Harmer (2009) Land- based versus water-based rehabilitation following total knee replacement

Land-Based Versus Water-Based Rehabilitation
Following Total Knee Replacement:
A Randomized, Single-Blind Trial
ALISON R. HARMER,1JUSTINE M. NAYLOR,2JACK CROSBIE,1ANDTYSON RUSSELL3
Objective.To compare outcomes between land-based and water-based exercise programs delivered in the early subacute
phase up to 6 months after total knee replacement (TKR).
Methods.Two weeks after surgery (baseline), 102 patients were randomized to participate in either land-based (n49)
or water-based (n53) exercise classes. Treatment parameters were guided by current clinical practice protocols.
Therefore, each study arm involved 1-hour sessions twice a week for 6 weeks, with patient-determined exercise intensity.
Session attendance was recorded. Outcomes were measured at baseline and at 8 and 26 weeks postsurgery. Outcomes
included distance on the 6-Minute Walk test, stair climbing power (SCP), the Western Ontario and McMaster Universities
(WOMAC) Osteoarthritis Index (n85 English-procient patients), visual analog scale for joint pain, passive knee range
of motion, and knee edema (circumference). Planned orthogonal contrasts, with an intent-to-treat approach, were used
to analyze the effects of time and time-group interactions.
Results.Compliance in both groups was excellent with 81% attending 8 or more sessions. Loss to followup was 5%.
Signicant improvements were observed across time in all outcomes at 8 weeks, with further improvements evident in
all variables (except WOMAC pain) at 26 weeks. Minor between-group differences were evident for 4 outcomes (SCP,
WOMAC stiffness, WOMAC function, and edema) but these appear clinically insignicant.
Conclusion.A short-term, clinically pragmatic program of either land-based or water-based rehabilitation delivered in
the early phase after TKR was associated with comparable outcomes at the end of the program and up to 26 weeks
postsurgery.
INTRODUCTION
The provision of physiotherapeutic rehabilitation in the
weeks immediately following total knee replacement
(TKR), either in an outpatient or inpatient facility, has
been reported to be common practice, if not routine, in
Australia (1), Canada (2), the UK, and the US (3). However,the evidence on which the provision of this service is
based is limited and wide variation is evident in the modes
of physiotherapy employed (1). The evaluation of rehabil-
itation after TKR should be a priority, particularly in light
of the 210-fold global increases in the incidence of TKR
over the last 10 20 years (4 7).
Physiotherapy after TKR includes use of modalities to
reduce pain and edema, specic exercises to enhance joint
range of motion, exercises to improve muscle strength and
endurance, and transfer and gait retraining that enhance
physical function and quality of life (1,8). Typically out-
come measures reect these priorities. However, the abil-
ity of the postsurgical patient to participate in weight-
bearing exercise may be limited by pain and edema of the
knee joint and by pre-existing comorbid conditions. An
exercise medium that can unload the operated knee and
other painful joints may allow patients to exercise more
effectively with less pain and swelling. Therefore, the in-
herent buoyancy and increased hydrostatic pressure of
water make it a potentially attractive treatment after TKR.
Hydrotherapy appears to provide short-term benet for
patients with knee or hip osteoarthritis, and therefore has
been suggested as an initial treatment option, particularly
in patients with severe disease (9). It is utilized in isolation
ACTRN: 12608000100392.
Supported by a grant from the Ingham Research Institute
(formerly the Health Research Foundation Sydney South
West Health Service).
1Alison R. Harmer, PhD, BAppSc, Jack Crosbie, PhD,
MSc: University of Sydney, Sydney, New South Wales, Aus-
tralia;
2Justine M. Naylor, PhD, BAppSc: Liverpool Hospital,
University of Sydney, and University of New South Wales,
Sydney, New South Wales, Australia;
3Tyson Russell,
BAppSc: Faireld Hospital and University of Sydney, Syd-
ney, New South Wales, Australia.
Address correspondence to Alison R. Harmer, PhD, Clin-
ical and Rehabilitation Sciences Research Group, Physio-
therapy, Faculty of Health Sciences, University of Sydney,
PO Box 170, Lidcombe, New South Wales, Australia, 1825.
E-mail: A.Harmer@usyd.edu.au.
Submitted for publication March 10, 2008; accepted in
revised form September 25, 2008.
