Journal of Physiotherapy 65 (2019) 136–143

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j p hy s

Research

Incorporating hip abductor strengthening exercises into a rehabilitation program

did not improve outcomes in people following total knee arthroplasty:
a randomised trial
Margaret B Schache a,b, Jodie A McClelland a, Kate E Webster a
a
School of Allied Health, La Trobe University, Melbourne; b Physiotherapy Department, Donvale Rehabilitation Hospital, Melbourne, Australia

K E Y W O R D S A B S T R A C T

Randomised controlled trial Question: In adults following primary total knee arthroplasty, does the incorporation of hip abductor
Arthroplasty strengthening exercises into a 6-week rehabilitation program improve muscle strength, functional
Knee replacement
performance and patient-reported outcomes at the end of rehabilitation and at 26 weeks? Design: Rando-
Muscle strength
mised controlled trial with concealed allocation, blinded assessors and intention-to-treat analysis.
Exercise therapy
Treatment outcome Participants: One hundred and five adults admitted to an inpatient rehabilitation facility immediately
following total knee arthroplasty. Intervention: Participants in both groups attended 12 days of inpatient
physiotherapy followed by 6 weeks of outpatient physiotherapy, which aimed to improve knee range of
movement, strength and mobility. The experimental group completed a standard rehabilitation protocol with
the addition of hip abductor strengthening. The control group completed the same standard rehabilitation
protocol, with the addition of 15 minutes of general functional exercises. Outcome measures: Primary
outcomes were the Knee Injury and Osteoarthritis Outcome Score (KOOS) and isometric hip abductor muscle
strength normalised to body mass index. Secondary outcome measures included the stair climb test, 6-
minute walk test, Timed Up and Go test, 40-m fast-paced walk test, 30-second chair stand test, step test,
isometric quadriceps muscle strength, Lower Extremity Functional Scale, and Short Form-12. Results: The
experimental intervention did not result in significantly greater improvements in hip strength, KOOS or any
of the secondary outcome measures than the control intervention at 6 weeks or 26 weeks. Conclusion:
Similar improvements in muscle strength, functional performance and patient-reported outcomes were
observed whether specific hip-strengthening exercises were incorporated or general functional exercises
were continued instead as part of a postoperative rehabilitation program for participants after total knee
arthroplasty. Registration: ANZCTR 12615000863538. [Schache MB, McClelland JA, Webster KE (2019)
Incorporating hip abductor strengthening exercises into a rehabilitation program did not improve
outcomes in people following total knee arthroplasty: a randomised trial. Journal of Physiotherapy
65:136–143]
© 2019 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction and hip abductor muscles can persist well beyond 3 years after total
knee arthroplasty.6 Patients report greater difficulty in completing
Total knee arthroplasty remains the most effective treatment for daily tasks and perform worse than normal on a range of functional
end-stage knee osteoarthritis.1 The number of procedures performed outcomes.5–7 It is therefore justified that postoperative rehabilitation
per year has increased, with approximately 62 000 primary total knee programs address the persistent muscle weakness and functional
arthroplasties performed on Australian adults in 2017, which is an deficits for this growing population.
increase of 4.3% compared with 2016.2 Total knee arthroplasty is Restoring quadriceps strength is a common goal of most post-
successful in relieving the symptoms of osteoarthritis and restoring operative rehabilitation programs; however, functional deficits
walking speed and stair climbing ability to preoperative levels by 6 remain even after completion of these programs. Increasing hip
months after surgery;3,4 however, it is important to seek improve- strength may improve outcomes after total knee arthroplasty, and
ment beyond preoperative levels towards normal function. there is some evidence that hip abductor strength influences physical
Despite symptomatic improvement following total knee arthro- function following total knee arthroplasty.8–10 Hip abductor strength
plasty, some patients experience persistent muscle weakness, has also been shown to be more strongly associated with functional
ongoing functional difficulties, and pain when compared with healthy performance than quadriceps strength.10 Furthermore, in a small pilot
age-matched controls.1,5–7 Muscle weakness affecting the quadriceps study, the addition of hip abductor strength exercises to rehabilitation

https://doi.org/10.1016/j.jphys.2019.05.008
1836-9553/© 2019 Australian Physiotherapy Association. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
 Research 137

Patients admitted for rehabilitation and

screened for eligibility (n = 315)

Excluded (n = 210)
• not returning as an outpatient (n = 68)
• comorbidities, unwell, hip pain (n = 67)
• revision TKA (n = 32)
• unicompartmental TKA (n = 19)
• did not speak English (n = 10)
• declined to participate (n = 7)
• bilateral TKA (n = 5)
• previously enrolled in this study with
contralateral TKA (n = 2)

Measured pain, range of movement, strength, chair stand test, stair climb test, 40-m fast
paced walk test, Timed Up and Go test, step taps, 6-minute walktest, and the KOOS, LEFS
and SF-12 questionnaires
Week 0
Randomised (n = 105)
(n = 54) (n = 51)

Experimental Group Control Group

Lost to follow-up • 45-minute exercise therapy • 45-minute exercise
• withdrew due to sessions with progressive therapy sessions
transport hip abductor strengthening • manual therapy
difficulties (n = 1) • manual therapy • eleven sessions/week
• eleven sessions/week until until discharge, then 2
discharge, then 2 outpatient
outpatient sessions/week sessions/week
• 6 to 8 weeks • 6 to 8 weeks

Measured pain, range of movement, strength, chair stand test, stair climb test, 40-m fast
Week 6 paced walk test, Timed Up and Go test, step taps, 6-minute walk test, and the KOOS, LEFS
and SF-12 questionnaires
(n = 53) (n = 51)

