Intervention SoE Conclusions
(Participants) Replication
Konnyu et al.

Body structure and Pain 11 (1297) Moderate Consistent Precise Direct All unique Low Similar pain
function
ROM 3 (178) Moderate Inconsistent Precise Direct All unique Insufficient No conclusion

Strength 5 (370) Moderate Consistent Precise Direct All unique Low Similar strength

Activity and participation ADLs 12 (1247) Moderate Consistent Precise Direct All unique Low Similar ADL

Other outcomes Satisfaction with care 2 (372) Moderate Consistent Precise Direct All unique Insufficient No conclusion

QoL 4 (335) Moderate Consistent Precise Direct All unique Low Similar QoL

Healthcare utilization Need for postoperative 0 NA NA NA NA NA Insufficient No evidence

procedures

Harms Harms from 4 (534) Moderate Consistent Precise Direct All unique Low Similar harms
prehabilitation

Abbreviations: ADLs = activities of daily living, QoL = health-related quality of life, NA = not applicable, QoL = quality of life, RoB = risk of bias, ROM = range of motion, SoE = strength of evidence.

Am J Phys Med Rehabil. Author manuscript; available in PMC 2024 January 01.
Page 18