Arthritis & Rheumatism (Arthritis Care & Research)
Vol. 61, No. 2, February 15, 2009, pp 184 191
DOI 10.1002art.24420
2009, American College of Rheumatology
ORIGINAL ARTICLE
184

or to supplement land-based therapy, and is provided as
either group or individualized therapy (1) by providers
who have access to either an in-house or a community
pool. Randomized trials have compared the effectiveness
or cost-effectiveness of hydrotherapy versus land-based
programs for patients with lower extremity conditions in-
cluding severe hip (10 14) or knee osteoarthritis (1114),
and juvenile idiopathic arthritis (15). Although associated
with improvements in function, pain relief, andor quality
of life (10 15), hydrotherapy has not been shown to offer
a clear advantage over land-based programs, especially in
view of the expense if provided in house (15). However, if
offered in a community pool, hydrotherapy is considered
to be a cost-effective treatment option (13,15). This, to-
gether with evidence that community dwelling patients
with arthritis appear content to continue with hydrother-
apy after being exposed to it (16), indicates that research
concerning the value of hydrotherapy has clinical appli-
cation even in a resource-poor health system. Despite the
potential benets of hydrotherapy in the early rehabilita-
tion stage, no randomized trials have investigated the com-
parative benets of land-based versus water-based therapy
after TKR.
The present study was pragmatically designed, based on
current clinical practice in Australia (1), and compared a
land-based group exercise program with a water-based
group exercise program for patients following TKR. We
hypothesized that the unloading of painful, swollen joints
in water may assist resorption of edema. Therefore, water-
based exercise after TKR will achieve greater improve-
ments in physical function and range of motion than land-
based therapy.
PATIENTS AND METHODS
Study design.A randomized single-blind trial of pa-
tients undergoing physiotherapy after primary TKR was
conducted in a metropolitan public hospital. All patients
provided voluntary, written informed consent prior to
study enrollment. Outcome measures were assessed at
baseline (2 weeks postsurgery), after 6 weeks of rehabili-
tation treatment, and at 26 weeks postsurgery. Ethical ap-
proval for the study was granted by the Institutional Re-
view Board.
Recruitment of participants.Consecutive, eligible pa-
tients attending the preoperative clinic were informed of
the study. Patients were rescreened for eligibility after
surgery, when they returned for the standard review fol-
lowing discharge from acute care. Exclusion criteria in-
cluded postoperative deep joint infection, bilateral joint
surgery or surgery planned for another joint within 6
months, and documented dementia or other neurologic
condition that precluded informed consent. Eligible con-
senting patients were assessed prior to allocation to treat-
ment group. Random allocation (using a random number
generator) to either land-based treatment (the hospital
physiotherapy gymnasium) or water-based treatment
(community pool) was achieved by the drawing of consec-
utively numbered, sealed envelopes by a person unin-volved in the study. Using data obtained from an earlier
TKR rehabilitation trial (17), calculations determined that
a sample size of 40 patients per group would provide 80%
power to detect a 20% difference between groups in
6-minute walk distance, at a signicance level ofP0.05.
Therefore, 102 patients were recruited to allow for a 25%
dropout rate.
Assessments and outcomes.Patients were assessed by a
blinded assessor, but on a few occasions unblinding oc-
curred due to assessor absence. At baseline, demographic
and anthropometric data, surgical details, and presence of
comorbid conditions were recorded. Anthropometric mea-
surements included body mass (kg) and height (m). Body
mass index (kgm
2) was calculated from these data. Out-
comes were selected on the basis of their importance and
relevance to patients and clinicians. The primary outcome
measured at each assessment was the 6-Minute Walk test.
Secondary outcomes were stair climbing power (SCP), the
Western Ontario and McMaster Universities (WOMAC)
Osteoarthritis Index, a visual analog scale (VAS) for pain
in the operated knee, passive knee joint range of motion
(ROM), and edema of the knee. The use of walking aids
was also recorded to assess the level of gait independence.
The 6-Minute Walk test was conducted on a standard
25-meter track and the distance walked in meters re-
corded. The 6-Minute Walk test was chosen as the primary
outcome because it has excellent testretest reliability and
responsiveness after TKR and permits decision making at
an individual patient level (18,19). The time to ascend 18
stairs (ights of 8 and 10 stairs, separated by a small
landing) as rapidly as possible, using handrails and walk-
ing aids as required, was recorded. SCP was calculated
using body mass, total stair height, and ascent time (20).
This functional test for mobility-impaired older adults, in
whom leg power is more important than strength for mo-
bility, may be an optimal tool for measuring the magnitude
of impairment and effects of therapy (20), and SCP has
good reliability and responsiveness after TKR (18).
The WOMAC has excellent reliability, validity, and re-
sponsiveness when used with patients after TKR
(19,21,22). The questionnaire (version LK3.1) was admin-
istered by the assessor to ensure completeness of the sur-
vey responses. The WOMAC comprises 3 subscales: pain
(5 items), stiffness (2 items), and function (difculty; 17
items) with each item scored on a 0 4 Likert scale. Each
subscale is summed (maximum 20 for pain, 8 for stiffness,
and 68 for function), with a higher score indicating worse
status (23). Raw scores for each subscale were analyzed
without transformation. Improvements within this scale
are intended to reect specic joint improvement since the
items pertain to specic joint behavior. Patients who were
not procient in English were excluded (8 in the land-
based, 9 in the water-based group), therefore WOMAC
responses are reported for 85 patients. Patients recorded
the highest intensity of pain experienced in their operated
knee in the previous 24 hours using a 0 10 cm VAS.
