VOL. 81-B, NO. 1, JANUARY 199927
C. N. A. Esler, FRCS (Tr & Orth), Senior Registrar in Orthopaedic
K. Lock, Grad Dip Phys, MCSP, Orthopaedic Research Physiotherapist
W. M. Harper, MD, FRCS Ed, Professor of Orthopaedic and Trauma
Department of Orthopaedic Surgery, Glenfield General NHS Trust, Groby
Road, Leicester LE3 9QP, UK.
P. J. Gregg, MD, FRCS, Professor of Orthopaedic Surgery
The Medical School, Framlington Place, Newcastle upon Tyne NE1 7RU,
Correspondence should be sent to Mr C. N. A. Esler.
©1999 British Editorial Society of Bone and Joint Surgery
Manipulation of total knee replacements
IS THE FLEXION GAINED RETAINED?
C. N. A. Esler, K. Lock, W. M. Harper, P. J. Gregg
From the Glenfield Hospital, Leicester, England
manipulation under anaesthesia in 47 knees.
Manipulation was considered when intensive
physiotherapy failed to increase flexion to more than
80°. The mean time from arthroplasty to manipulation
was 11.3 weeks (median 9, range 2 to 41). The mean
active flexion before manipulation was 62° (35 to 80).
One year later the mean gain was 33° (Wilcoxon
signed-rank test, range -5 to 70, 95% CI 28.5 to 38.5).
Definite sustained gains in flexion were achieved even
when manipulation was performed four or more
months after arthroplasty (paired t-test, p < 0.01, CI
8.4 to 31.4).
A further 21 patients who met our criteria for
manipulation declined the procedure. Despite
continued physiotherapy, there was no significant
increase in flexion in their knees. Six weeks to one
year after TKR, the mean change was 3.1° (paired
t-test, p = 0.23, CI -8.1 to +2).
s part of a prospective study of 476 total knee
replacements (TKR), we evaluated the use of
J Bone Joint Surg [Br] 1999;81-B:27-9.
Received 23 February 1998; Accepted after revision 16 June 1998
While the primary aims of total knee arthroplasty are relief
of pain and restoration of mobility, an adequate range of
movement (ROM) is also desirable. Laubenthal, Smidt and
Kettelkamp1assessed the amount of flexion necessary for
everyday activities and found that the mean flexion
required to climb stairs, to sit, and to tie a shoelace was
83°, 93° and 106°, respectively. The ROM attained after
total knee replacement (TKR) depends not only on such
factors as prosthetic design and soft-tissue balance, but also
on patient morphology, the preoperative ROM and motiva-
tion.2-4Some patients may be satisfied with less flexion as
long as they have relief from pain.
The long-term benefits of manipulation under anaes-
thesia (MUA) after TKR have been questioned.5The
known complications of manipulation, including supra-
condylar fracture, avulsion of the patellar tendon, myositis
ossificans and wound breakdown, may further compromise
poor results. These occur, however, in fewer than 3% of
patients.6Our aim was to evaluate the use of MUA in
patients whose maximum flexion was less than 80° despite
Patients and Methods
In 1987, we began a prospective, randomised study com-
paring 476 cemented and uncemented posterior-cruciate-
retaining TKRs (PFC; Johnson & Johnson, Bracknell, UK)
which had been performed or directly supervised by the
senior author (PJG). None of the patellae had been resur-
faced but their osteophytes had been excised. The deep and
superficial layers had been closed with interrupted sutures.
Under the supervision of a physiotherapist, knee flexion
began when the wound drains were removed 48 hours after
surgery. Continuous passive motion (CPM) was not used
except after manipulation. A goniometer was used to meas-
Manipulation was considered in patients in whom max-
imum flexion remained less than 80° despite intensive
physiotherapy. This was performed on 47 (group 1) out of
68 knees. The remaining 21 patients (group 2) declined
manipulation but continued with physiotherapy.
Group 1 (manipulation group). There were 18 men and
24 women with a mean age of 67 years (51 to 88). Five
patients had bilateral replacements, two of them simultane-
ously, and required manipulation of both knees. In all but
one, who had rheumatoid arthritis (RA), the underlying
pathology was osteoarthritis (OA). High tibial osteotomy
had been performed on two knees: these were the only
previous surgical procedures in this group. There were
varus deformities in 38 knees. In 19 cases neither the
femoral nor the tibial component had been cemented.
Before arthroplasty, the mean active flexion was 102° (60
to 135). The mean time from implantation to manipulation
was 11.3 weeks (median 9, range 2 to 41). Before manip-
ulation, the mean flexion was 62° (median 65, range 35 to
Group 2 (patients who declined manipulation). There
were nine men and 12 women with a mean age of 67 years
(46 to 81). The primary pathology was OA in 19 patients
and RA in two. None had previous knee surgery. Before
arthroplasty, there were varus deformities in 20 knees. In
five, the implants were uncemented. The mean active flex-
ion before arthroplasty was 84.7° (60 to 110). The physio-
therapist continued to treat these patients.