Lost to follow-up
• contralateral Lost to follow-up
TKA (n = 1) • contralateral TKA
• cardiac valve (n = 1)
repair (n= 1) • working (n = 1)
• too busy (n = 1) • family illness (n = 1)
• working (n = 1)
• declined (n = 1)

Measured pain, range of movement, strength, chair stand test, stair climb test, 40-m fast
paced walk test, Timed Up and Go test, step taps, 6-minute walk test, and the KOOS, LEFS
Week 26 and SF-12 questionnaires

(n = 48) (n = 48)

Figure 1. Design and flow of participants through the trial. TKA = total knee arthroplasty, KOOS = Knee Injury and Osteoarthritis Outcome Score, LEFS = Lower Extremity Functional
Scale, SF-12 = Short Form-12 version 1.

after total knee arthroplasty achieved significant improvement in Therefore, the research question for this randomised controlled
functional outcomes after total knee arthroplasty.11 trial was:
The addition of targeted hip abductor strengthening in other
populations with knee pain, such as those with knee osteoarthritis In adults following primary total knee arthroplasty, does the
and patellofemoral pain, has been shown to reduce pain and improve incorporation of hip abductor strengthening exercises into a 6-
physical function and quality of life.12–16 These effects are greater week rehabilitation program improve muscle strength, func-
when compared with quadriceps strengthening alone.13–15 It is tional performance and patient-reported outcomes at the end of
therefore reasonable to hypothesise that the addition of hip rehabilitation and at 26 weeks?
strengthening exercises to total knee arthroplasty rehabilitation
programs would also result in greater improvements in patient- Method
reported and functional outcomes. Larger randomised controlled
trials are required to determine the benefit of incorporating targeted Design
hip abductor strengthening exercises into postoperative total knee
arthroplasty rehabilitation, which is the primary aim of the current A randomised controlled trial was conducted with concealed
trial. allocation, blinded assessors, and intention-to-treat analysis.
138 Schache et al: Hip abductor strengthening following knee arthroplasty

Table 1 their treating physiotherapist. The experience of the physiotherapists

Characteristics of participants at baseline. who provided the intervention ranged from 2 to 26 years.
Characteristic Exp Con Participants were eligible for inclusion if they were at least 50
(n = 54) (n = 51) years of age and had undergone a primary unilateral total knee
Age (yrs), mean (SD) 70 (7) 69 (7) arthroplasty for end-stage knee osteoarthritis in the previous 2
Gender, n female (%) 39 (72) 30 (58) weeks. Participants were excluded if they had: unstable medical
Height (cm), mean (SD) 166 (9) 167 (10) conditions, such as uncontrolled cardiovascular disease or uncon-
Mass (kg), mean (SD) 84 (18) 87 (17)
Body mass index (kg/m2), mean (SD) 30 (6) 31 (6)
trolled diabetes; a history of ipsilateral hip replacement, ipsilateral
Time since surgery (d), mean (SD) 8 (2) 7 (2) hip osteoarthritis or lateral hip pain; or neurological or any other
Side of surgery, L:R 27:27 33:18 conditions affecting strength or function of the lower limbs.
Prosthesis and patella resurfacing, n (%)
cruciate retaining 1 patella resurfaced 17 (31) 13 (25)
cruciate retaining – patella resurfaced 18 (33) 19 (37) Intervention
posterior stabilised 1 patella resurfaced 16 (30) 19 (37)
posterior stabilised – patella resurfaced 3 (6) 0 (0) Standard rehabilitation
Previous joint replacements, n TKA:THA 11:3 23:6
Participants attended either two 45-minute sessions of physio-
Con = control group, exp = experimental group, THA = total hip arthroplasty, therapy or one 45-minute session of physiotherapy and one 45-
TKA = total knee arthroplasty.
minute session of hydrotherapy each week day. All participants also
attended one 45-minute session of either physiotherapy or hydro-
Participants in the experimental group were allocated standard therapy on the weekend. All participants were given an exercise
rehabilitation plus 15 minutes of additional hip strengthening exer- program to complete whilst admitted to the rehabilitation hospital.
cises, whereas participants in the control group were allocated These exercises included exercises targeted at improving quadriceps,
standard rehabilitation plus 15 minutes of additional general func- hamstring and calf strength, increasing knee range of movement, and
tional exercise; these interventions are described further below and improving walking and stair-climbing ability. These exercises have
in the published protocol of this study.17 Although participants were been described in detail previously.17 Manual therapy, including joint
aware of their allocated intervention, they were not informed of what mobilisation and massage, was also used. Participants were dis-
the other group had been allocated. charged from the inpatient rehabilitation hospital facility after
Participants were randomised by a computer-generated list of approximately 12 days when they could mobilise independently and
random numbers concealed in sealed, opaque envelopes that were negotiate stairs safely.
prepared by a person not involved in delivering the treatment or Once discharged from the inpatient facility, participants returned
testing. The intervention commenced on admission to inpatient for weekly outpatient physiotherapy rehabilitation for 6 to 8 weeks
rehabilitation. A blinded assessor measured all outcomes immedi- for one 45-minute session of physiotherapy and either one 45-minute
ately before the intervention (Week 0), at the end of the intervention session of hydrotherapy (if indicated) or land-based exercise (if hy-
period (Week 6) and at follow-up 20 weeks later (Week 26). drotherapy not indicated). After 6 to 8 weeks of outpatient physio-
therapy, participants were discharged and encouraged to continue
their home exercise program.
Participants, therapists, centres
Experimental group
The trial was conducted in a private rehabilitation hospital. All Participants in the experimental group received standard reha-
patients admitted to inpatient rehabilitation were screened for bilitation plus additional exercises designed specifically to
eligibility to participate in the trial. Participants were recruited by target strengthening of the hip abductor muscles.12,16,18–21 The

Table 2
Intention-to-treat analysis of mean (SD) of groups, mean (SD) within-group difference, and mean (95% CI) between-group difference for pain and objectively measured outcomes.