Passive joint range limitations are clinically important,
resulting in knee pain and disordered gait (24). Passive
knee exion was measured using a long-arm universal
Exercise Program Outcomes Following Total Knee Replacement185

goniometer and the recommended leg landmarks (25)
while the patient was seated upright on a standard height
chair. Passive knee extension was measured with the pa-
tient supine. Knee edema was estimated by circumferen-
tial measurements (cm) at 4 locations (apex, midline, su-
perior border, and 4 cm proximal to the superior border of
the patella) with the knee in extension. The 4 measures
were averaged for each knee and used as the knee edema
outcome measure. Additionally, complications were mon-
itored up to 26 weeks postsurgery using a standardized
question form administered through patient interview at
review clinics or via a followup phone call after discharge.
Treatment.Treatment parameters for both groups were
selected so as to reect the frequency, intensity, and du-
ration of current rehabilitation protocols for such patients
(1). Patients attended sessions twice a week for 6 weeks
and attendance was recorded at each session. Because all
classes were supervised, class attendance was considered
equivalent to compliance. Treatment sessions for each
mode were 60 minutes long, which included 5-minute
warm-up and cool-down periods. Per clinical practice (1),
patients controlled the exercise intensity, but were in-
structed to exercise to their tolerance level. The same
physiotherapist conducted all land-based and water-based
classes. This, together with a highly prescribed exercise
program, was intended to ensure the classes were similar
in content and intensity despite the differing media. In
addition to the supervised classes, all patients were in-
structed on a simple home exercise program, which was
recommended to be undertaken daily and consisted of
general active ROM exercises and walking as tolerated.
Home program compliance was not monitored.
Land-based (gymnasium) treatment comprised cycling
on a stationary ergometer, walking on a motor driven
treadmill, climbing stairs (stationary stepper machine and
a set of 5 stairs with handrails), standing isometric, bal-
ance and knee ROM exercises at a bar, and sit to stand
exercises from chairs of varying heights. Water-based treat-
ment was conducted in a community pool that ranged
from 0.51.6 meters deep, which allowed the water level
to be approximately waist high for each participant. The
water was heated to a meanSD temperature of 253C.
A waterproof dressing was applied to the surgical site on
each patients knee immediately before immersion. Each
session included repetitions of walking forward and back-
ward, stepping sideways, step-ups, jogging, jumping, kick-
ing, knee ROM exercises, lunges, and combined squats and
upper extremity exercises.
Statistical analysis.Between-group differences for
baseline demographic, anthropometric, and surgical data
were analyzed using independentt-tests for continuous
data or chi-square test for independent dichotomous data.
The primary and secondary outcome measures were ana-
lyzed using planned orthogonal contrasts for repeated
measures. Statistical Package for the Social Sciences soft-
ware, version 14 (SPSS, Chicago, IL) was used. Contrasts
were performed with a Helmert comparison (baseline com-
pared with 8 and 26 weeks, 8 weeks compared with 26
weeks) and yielded effects for time, time-by-group inter-action, and an overall effect of group. All analyses were
performed per intent-to-treat. Compliance (attendance at
treatment sessions) was dichotomized into those who at-
tended less than 8 sessions (noncompliant) and those who
attended 8 or more sessions (compliant).Pvalues less than
0.05 were considered signicant and were further assessed
by condence interval analysis where appropriate. The
odds ratio (OR) of requiring a walking stick for ambulation
at 26 weeks was calculated. Data were reported as mean
SD or as mean difference (95% condence interval 95%
CI).
RESULTS
Participant recruitment.Of the 216 patients who un-
derwent surgery between November 2005 and November
2006, 102 (47%) consented and, therefore, were available
for randomization (Figure 1). Of these, 1 patient from the
water-based group died from a cardiac event 4 weeks post-
surgery. Two patients from the water-based group with-
drew from the intervention during the treatment phase and
declined the 8-week assessment; however, they returned
for the nal assessment. Two patients were unable to at-
tend the nal assessment (one was admitted to the hospital
with a deep surgical site infection and the other relocated).
Figure 1.Participant ow in land-based and water-based rehabil-
itation groups after total knee replacement (TKR) surgery.
186Harmer et al

Baseline data.No between-group differences were evi-
dent at baseline, except for a higher frequency of heart
disease in the land-based group (Table 1). Comorbidity
was common, with hypertension (71%) and obesity (52%)
featuring most frequently. Eight percent of the cohort were
morbidly obese (BMI40), and 51% had3 comorbid
conditions. Cement xation of the prosthesis was almost
universal.
Numbers analyzed.Consistent with the intent-to-treat
analyses employed, missing data for the 5 patients who
missed assessments (Figure 1) were handled by carrying
forward the previous value for that patient. For 1 patient,
the 6-Minute Walk test, the SCP test, and joint ROM tests
were not conducted at baseline and the group mean was
used to estimate baseline values. Three patients did not
perform the SCP test at baseline due to resting tachycardia
(2 in the land-based and 1 in the water-based group). These
patients were assigned the mean baseline power for the
group. Nine patients could only complete 1 ight of stairs
at baseline. For these patients, stair ascent time for the
complete test (2 ights) was estimated from a regression
equation of all paired single ight times versus double
ight times, from which each patients power was then
calculated. Analyses for SCP included and excluded these
9 patients, and there was no difference in the results. Data
for all 102 patients are presented.
Compliance.Compliance did not differ between treat-
ment groups (mean difference 0.18 sessions, 95% CI0.7,
1.1). Overall, 81% of patients attended8 of 12 sessions
(meanSD 9.32.3).Outcomes.The distance walked during the 6-Minute
Walk test increased signicantly up to 26 weeks postsur-
gery (201 meters, 95% CI 184, 218 meters;P0.000), with
no between-group differences (mean difference 3 meters,
95% CI32, 38 meters) (Figure 2A).
SCP continued to improve up to 26 weeks postsurgery
(91 watts, 95% CI 81, 101 watts;P0.000) (Figure 2B).
The water-based group improved more than the land-
based group between 8 and 26 weeks postsurgery (P
0.005).
For the WOMAC subscales (n85), WOMAC pain was
reduced at 8 weeks postsurgery (6.5 points, 95% CI 5.6,
7.5 points) and reached a plateau thereafter, with no group
differences (0.4 points, 95% CI0.5, 1.3 points) (Figure
3A). WOMAC stiffness was reduced up to 26 weeks post-
surgery (2.9, 95% CI3.2,2.5 points;P0.000), with
the land-based group reporting less stiffness overall than
the water-based group (0.5 points, 95% CI0.1,1.0
Figure 2.Physical performance tests for land-based (circles) and
water-based (triangles) rehabilitation groups at 2, 8, and 26 weeks
after total knee replacement surgery inA,6-Minute Walk test and
B, Stair climbing power. The land-based group is slightly offset to
the left, the water-based group to the right at each time point to
allow a clear representation of the mean. ***Main effect for
time, 8 and 26 weeks combined2 weeks, and 268 weeks;P
0.001.ainteraction effect, 8 vs. 26 weeks, water-basedland-
based;P0.005.
Table 1. Baseline demographic, anthropometric,
comorbidity, and surgical details*
Land-based
(n49)Water-based
(n53)
Age, meanSD years 67.86.3 68.79.1
Women 28 (57) 30 (57)
Body mass, meanSD kg 81.013.7 84.016.0
Height, meanSD meters 1.630.09 1.630.11
BMI, meanSD kgm
230.65.0 31.65.8
Obese 25 (51) 28 (53)
Comorbidities
Endocrine 13 (27) 21 (40)
Gastrointestinal 9 (18) 15 (28)
Heart disease 18 (37) 9 (17)
Hypertension 37 (76) 35 (66)
Other joint disease 6 (12) 12 (23)
Respiratory 12 (24) 16 (30)
3 comorbidities 25 (51) 27 (51)
Surgical details
Surgery time, mean SD
minutes10817 11421
Cement used 49 (100) 51 (96)
Patella resurfaced 12 (25) 13 (25)
* Values are the number (percentage) unless indicated otherwise.
P0.05; 95% condence interval 336.
Exercise Program Outcomes Following Total Knee Replacement187

points;P0.02), but improving less from 8 to 26 weeks
(P0.02) (Figure 3B). WOMAC function (difculty) de-
creased up to 26 weeks postsurgery (27.6 points, 95% CI
30.0,25.2 points;P0.000) with the water-based
group showing more improvement from 8 to 26 weeks
compared with the land-based group (P0.04) (Figure
3C).
Reported VAS pain intensity in the operated joint con-
tinued to decrease up to 26 weeks postsurgery (3.2
points, 95% CI 2.7, 3.8 points;P0.000) with no between-
group differences evident (0.5 points, 95% CI0.2, 1.1
points) ( Figure 3D).
Knee joint exion ROM continued to improve to 26
weeks postsurgery (24, 95% CI 21, 27;P0.000), with
no group differences evident (3, 95% CI1, 7) (Figure
4A). Knee extension also continued to improve to 26
weeks postsurgery (7, 95% CI5,9;P0.000), with
no group differences (1, 95% CI1, 3) (Figure 4A).