Outcome Groups Within-group difference Between-group difference

Week 0 Week 6 Week 26 Week 6 Week 26 Week 6 Week 26

minus minus minus minus
Week 0 Week 0 Week 0 Week 0

Exp Con Exp Con Exp Con Exp Con Exp Con Exp minus Exp minus Con
(n = 54) (n = 51) (n = 52) (n = 50) (n = 48) (n = 48) Con

Pain 3 4 1 2 0 1 22 22 23 23 0 0
(0 to 10) (2) (2) (2) (2) (1) (1) (2) (2) (2) (2) (21 to 1) (21 to 1)
Flexion ROM 83 81 109 107 121 118 27 26 38 37 1 1
(deg) (10) (12) (8) (10) (6) (9) (12) (10) (12) (12) (24 to 5) (24 to 6)
Extension ROM 25 25 22 22 0 0 3 3 4 5 0 21
(deg) (6) (5) (3) (3) (1) (2) (5) (4) (5) (5) (22 to 2) (23 to 2)
Hip strength 1.9 2.1 3.5 3.6 4.0 4.1 1.6 1.5 2.0 2.0 0.0 0.0
(N/kg/m2) (0.9) (1.1) (1.4) (1.8) (1.6) (1.7) (1.1) (1.2) (1.1) (1.1) (20.4 to 0.5) (20.5 to 0.5)
Quadriceps strength 1.5 1.4 4.9 5.3 6.6 6.9 3.4 3.9 5.1 5.5 20.5 20.4
(N/kg/m2) (0.9) (1.0) (2.3) (2.9) (3.0) (3.4) (1.9) (2.6) (2.7) (3.2) (21.4 to 0.4) (21.6 to 0.8)
Chair stand test 7 8 13 14 15 15 6 6 7 8 0 0
(n in 30 s) (3) (3) (4) (5) (4) (5) (2) (3) (3) (4) (21 to 1) (22 to 1)
Stair climb test 25 22 8 8 7 7 216 214 217 215 22 22
(s) (10) (8) (3) (3) (2) (2) (9) (6) (8) (7) (25 to 1) (25 to 1)
40-m fast-paced walk 95 85 34 35 29 29 261 252 255 255 29 0
(s) (60) (41) (10) (12) (9) (10) (57) (37) (39) (36) (228 to 10) (215 to 16)
Timed Up and Go 29 25 9 10 8 8 220 216 219 217 24 22
(s) (14) (11) (3) (4) (2) (3) (12) (9) (11) (9) (28 to 0) (26 to 2)
Step taps 5 5 16 17 17 18 11 12 12 13 21 21
(n) (5) (6) (4) (5) (4) (5) (4) (6) (5) (6) (23 to 1) (23 to 1)
6-minute walk test 178 194 411 420 474 477 233 226 285 281 7 4
(m) (75) (101) (97) (126) (106) (128) (78) (88) (87) (95) (226 to 40) (233 to 41)

Con = control group, Exp = experimental group, ROM = range of movement.

 Research 139

16 muscle strength.23 The KOOS is a self-reported questionnaire with 42

items in five separately analysed subscales of pain, other symptoms,
Hip abductor strength (N.kg‫־‬¹.m‫־‬²)

Experimental function in daily living, function in sport and recreation, and knee-
12 Control related quality of life.22 (The sport and recreation subscale was not
considered in this study.) Isometric hip abductor muscle strength was
measured in supine with a handheld dynamometer, recorded in New-
8
tons and normalised to body mass index.23 The measurement of iso-
metric hip abductor strength has previously been described in detail.17
Secondary outcomes: Secondary outcome measures included the stair
4
climb test (ie, time taken to climb four steps in seconds),24,25 the 6-
minute walk test (measured in metres),24,26 Timed Up and Go
0 (measured in seconds),27 the 40-m fast-paced walk test (measured in
0 6 26 seconds),24,28 30-second chair stand test (measured as number of
Time (weeks) stands),24,29 step taps (measured as number of taps),30 isometric
quadriceps muscle strength (measured in Newtons and normalised to
Figure 2. Hip abductor strength by time. Symbols show individual participants’ out-
comes. Lines join group means at baseline and at Weeks 6 and 26. Experimental and body mass index),17,31,32 passive knee range of movement (measured
control group data have been offset slightly for clarity. in degrees),17,31 the Lower Extremity Functional Scale,33 and Short
Form-12 version 1.34

participants initially performed non-weight-bearing, antigravity, hip- Data analysis

strengthening exercises (side lying hip abduction, prone hip exten- The sample size calculation was based around the primary
sion and standing hip abduction). The exercises were progressed to outcome measure: KOOS.22 We nominated a between-group differ-
weight-bearing, gravity-resisted and TheraBand-resisted exercises ence of 10 as the smallest worthwhile effect,35 and 17 as the antici-
(sideways walking, hip hitching and hip abduction whilst standing on pated standard deviation.36 To ensure statistical power of 0.8 at a
the operated leg) based on pre-defined criteria. These exercises have significance level of a = 0.05, a minimum sample of 94 participants
previously been described in detail.17 (47 per group) would be needed. To allow for 10% loss to follow-up at
6 months,37–39 the aim was to recruit a sample of 105 participants.
Control group For all primary and secondary outcomes, between-group com-
Participants in the control group received standard rehabilitation parisons at 6 weeks and at 26 weeks were reported as mean differ-
plus an additional 15 minutes of general functional exercises at each ence with 95% confidence intervals. Statistical analyses were
physiotherapy session to serve as a time control for the additional hip conducted according to the intention-to-treat principle, with all
strengthening exercises prescribed for the experimental group. These available data analysed according to group allocation.
exercises were chosen to replicate functional activities and included In addition, a per-protocol analysis was conducted, in which par-
sit-to-stand, marching and walking around a pre-measured circuit. If ticipants were considered to have adhered to the program if they
the participant received hydrotherapy, the same additional exercises received the allocated intervention, attended one exercise session per
were performed in the water. day as an inpatient, and attended six consecutive weeks as an
outpatient. Apart from analysing only the adherent participants, the
Outcome measures per-protocol analysis used the same statistical methods as the
Demographic information recorded at baseline included name, intention-to-treat analysis.
age, gender, height, weight and body mass index. The type of pros-
thesis used (including patellar resurfacing), left or right total knee Results
arthroplasty, previous joint replacements and co-interventions (eg,
hydrotherapy) were recorded. Length of hospital stay, number of Compliance with the study protocol
treatment sessions, and adverse effects were also recorded.
Primary outcomes: The primary outcomes were the Knee Injury and There were no important deviations from the study protocol. The
Osteoarthritis Outcome Score (KOOS)22 and isometric hip abductor interventions were delivered as allocated, with hydrotherapy sessions