For knee edema, mean circumference of the nonoper-
ated knee (41.4 cm, 95% CI 40.6, 42.3 cm) did not change
over time and did not differ between groups (0.1 cm, 95%
CI1.9, 1.6 cm,P0.89) (Figure 4B). Edema in the
operated knee continued to decrease up to 26 weeks post-
surgery (2.8 cm, 95% CI3.2,2.3 cm;P0.000) witha slightly greater reduction in the water-based group be-
tween 8 weeks and 26 weeks (P0.03) (Figure 4B).
At baseline, 97% of patients required a walking aid at
any time. By 26 weeks, 27% required an aid. There was a
small increase in risk of requiring a walking aid at 26
weeks in the water-based group (OR 1.2, 95% CI 0.5, 2.9).
However, this was not signicant on chi-square test.
Adverse events and complications.One patient died of
a cardiac event that appeared to be unrelated to rehabili-
tation treatment (the event occurred over the weekend
between the second and third treatment sessions). Thirty-
two patients experienced 34 complications during the year
and 18 of those complications occurred after discharge
from acute care (Table 2). Six patients were readmitted to
the hospital; 4 of the 6 required manipulation under anes-
thetic, 1 had a deep surgical site infection, and 1 had a
documented hemarthrosis.
DISCUSSION
This is the rst randomized trial comparing land-based
and water-based therapy after TKR. The value of water-
Figure 3. AC,Western Ontario and McMaster Universities (WOMAC) Osteoarthritis Index subscale
scores (n85), andD,visual analog scale (VAS) pain score for land-based (circles) and water-based
(triangles) rehabilitation groups at 2, 8, and 26 weeks after total knee replacement surgery.A, WOMAC
pain. ***main effect for time, 8 and 26 weeks combined2 weeks;P0.001.B,WOMAC stiffness.
***main effect for time, 8 and 26 weeks combined2 weeks, and 268 weeks;P0.001.a
interaction effect (time-by-group) 8 vs. 26 weeks, improvement greater in water-based group;P0.02.b
group effect, water-basedland-based;P0.038.C,WOMAC difculty (function). ***main effect for
time, 8 and 26 weeks combined2 weeks, and 268 weeks;P0.001.ainteraction effect (time-by-
group) 8 vs. 26 weeks, improvement greater in water-based group,P0.04.D,VAS pain in operated knee
over preceding 24 hours. ***main effect for time 8 and 26 weeks combined2 weeks;P0.001.
**268 weeks;P0.01.
188Harmer et al

based therapy is particularly interesting since this treat-
ment is associated with comparatively high in-house op-
erational costs. Patients in the 2 groups improved equally
well in most outcome measures with few apparent between-
group differences. Therefore, neither mode was clearly
superior to the other. Contemporary health care relies
heavily on cost-neutral service enhancement, therefore if
water-based therapy cannot be provided without higher
underlying costs, patient access to this medium may re-
main limited to facilities already utilizing onsite pools for
multiple purposes, facilities supported by user-pay sys-
tems (such as private hospitals), or facilities in close prox-
imity to community-based pools.
Exercise after TKR is effective in improving function
(17,26) and in our study, land-based and water-based ex-
ercise programs were associated with comparable im-
provements in timed walk distance, joint pain, and rangeof motion, immediately after the program and up to 6
months postsurgery. Improvements were also evident in
stair climbing power, stiffness, function (difculty), and
edema, albeit with minor between-group differences. Re-
sults of the 6-Minute Walk test at 26 weeks were similar to
those achieved in other trials of land-based therapy after
TKR (26,27) and approached that (448 meters) of age-
matched controls (28). However, SCP results were below
those of a comparable age-matched, mobility-limited co-
hort (284 watts) (20). Our cohorts knee exion range re-
mained well below age-matched controls (143) (29) at 6
months postsurgery but was consistent with other ranges
of post-TKR cohorts (26,30 32). This may affect the po-
tential for further improvement in the WOMAC difculty
subscale, as could the presence of severe other extremity
or lumbar dysfunction (33). To date, there are no pub-
lished Australian age-matched normative data for
WOMAC outcomes. However, our patients reported low
levels of pain and stiffness 6 months after surgery, there-
fore, only minor further gains would be possible. Never-
theless, we acknowledge that although the WOMAC re-
sults accurately reect recovery, the absolute scores may
be an over-estimate of patient perceptions given the survey
was observer-administered. The potential for further im-
provement in our cohort, based on the WOMAC scores, is
arguably uncertain.
The clinical relevance and mechanisms underlying the
minor between-group differences we detected are unclear.