Table 3
Intention-to-treat analysis of mean (SD) of groups, mean (SD) within-group difference, and mean (95% CI) between-group difference for questionnaires.

Outcome Groups Within-group difference Between-group difference

Week 0 Week 6 Week 26 Week 6 Week 26 Week 6 minus Week 26 minus

minus minus Week 0 Week 0
Week 0 Week 0

Exp Con Exp Con Exp Con Exp Con Exp Con Exp minus Exp minus
(n = 54) (n = 51) (n = 52) (n = 49) (n = 47) (n = 48) Con Con

KOOS (0 to 100)
symptoms 45 46 66 63 82 79 21 18 36 34 3 2
(13) (16) (16) (17) (13) (14) (15) (19) (14) (18) (24 to 10) (24 to 9)
pain 46 47 73 71 87 71 27 23 40 39 4 1
(14) (19) (16) (15) (11) (15) (18) (17) (13) (18) (23 to 10) (25 to 8)
ADL 54 55 82 78 90 88 27 23 35 33 5 3
(18) (22) (15) (15) (11) (13) (20) (19) (15) (22) (23 to 12) (25 to 11)
quality of life 29 28 59 55 73 70 30 27 45 41 3 3
(17) (21) (18) (22) (19) (21) (23) (23) (22) (24) (26 to 12) (26 to 13)
LEFS 22 21 42 43 53 54 20 22 30 32 22 22
(0 to 80) (12) (13) (11) (13) (12) (12) (14) (13) (13) (13) (27 to 4) (27 to 3)
SF-12 physical 30 29 40 38 47 46 11 9 17 16 2 1
(0 to 100) (7) (6) (10) (9) (8) (9) (10) (9) (10) (9) (22 to 6) (23 to 5)
SF-12 mental 51 49 55 52 57 55 3 2 6 5 1 1
(0 to 100) (11) (11) (8) (10) (6) (8) (9) (12) (11) (10) (23 to 5) (24 to 5)

ADL = activities of daily living, Con = control group, Exp = experimental group, LEFS = Lower Extremity Functional Scale, ROM = range of movement, SF-12 = Short Form 12 quality of
life questionnaire component summary scores.
140 Schache et al: Hip abductor strengthening following knee arthroplasty

a 100
b 100
KOOS symptoms (0 to 100)

80 80

KOOS pain (0 to 100)

60 60

40 40

Experimental 20 Experimental
20
Control Control

0 0
0 6 26 0 6 26
Time (weeks) Time (weeks)

c d
KOOS activities of daily living (0 to 100)

100 100

KOOS quality of life (0 to 100)

80 80

60 60

40 40

Experimental 20 Experimental
20
Control Control
0 0
0 6 26 0 6 26
Time (weeks) Time (weeks)

Figure 3. Knee Injury and Osteoarthritis Outcome Score (KOOS) by time, for the subscales (a) symptoms, (b) pain, (c) activities of daily living, and (d) quality of life. Symbols show
individual participants’ outcomes. Lines join group means at baseline and at Weeks 6 and 26. Experimental and control group data have been offset slightly for clarity.

during the rehabilitation period being completed by 74% of the The KOOS subscales all showed improvement over the 26-week
experimental group and 71% of the control group. All prospectively study period (Figure 3). Again, however, there were no significant
registered primary and secondary outcomes were reported. The differences for any of the KOOS subscales between the experimental
planned assessment at 3 weeks was not undertaken, but the regis- and control groups at either Week 6 or Week 26 (Table 3). The mean
tered assessments at Weeks 0, 6 and 26 are reported here. estimates of the between-group differences ranged between 1 and 5
on the 100-point subscales, and were therefore all well below the
smallest worthwhile effect of 10. In some subscales, the confidence
Flow of participants through the study
intervals around these between-group differences did include the
smallest worthwhile effect of 10 (Table 3).
Between October 2015 and August 2016, 315 patients were
screened for eligibility. One hundred and twelve patients were
eligible and 105 patients (69 females, 36 males) agreed to participate. Secondary outcomes
These participants were randomly allocated to experimental and
control groups (Figure 1). Although both groups experienced improvements in function and
patient-reported outcomes, the incorporation of targeted hip
Characteristics of the participants strengthening did not result in significantly greater improvements
than the control intervention at either Week 6 or Week 26 (Tables 2
At baseline, the demographic characteristics of the two groups and 3).
were similar, as shown in Table 1. The control group had a higher
proportion of participants with a previous arthroplasty of the other
Per-protocol analysis
knee and other hip, but the key baseline clinical characteristics of the
groups were similar (as shown in Table 1 and the first two columns of
Thirty-six participants from the experimental group and 43 par-
data in Table 2 and Table 3).
ticipants from the control group were deemed to be adherent and
were analysed in the per-protocol analysis. There were no significant
Primary outcomes between-group differences in most of the primary and all of the
secondary outcomes at Week 6 and Week 26 (Tables 4 and 5). Two
Hip abductor strength improved in both the experimental and KOOS domains (symptoms and activities of daily living) had a sig-
control groups over the 26-week study period, with most improve- nificant between-group difference in favour of the experimental
ment occurring during the initial 6-week supervised exercise period group at Week 6. The mean differences and lower limits of the con-
(Figure 2). However, the average amount of improvement in hip fidence intervals were below the smallest worthwhile effect, indi-
strength did not differ between the experimental and control groups cating uncertainty about whether the effect may be clinically
(Table 2), with mean between-group differences of 0 N/kg/m2 both at worthwhile. Neither effect remained statistically significant at Week
Week 6 and Week 26. The confidence intervals around these 26 (Table 5).
between-group differences extended no further than 0.5 N/kg/m2 in Individual participant data are presented in Table 6, which is
either direction (Table 2). available on the eAddenda.
 Research 141

Table 4
Per-protocol analysis of mean (SD) of groups, mean (SD) within-group difference, and mean (95% CI) between-group difference for pain and objectively measured outcomes.

Outcome Groups Within-group difference Between-group difference

Week 0 Week 6 Week 26 Week 6 minus Week 26 Week 6 minus Week 26 minus
Week 0 minus Week 0 Week 0 Week 0

Exp Con Exp Con Exp Con Exp Con Exp Con Exp minus Exp minus
(n = 36) (n = 43) (n = 36) (n = 42) (n = 33) (n = 40) Con Con

Pain 3 4 1 2 0 0 22 22 23 23 0 0
(0 to 10) (2) (2) (2) (2) (1) (1) (2) (2) (2) (3) (21 to 1) (21 to 2)
Flexion ROM 81 81 109 106 120 117 28 25 39 36 3 3
(deg) (8) (12) (8) (10) (6) (9) (10) (10) (11) (12) (21 to 8) (22 to 8)
Extension ROM 25 26 22 22 0 21 3 4 5 5 21 21
(deg) (6) (5) (3) (3) (2) (2) (5) (5) (6) (5) (23 to 2) (23 to 2)
Hip strength 1.8 2.2 3.5 3.8 4.2 4.2 1.7 1.6 2.3 1.9 0.1 0.3
(N/kg/m2) (0.9) (1.1) (1.5) (1.8) (1.6) (1.7) (1.0) (1.2) (1.2) (1.3) (20.4 to 0.6) (20.2 to 0.9)
Quadriceps strength 1.6 1.5 5.0 5.6 6.8 7.3 3.4 4.0 5.2 5.7 20.6 20
(N/kg/m2) (0.9) (1.1) (2.3) (3.0) (2.8) (3.5) (1.9) (2.7) (2.5) (3.2) (21.7 to 0.5) (21.9 to 0.8)
Chair stand test 7 8 13 14 15 15 6 6 8 8 0 0
(n in 30 s) (3) (2) (3) (5) (4) (5) (2) (4) (3) (4) (21 to 2) (22 to 2)
Stair climb test 25 22 8 8 7 7 216 214 218 216 22 22
(s) (10) (9) (3) (3) (2) (2) (10) (7) (9) (7) (26 to 2) (26 to 2)
40-m fast-paced walk 91 85 34 35 28 30 257 251 252 254 26 2
(s) (55) (44) (11) (13) (7) (11) (52) (39) (28) (38) (227 to 14) (213 to 18)
Timed Up and Go 29 25 10 10 8 8 220 215 219 216 25 23
(s) (13) (11) (3) (4) (2) (3) (12) (9) (10) (9) (29 to 0) (27 to 2)
Step taps 5 5 16 17 17 18 10 12 11 13 22 22
(n) (5) (6) (4) (6) (4) (5) (4) (7) (5) (6) (24 to 1) (24 to 1)
6-minute walk test 178 198 415 420 479 476 237 222 292 276 15 17
(m) (74) (109) (91) (133) (101) (138) (73) (93) (77) (100) (223 to 53) (226 to 59)

Con = control group, Exp = experimental group, ROM = range of movement.

Discussion There is limited research against which to compare these results.