Water-based therapy elicited greater improvements in stair
climbing power and edema resolution at 6 months post-
surgery, but these effects were not evident immediately
after treatment cessation. Further, these improvements did
not translate into better function or timed mobility, or less
reliance on walking aids. Therefore, even if the water
medium could accelerate resolution of lower extremity
edema or temporarily unload pain-sensitive structures
(and we are unable to provide evidence that conrms this),
it did not translate into perceptible clinical gains. The
Figure 4.Passive knee joint exion and extension range of motion
and knee edema for land-based (circles) and water-based (trian-
gles) rehabilitation groups at 2, 8, and 26 weeks after total knee
replacement.A,Knee exion (upper data set) and extension
(lower data set). ***main effect for time, 8 and 26 weeks
combined2 weeks, and 268 weeks;P0.001.B. Knee edema
(mean circumference) on the operated leg (upper data set) and
nonoperated leg (lower data set). ***main effect for time in
operated leg, 8 and 26 weeks combined2 weeks, and 268
weeks;P0.001.ainteraction effect for group-by-time be-
tween 8 and 26 weeks;P0.03.
Table 2. Number of complications and adverse events in
land-based and water-based rehabilitation groups
Event Land-based Water-based
Death 1
Manipulation under
anesthetic22
Deep surgical site infection 1
Hemarthrosis 1
Thromboembolism 1 2*
Supercial surgical site
infection56
Symptomatic anemia 1 2
Cardiac 1 1
Respiratory 1 1
Urinary tract infection 1
Hematuria 1
Prolonged febrile state of
undened cause22
Electrolyte disturbance 1
* 1 patient had both a deep venous thrombosis and a pulmonary
embolus.
Exercise Program Outcomes Following Total Knee Replacement189

lesser improvement beyond 8 weeks in SCP in the land-
based group may be attributable to functional limitations
secondary to cardiac disease which was more frequent in
this group. Land-based therapy produced greater improve-
ments in patient-perceived stiffness overall; however sim-
ilar end points were achieved and ROM improved simi-
larly in both groups.
The broad inclusion criteria of the present study en-
hance the generalizability of our results to other TKR pop-
ulations. Patients were not excluded on the basis of signif-
icant comorbidity, age, or surgical approach. The high
level of compliance achieved is also noteworthy and com-
pares favorably with other TKR rehabilitation trials
(17,26,30). This is in part attributable to phone call fol-
lowup by the treating physiotherapist when a patient
failed to attend. Replication of this compliance may de-
pend on the ability to monitor attendance. Although we
employed intent-to-treat analysis, the high compliance
added further robustness. The pragmatic design of the
interventions was deliberate so that results would be im-
mediately applicable to, and the programs easily adoptable
by, rehabilitation providers.
This study followed our usual protocol for rehabilitation
after TKR and compared the outcomes associated with 2
different interventions in the absence of a control arm.
Consequently, although we have demonstrated that nei-
ther program appears to result in superior outcomes at 6
months, the contribution attributable to natural recovery
after surgery cannot be determined. It is not typical in the
vast majority of facilities for patients to receive unsuper-
vised rehabilitation after TKR (1,2). We acknowledge,
however, that a control arm would help partition the ef-
fects of natural recovery from any supplementary effects
from rehabilitation. Since we did not monitor engagement
in regular physical activity after treatment cessation, it is
not clear if the continued improvement in all outcomes
was due to the programs instilling favorable exercise hab-
its in the cohort or a manifestation of natural recovery. It is
conceivable that the outcome measures we assessed had
not stabilized (e.g., knee edema) by 6 months postsurgery,
and that improvements may have been greater had we
followed patients at 1 year, or if patients had been exposed
to more protracted rehabilitation. It is unlikely, however,
that given the similar trajectory of recovery in both groups
up to 6 months postsurgery that clinically relevant differ-
ences between groups would emerge at 1 year.
Consistent with current clinical practice in Australia,
we did not formally measure exercise intensity or compli-
ance with the home program. Therefore we cannot be
certain that the exercise dosage was similar between the
land-based and water-based groups or even of sufcient
dosage to generate physiologic change. A criticism of most
studies to date evaluating the value of rehabilitation after
TKR and hip replacement is that the intensity of the pre-
scribed exercise programs may fall short of what is neces-
sary to induce meaningful adaptations (17,34). Neverthe-
less, every effort was made to ensure the intensity of both
programs was similar by employing the same therapist for
both groups and providing the same instructions to all
patients regarding working to their tolerance level.
The temperature of the pool we used was cooler (25C)than the temperature generally utilized (3236C) for hy-
drotherapy for patients with knee and hip arthritis (35)
and recommended for reducing musculoskeletal stiffness
(36). It is theoretically possible that hydrotherapy could
have produced more favorable results if the pool temper-
ature was warmer, but previous trials involving arthritis
patients (10 15) that used warmer water have not shown
hydrotherapy to be superior to land-based alternatives. An
in-depth cost-effectiveness analysis was not conducted
given the earlier comprehensive studies by Epps et al (15)
and Cochrane et al (13).