In a small pilot study by Harikesavan et al,11 the experimental group
This is the first large, longitudinal, randomised controlled trial to achieved significantly greater improvements in hip abductor and
investigate the effects of incorporating hip abductor strengthening quadriceps strength, and some of the physical performance measures
exercises in rehabilitation after total knee arthroplasty. In each group, compared with the control group 12 months after total knee
within-group improvements were observed in muscle strength, arthroplasty, which did not occur in the current study. One notable
function and patient-reported outcomes following total knee difference between the Harikesavan study and this trial is the content
arthroplasty. However, incorporating targeted hip strengthening did of the control group, where the inclusion of simple step-ups and stair
not result in a substantially greater improvement of the primary climbing differed to the usual care program included in this trial
outcomes when compared with the control intervention at either the where participants performed higher intensity single leg exercises.
end of the 6-week supervised rehabilitation period or 20 weeks later. These differences in the content of the control group exercise pro-
The first primary outcome was hip abductor strength. The gram likely explain why there were between-group differences in the
equivalence of the effects of the experimental and control in- Harikesavan study. The pilot study had higher frequency of outpatient
terventions on hip abductor strength was demonstrated fairly clearly physiotherapy sessions with four to five sessions per week for 4
by the absence of any mean between-group difference accompanied weeks, followed by two to three sessions per week for 12 weeks. This
by confidence intervals that excluded effects . 0.5 N/kg/m2 in either could have contributed to the significantly greater improvements in
direction. muscle strength and physical performance. A higher volume of
The other primary outcome (KOOS) indicated a mean estimate of therapy has been shown to better improve pain and functional out-
the between-group difference between 1 and 5 on the 100-point comes among patients following unilateral total knee arthroplasty.40
subscales. While these mean differences would not be considered The frequency of physiotherapy sessions in our study is consistent
worthwhile and these results were statistically non-significant, some with current practice. A higher frequency of visits similar to the pilot
of the confidence intervals did extend to or slightly beyond the study would significantly increase the cost of the rehabilitation and
smallest worthwhile effect of 10, indicating that the possibility of a limit accessibility to rehabilitation due to the high cost. The results of
worthwhile effect was not completely excluded by the data in this this study are similar to the results in other trials where the addition
study. These were also the only effects to identify a statistically sig- of specific exercises to usual care have not resulted in greater
nificant benefit on the per-protocol analysis, suggesting that if improvement in outcomes following total knee arthroplasty.39,41
adherence could be improved, an effect might be more evident. The current study showed that both the experimental and control
However, with 18 outcomes being assessed at two time points for groups experienced improvements in hip abduction strength. The
both intention-to-treat and per-protocol analyses, the possibility that rationale for including the targeted hip strengthening exercises was
these statistically significant findings were Type-I errors must be based on maximal activation of gluteal muscles. The weight-bearing
acknowledged. Therefore, overall, there is no robust evidence of an nature of the functional exercises, which recruited the hip abduc-
effect on the KOOS. tors to stabilise the pelvis, also resulted in significant increases in hip
Across the secondary outcomes, the mean estimates of effect were abductor strength after total knee arthroplasty. Functional exercises
generally very small, with more than half the secondary outcomes may increase hip abductor muscle strength as effectively as targeted
having a mean between-group difference with a magnitude of  1 hip abductor exercises, and the strength gains of the hip abductors
unit on whatever scale was used. Although a formal smallest during the non-specific functional exercises may have been under-
worthwhile effect was not nominated for each secondary outcome estimated. Both programs in the current study were progressive in
measure, the upper (ie, most favourable) limits of the confidence dosage. The loads imposed on the hip abductors by the program
intervals was  10% of the available scale for most of the secondary that the control group experienced were similar to the loads expe-
outcome measures. rienced by the experimental group. This presumably resulted in
142 Schache et al: Hip abductor strengthening following knee arthroplasty

Table 5
Per-protocol analysis of mean (SD) of groups, mean (SD) within-group difference, and mean (95% CI) between-group difference for questionnaires.

Outcome Groups Within-group difference Between-group difference

Week 0 Week 6 Week 26 Week 6 minus Week 26 Week 6 minus Week 26 minus
Week 0 minus Week 0 Week 0 Week 0

Exp Con Exp Con Exp Con Exp Con Exp Con Exp minus Exp minus
(n = 36) (n = 43) (n = 36) (n = 41) (n = 33) (n = 40) Con Con

KOOS (0 to 100)
symptoms 44 47 68 62 81 79 24 15 36 32 9 4
(13) (16) (14) (17) (13) (15) (15) (18) (14) (17) (1 to 16) (22 to 11)
pain 45 49 73 71 87 85 28 22 40 37 6 3
(14) (19) (15) (15) (10) (14) (19) (16) (13) (18) (22 to 14) (24 to 11)
ADL 52 56 83 77 91 88 31 21 38 31 10 7
(17) (22) (15) (15) (9) (15) (20) (19) (14) (22) (1 to 19) (22 to 16)
quality of life 26 27 59 55 73 69 33 28 47 41 5 5
(15) (21) (18) (23) (17) (22) (21) (22) (21) (24) (25 to 15) (25 to 16)
LEFS 22 22 43 42 56 53 21 19 33 30 2 3
(0 to 80) (12) (14) (11) (13) (11) (13) (15) (13) (12) (13) (24 to 8) (23 to 9)
SF-12 physical 29 29 40 38 49 45 12 9 20 16 3 4
(0 to 100) (7) (7) (10) (9) (7) (9) (10) (9) (9) (9) (21 to 7) (21 to 8)
SF-12 mental 52 49 54 52 58 54 3 2 6 5 0 1
(0 to 100) (11) (12) (8) (10) (5) (8) (9) (12) (11) (10) (25 to 5) (24 to 6)

ADL = activities of daily living, Con = control group, Exp = experimental group, LEFS = Lower Extremity Functional Scale, ROM = range of movement, SF-12 = Short Form 12 quality of
life questionnaire component summary scores.

improvements in hip strength in both the control group and the outcomes and patient-reported outcomes were observed in both the
targeted hip strengthening group, and therefore no difference in hip experimental and control groups.
strength in the per-protocol analysis was observed.
This study has implications for clinical practice. Hip abductor
strengthening occurred during general functional exercise as well as What was already known on this topic: Muscle weakness
targeted hip strengthening exercises. Specific hip abductor persists following total knee arthroplasty. Hip abductor strength
strengthening exercises may therefore not be necessary to improve is associated with functional performance following total knee
hip strength and other outcomes following total knee arthroplasty. arthroplasty.
An equally successful outcome can be achieved from performing What this study adds: Similar improvements in muscle
strength, functional performance and patient-reported outcomes
functional exercises in this population. This has been demonstrated in
are observed when targeted hip strengthening exercises are
people following total hip arthroplasty who have had a similar
incorporated into rehabilitation in adults following total knee
diagnosis of osteoarthritis and subsequent joint replacement.42 This arthroplasty in place of general functional exercises.
is particularly useful in the early postoperative period when pain and
swelling can be a barrier to exercise progression and emphasis is on
regaining functional movements such as getting out of a chair and
eAddenda: Table 6 can be found online at: https://doi.org/10.1016/
walking.
j.jphys.2019.05.008.
A possible research implication from the study is suggested by the
Ethics approval: The La Trobe University Faculty Human Ethics
more positive findings on the per-protocol analysis. That is, the hip
Committee approved this study (FHEC 14/256). All participants gave
abductor strengthening exercises might be worthy of further research
written informed consent before data collection began.
if a way could be found to improve either their intensity or the pa-
Competing interests: Nil.
tients’ adherence to them.
Source(s) of support: Nil.
There were a number of limitations in this study. Participants in
Acknowledgements: Nil.
the experimental group were not isolated from those in the control
Provenance: Not invited. Peer reviewed.
group, as both groups used the same gym. However, the gym was
Correspondence: Margaret Schache, Physiotherapy Department,
large and included many patients with other conditions. A number of
Donvale Rehabilitation Hospital, Melbourne, Australia. Email:
different therapists with varying years of experience delivered the
[email protected]
treatments. The therapists were instructed in the rehabilitation pro-
tocols prior to commencement of the trial, with opportunities for
clarification throughout the trial as well as regular verbal encour- References
agement for the duration of the trial. The compliance level of the
1. Bade MJ, Kohrt WM, Stevens-Lapsley JE. Outcomes before and after total knee
participants was high, although not perfect because the trial was arthroplasty compared to healthy adults. J Orthop Sports Phys Ther. 2010;40:559–
conducted over a 6-month period and some of the patients had 567.
holidays during this period. This trial was conducted at one centre 2. Australian Orthopaedic Association National Joint Replacement Registry. Annual
Report. 2018.
with privately insured patients only. Therefore, the generalisability of 3. Mizner RL, Petterson SC, Snyder-Mackler L. Quadriceps strength and the time
the results to the public sector or other centres may be limited. Note course of functional recovery after total knee arthroplasty. J Orthop Sports Phys
that the demographics of the cohort in this study match those re- Ther. 2005;35:424–436.
4. Petterson SC, Mizner RL, Stevens JE, Raisis L, Bodenstab A, Newcomb W, et al.
ported in previous investigations.10,39 Improved function from progressive strengthening interventions after total knee
Overall, this well-designed and appropriately powered study arthroplasty: A randomized clinical trial with an imbedded prospective cohort.
showed that the incorporation of targeted hip strengthening into Arthritis Care Res. 2009;61:174–183.
5. Mizner RL, Snyder-Mackler L. Altered loading during walking and sit-to-stand is
standard rehabilitation after total knee arthroplasty did not clearly affected by quadriceps weakness after total knee arthroplasty. J Orthop Res.
demonstrate superior muscle strength, functional outcomes or 2005;23:1083–1090.
patient-reported outcomes when compared with a control group at 6. Schache MB, McClelland JA, Webster KE. Lower limb strength following total knee
arthroplasty: a systematic review. Knee. 2014;21:12–20.
the end of a 6-week period of supervised rehabilitation or 20 weeks
7. Walsh M, Woodhouse LJ, Thomas SG, Finch E. Physical impairments and functional
later. The relevance of specific hip strengthening could be questioned, limitations: a comparison of individuals 1 year after total knee arthroplasty with
as very similar improvements in hip abductor strength, functional control subjects. Phys Ther. 1998;78:248–258.
 Research 143