In conclusion, in both the land-based and water-based
rehabilitation groups, improvement was evident in nearly
all outcome measures up to 6 months after TKR. Because
our original hypothesis was not supported, this study adds
to the body of work demonstrating that hydrotherapy,
although associated with clinically relevant improve-
ments, does not generally provide superior outcomes
when compared with alternative treatments for lower ex-
tremity arthritis.
ACKNOWLEDGMENTS
We are grateful for the skilled assistance of research assis-
tants Dr. Rola Ajjawi, David Guest, and Sue Skillen; and to
the Faireld Hospital Physiotherapy Manager, Lesley
Innes.
AUTHOR CONTRIBUTIONS
Dr. Harmer had full access to all of the data in the study and
takes responsibility for the integrity of the data and the accuracy
of the data analysis.
Study design.Harmer, Naylor, Crosbie.
Acquisition of data.Naylor.
Analysis and interpretation of data.Harmer, Naylor, Crosbie.
Manuscript preparation.Harmer, Naylor, Crosbie, Russell.
Statistical analysis.Harmer, Naylor.
Conducted exercise interventions.Russell.
REFERENCES
1. Naylor JM, Harmer AR, Fransen M, Crosbie J, Innes L. The
status of physiotherapy rehabilitation following total knee
replacement in Australia. Physiother Res Int 2006;11:35 47.
2. The Medical Advisory Secretariat. Physiotherapy rehabilita-
tion after total knee or hip replacement: health technology
literature review. Ontario: Ministry of Health and Long-Term
Care; 2005.
3. Lingard EA, Berven S, Katz JN, and the Kinemax Outcomes
Group. Management and care of patients undergoing total
knee arthroplasty: variations across different health care set-
tings. Arthritis Care Res 2000;13:129 36.
4. Australian Orthopaedic Association National Joint Replace-
ment Registry: annual report. Adelaide: Australian Orthopae-
dic Association; 2007.
5. Dixon T, Shaw M, Ebrahim S, Dieppe P. Trends in hip and
knee joint replacement: socioeconomic inequalities and pro-
jections of need. Ann Rheum Dis 2004;63:82530.
6. Sokka T, Kautiainen H, Hannonen P. Stable occurrence of
knee and hip joint replacement in Central Finland between
1986 and 2003: an indication of improved long-term out-
comes of rheumatoid arthritis. Ann Rheum Dis 2007;66:
341 4.
7. Katz BP, Freund DA, Heck DA, Dittus RS, Paul JE, Wright J, et
al. Demographic variation in the rate of knee replacement: a
multi-year analysis. Health Serv Res 1996;31:125 40.
190Harmer et al

8. Brander VA, Stulberg SD, Chang RW. Rehabilitation following
hip and knee arthroplasty. Phys Med Rehabil Clinics North
Am 2004;5:81536.
9. Bartels EM, Lund H, Hagen KB, Dagnrud H, Christensen R,
Danneskiold-Samsoe B. Aquatic exercise for the treatment of
knee and hip osteoarthritis. Cochrane Database Sys Rev 2007;
4:CD005523.
10. Stener-Victorin E, Kruse-Smidje C, Jung K. Comparison be-
tween electro-acupuncture and hydrotherapy, both in combi-
nation with patient education and patient education alone, on
the symptomatic treatment of osteoarthritis of the hip. Clin J
Pain 2004;20:179 85.
11. Green J, McKenna F, Redfern EJ, Chamberlain MA. Home
exercises are as effective as outpatient hydrotherapy for os-
teoarthritis of the hip. Br J Rheumatol 1993;32:8125.
12. Fransen M, Nairn L, Winstanley J, Lam P, Edmonds J. Phys-
ical activity for osteoarthritis management: a randomized con-
trolled clinical trial evaluating hydrotherapy or Tai Chi
classes. Arthritis Rheum 2007;57:40714.
13. Cochrane T, Davey RC, Matthes Edwards SM. Randomised
controlled trial of the cost-effectiveness of water-based ther-
apy for lower limb osteoarthritis. Health Technol Assess
2005;9:1114.
14. Foley A, Halbert J, Hewitt T, Crotty M. Does hydrotherapy
improve strength and physical function in patients with
osteoarthritis: a randomised controlled trial comparing a gym
based and a hydrotherapy based strengthening programme.
Ann Rheum Dis 2003;62:11627.
15. Epps H, Ginnelly L, Utley M, Southwood T, Gallivan S, Scul-
pher M, et al. Is hydrotherapy cost-effective? A randomised
controlled trial of combined hydrotherapy programmes com-
pared with physiotherapy land techniques in children with
juvenile idiopathic arthritis. Health Technol Assess 2005;9:
159.
16. Lineker SC, Badley EM, Hawker G, Wilkins A. Determining
sensitivity to change in outcome measures used to evaluate
hydrotherapy exercise programs for people with rheumatic
diseases. Arthritis Care Res 2000;13:625.