8. Alnahdi AH, Zeni JA, Snyder-Mackler L. Hip abductor strength reliability and as- 26. Parent E, Moffet H. Comparative responsiveness of locomotor tests and question-
sociation with physical function after unilateral total knee arthroplasty: a cross- naires used to follow early recovery after total knee arthroplasty. Arch Phys Med
sectional study. Phys Ther. 2014;94:1154–1162. Rehabil. 2002;83:70–80.
9. Loyd BJ, Jennings JM, Judd DL, Kim RH, Wolfe P, Dennis DA, et al. influence of hip 27. Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basic functional mobility
abductor strength on functional outcomes before and after total knee arthroplasty: for frail elderly persons. J Am Geriatr Soc. 1991;39:142–148.
post hoc analysis of a randomized controlled trial. Phys Ther. 2017;97:896–903. 28. Wright AA, Cook CE, Baxter GD, Dockerty JD, Abbott JH. A comparison of 3
10. Piva SR, Teixeira PE, Almeida GJ, Gil AB, DiGioia AM, Levison TJ, et al. Contribution methodological approaches to defining major clinically important improvement of
of hip abductor strength to physical function in patients with total knee arthro- 4 performance measures in patients with hip osteoarthritis. J Orthop Sports Phys
plasty. Phys Ther. 2011;91:225–233. Ther. 2011;41:319–327.
11. Harikesavan K, Chakravarty RD, Maiya AG, Hegde SP, Shivanna Y. Hip abductor 29. Jones CJ, Rikli RE, Beam WC. A 30-s chair-stand test as a measure of lower body
strengthening improves physical function following total knee replacement: one- strength in community-residing older adults. Res Q Exerc Sport. 1999;70:113–119.
year follow-up of a randomized pilot study. Open Rheumatol J. 2017;11:30–42. 30. Hill KD, Bernhardt J, McGann AM, Maltese D, Berkovits D. A new test of dynamic
12. Bennell KL, Hunt MA, Wrigley TV, Hunter DJ, McManus FJ, Hodges PW, et al. Hip standing balance for stroke patients: reliability, validity and comparison with
strengthening reduces symptoms but not knee load in people with medial knee healthy elderly. Physiother Can. 1996;48:257–262.
osteoarthritis and varus malalignment: A randomised controlled trial. Osteoar- 31. Alnahdi AH. Outcome measures capturing ICF domains in patient with total knee
thritis Cartilage. 2010;18:621–628. arthroplasty. Int J Rehabil Res. 2014;37:281–289.
13. Lack S, Barton C, Sohan O, Crossley K, Morrissey D. Proximal muscle rehabilitation 32. Koblbauer IF, Lambrecht Y, van der Hulst ML, Neeter C, Engelbert RH, Poolman RW,
is effective for patellofemoral pain: a systematic review with meta-analysis. Br J et al. Reliability of maximal isometric knee strength testing with modified hand-
Sports Med. 2015;49:1365–1376. held dynamometry in patients awaiting total knee arthroplasty: useful in
14. Nascimento LR, Teixeira-Salmela LF, Souza RB, Resende RA. Hip and knee research and individual patient settings? A reliability study. BMC Musculoskelet
strengthening is more effective than knee strengthening alone for reducing pain Disord. 2011;12:249.
and improving activity in individuals with patellofemoral pain: a systematic re- 33. Binkley JM, Stratford PW, Lott SA, Riddle DL. The Lower Extremity Functional Scale
view with meta-analysis. J Orthop Sports Phys Ther. 2018;48:19–31. (LEFS): scale development, measurement properties, and clinical application.
15. Singh S, Pattnaik M, Mohanty P, Ganesh GS. Effectiveness of hip abductor North American Orthopaedic Rehabilitation Research Network. Phys Ther.
strengthening on health status, strength, endurance and six minute walk test in 1999;79:371–383.
participants with medial compartment symptomatic knee osteoarthritis. J Back 34. Ware J, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construc-
Musculoskeletal Rehabil. 2016;29:65–75. tion of scales and preliminary tests of reliability and validity. Med Care.
16. Sled EA, Khoja L, Deluzio KJ, Olney SJ, Culham EG. Effect of a home program of hip 1996;34:220–233.
abductor exercises on knee joint loading, strength, function, and pain in people 35. Roos EM, Lohmander LS. Knee injury and Osteoarthritis Outcome Score (KOOS):
with knee osteoarthritis: a clinical trial. Phys Ther. 2010;90:895–904. from joint injury to osteoarthritis. Health Qual Life Outcomes. 2003;1:64.
17. Schache MB, McClelland JA, Webster KE. Does the addition of hip strengthening 36. Roos EM, Toksvig-Larsen S. Knee injury and Osteoarthritis Outcome Score (KOOS) –
exercises improve outcomes following total knee arthroplasty? A study protocol validation and comparison to the WOMAC in total knee replacement. Health Qual
for a randomized trial. BMC Musculoskelet Disord. 2016;17:1. Life Outcomes. 2003;1:17.
18. Bolgla LA, Uhl TL. Electromyographic analysis of hip rehabilitation exercises in a 37. Lowe CJ, Barker KL, Holder R, Sackley CM. Comparison of postdischarge physio-
group of healthy subjects. J Orthop Sports Phys Ther. 2005;35:487–494. therapy versus usual care following primary total knee arthroplasty for osteoar-
19. Ekstrom RA, Donatelli RA, Carp KC. Electromyographic analysis of core trunk, hip, thritis: an exploratory pilot randomized clinical trial. Clin Rehabil. 2012;26:629–
and thigh muscles during 9 rehabilitation exercises. J Orthop Sports Phys Ther. 641.
2007;37:754–762. 38. Fransen M, Nairn L, Bridgett L, Crosbie J, March L, Parker D, et al. Post-
20. Jacobs CA, Lewis M, Bolgla LA, Christensen CP, Nitz AJ, Uhl TL. Electromyographic acute rehabilitation after total knee replacement: a multicenter randomized
analysis of hip abductor exercises performed by a sample of total hip arthroplasty clinical trial comparing long-term outcomes. Arthritis Care Res. 2017;69:192–
patients. J Arthroplasty. 2009;24:1130–1136. 200.
21. Thorborg K, Bandholm T, Petersen J, Weeke KMO, Weinold C, Andersen B, et al. Hip 39. Bade MJ, Struessel T, Dayton M, Foran J, Kim RH, Miner T, et al. Early high-intensity
abduction strength training in the clinical setting: With or without external versus low-intensity rehabilitation after total knee arthroplasty: a randomized
loading? Scand J Med Sci Sports. 2010;20:70–77. controlled trial. Arthritis Care Res. 2017;69:1360–1368.
22. Peer MA, Lane J. The Knee Injury and Osteoarthritis Outcome Score (KOOS): A 40. Brennan GP, Fritz JM, Houck LT, Hunter SJ. Outpatient rehabilitation care
review of its psychometric properties in people undergoing total knee arthroplasty. process factors and clinical outcomes among patients discharged home
J Orthop Sports Phys Ther. 2013;43:20–28. following unilateral total knee arthroplasty. J Arthroplasty. 2015;30:885–
23. Schache MB, McClelland JA, Webster KE. Reliability of measuring hip abductor 890.
strength following total knee arthroplasty using a hand-held dynamometer. Disabil 41. Henderson KG, Wallis JA, Snowdon DA. Active physiotherapy interventions
Rehabil. 2015;20:1–4. following total knee arthroplasty in the hospital and inpatient rehabilitation set-
24. Dobson F, Hinman RS, Roos EM, Abbott JH, Stratford P, Davis AM, et al. OARSI rec- tings: a systematic review and meta-analysis. Physiotherapy: Epub ahead of print.
ommended performance-based tests to assess physical function in people diagnosed 2017.
with hip or knee osteoarthritis. Osteoarthritis Cartilage. 2013;21:1042–1052. 42. Monticone M, Ambrosini E, Rocca B, Lorenzon C, Ferrante S, Zatti G. Task-oriented
25. Kennedy DM, Stratford PW, Wessel J, Gollish JD, Penney D. Assessing stability and exercises and early full weight-bearing contribute to improving disability after
change of four performance measures: a longitudinal study evaluating outcome total hip replacement: a randomized controlled trial. Clin Rehabil. 2014;28:658–
following total hip and knee arthroplasty. BMC Musculoskelet Disord. 2005;6:3. 668.