17. Moffet H, Collet JP, Shapiro SH, Paradis G, Marquis F, Roy L.
Effectiveness of intensive rehabilitation on functional ability
and quality of life after rst total knee arthroplasty: a single-
blind randomised controlled trial. Arch Phys Med Rehabil
2004;5:546 56.
18. Kennedy DM, Stratford PW, Wessel J, Gollish JD, Penney D.
Assessing stability and change of four performance measures:
a longitudinal study evaluating outcome following total hip
and knee arthroplasty. BMC Musculoskelet Disord 2005;6:3.
19. Parent E, Moffet H. Comparative responsiveness of locomotor
tests and questionnaires used to follow early recovery after
total knee arthroplasty. Arch Phys Med Rehabil 2002;83:70
80.
20. Bean JF, Kiely DK, LaRose S, Alian J, Frontera WR. Is stair
climb power a clinically relevant measure of leg power im-
pairments in at-risk older adults? Arch Phys Med Rehabil
2007;88:604 9.
21. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt
LW. Validation study of WOMAC: a health status instrument
for measuring clinically-important patient-relevant outcomes
following total hip or knee arthroplasty in osteoarthritis. J Or-
thop Rheumatol 1988;1:95108.22. Hawker G, Mel C, Paul J, Green R, Bombardier C. Compari-
son of a generic (SF-36) and a disease specic (WOMAC)
instrument in the measurement of outcomes after knee re-
placement surgery. J Rheumatol 1995;22:1193 6.
23. Bellamy N. WOMAC osteoarthritis index user guide. VII.
Queensland (Australia): 2004.
24. Brander VA, Stulberg SD. Rehabilitation after hip- and knee-
joint replacement: an experience- and evidence-based ap-
proach to care. Am J Phys Med Rehabil 2006;85(Suppl 11):
S98 118.
25. Edwards JZ, Greene KA, Davis RS, Kovacik MW, Noe DA,
Askew MJ. Measuring exion in knee arthroplasty patients. J
Arthroplasty 2004;19:369 72.
26. Kramer JF, Speechley M, Bourne R, Rorabeck C, Vaz M. Com-
parison of clinic- and home-based rehabilitation programs
after total knee arthroplasty. Clin Orthop Relat Res 2003;410:
22534.
27. Moffet H, Collet JP, Shapiro SH, Paradis G, Marquis F, Roy L.
Effectiveness of intensive rehabilitation on functional ability
and quality of life after rst total knee arthroplasty: a single-
blind randomized controlled trial. Arch Phys Med Rehabil
2004;85:546 56.
28. Steffen T, Hacker TA, Mollinger L. Age- and gender-related
test performance in community-dwelling elderly people: Six-
Minute Walk test, Berg Balance Scale, Timed Up & Go Test,
and gait speeds. Phys Ther 2002;82:128 37.
29. Shumway-Cook A, Brauer S, Woollacott M. Predicting the
probability for falls in community-dwelling older adults us-
ing the Timed Up and Go Test. Phys Ther 2000;80:897903.
30. Frost H, Lamb SE, Robertson S. A randomized controlled trial
of exercise to improve mobility and function after elective
knee arthroplasty: feasibility, results and methodological dif-
culties. Clin Rehabil 2002;16:200 9.
31. Rajan RA, Pack Y, Jackson H, Gillies C, Asirvatham R. No
need for outpatient physiotherapy following total knee
arthroplasty: a randomized trial of 120 patients. Acta Orthop
Scand 2004;75:713.
32. Codine P, Dellemme Y, Denis-Laroque F, Herisson C. The use
of low velocity submaximal eccentric contractions of the
hamstrings for recovery of full extension after total knee
replacement: a randomized controlled study. Isokinet Exerc
Sci 2004;12:215 8.
33. Naylor JM, Harmer AR, Heard RC. Severe other joint disease
and obesity independently inuence recovery after joint re-
placement surgery: a prospective, observational study. Aust J
Physiother 2008;54:57 64.
34. Suetta C, Magnusson SP, Rosted A, Aagard P, Jakobsen AK,
Larsen LH, et al. Resistance training in the early postoperative
phase reduces hospitalization and leads to muscle hypertro-
phy in elderly hip surgery patients: a controlled, randomized
study. J Am Geriatr Soc 2004;52:2016 22.
35. Bartels EM, Lund H, Hagen KB, Dagnrud H, Christensen R,
Danneskiold-Samsoe B. Aquatic exercise for the treatment of
knee and hip osteoarthritis. Cochrane Database Syst Rev 2007;
4:CD005523.
36. Elkayam O, Wigler I, Tishler M, Rosenblum I, Caspi D, Segal
R, et al. Effect of spa therapy in Tiberias on patients with
rheumatoid arthritis and osteoarthritis. J Rheumatol 1991;18:
1799 803.
Exercise Program Outcomes Following Total Knee Replacement191