Manipulation technique. Before manipulation, the ROM
of the knee and its resistance to further flexion were
assessed under general anaesthesia. The surgeon supported
the lower leg proximally and distally and an assistant
supported the thigh. A steady progressive force was applied
to the proximal tibia until the adhesions gave way. After
this, all the knees could be flexed, under anaesthesia, to at
least 90°. For 24 hours after MUA a CPM machine was
used. Patients were discharged when they had made sat-
Statistical analysis. The Wilcoxon signed-rank test was
used to determine whether the recorded changes in flexion
were significant. We used a paired t-test to compare chan-
ges in flexion within each group and a two-sample t-test to
compare the groups. The 95% confidence intervals (CI) are
Group 1. The mean improvement in passive flexion at
MUA was 34° (10 to 65, CI 30 to 37.5). In the week after
MUA, active flexion equalled or exceeded this in 31 of the
47 knees. By the following year, 33 patients had gained
flexion, ten had lost flexion and four showed no change.
The estimated median change for this period was 4° (-26 to
+37, CI 0 to 7.0). One year after MUA the mean increase in
active flexion was 33° (p < 0.01; CI 28.5 to 38.7) and the
mean active flexion was 95.6° (median 90, range 70 to
130). Three knees flexed only 10° more than they did
before MUA, and one showed no gain. There was no
appreciable change in flexion in the second year after
MUA. Ten patients were unable to flex to 90°. Four of
these had a late MUA (at 16, 17, 20 and 25 weeks). In three
of the ten, flexion before TKR had been less than 80°. Table
I and Figure 1 show the results.
In eight patients MUA was performed more than 16
weeks (17 to 41) after arthroplasty. Flexion improved in
seven who had a mean gain of 24° in active flexion (0 to
33; CI 13.5 to 33.0). One year later, five of the eight knees
were able to flex beyond 90° (mean 87, range 70 to 105).
Despite the delay before MUA, the gain in flexion at one
year was significant (p < 0.01; CI 8.4 to 31.4).
Whether the components had been cemented or not made
no difference to the gain in flexion (p = 0.35; CI -4 to 11),
or in retention of the increase gained (p = 0.84, CI -8.9 to
There were no complications as a result of the
Group 2. The mean active knee flexion was 73.6° (50 to
80) six weeks after TKR. There was no significant change
in flexion (p = 0.23; CI -8.1 to 2.0) between six weeks and
one year when the mean active flexion was 76.7° (62 to
90). Between one and two years, knee flexion did not
change appreciably (p = 0.37; CI -7.2 to 2.7). As shown in
Table II the mean at two years was 79° (55 to 90). At that
28 C. N. A. ESLER,K. LOCK, W. M. HARPER,P. J. GREGG
THE JOURNAL OF BONE AND JOINT SURGERY
Flexion (degrees) in patients in group 1 who had MUA
after MUAAt MUA
35 to 80
75 to 120
77 to 110
70 to 130
70 to 137
60 to 135
Flexion (degrees) in patients in group 2 who did not have
arthroplasty6 weeks1 year2 years
60 to 110
50 to 80
62 to 90
55 to 90
Maximum active knee flexion (degrees) related to the time after TKR. The
error bars show the SEM.
time 14 knees flexed less than 80° and only four flexed
more than 90°. Thirteen patients in this group had less
flexion one year after TKR than they had before (mean loss
21°; range 7 to 37).
Comparison of groups 1 and 2. The group of patients who
declined MUA had significantly worse flexion before TKR
(p < 0.01; CI -26 to -8.1). At the time at which MUA was
considered, however, flexion was significantly worse in the
group of patients who consented (p < 0.01; CI -16.5 to
-6.5). At one and two years there was significantly greater
flexion in the manipulated than in the non-manipulated
group (at one year p < 0.01, CI 13.7 to 24.0; at two years
p < 0.01, CI 12.1 to 23.5). Figure 1 shows the results for
There are few reports about manipulation after TKR. In a
series of 343 TKRs, Fox and Poss5reported the results of
manipulation of 23% of predominantly (70%) rheumatoid
patients with a mean age of 63 years. Manipulation was
performed two weeks after surgery if the patient had less
than 90° of flexion. They found that manipulation was more
likely to be necessary if patients were female, osteoarthritic
or over 70 years of age. At MUA, they achieved a mean
gain in flexion of 37°, but one week later only 17° of this
remained. After one year, the mean gain in flexion was only
13°. A retrospective study of 17 patients who had bilateral
TKRs but MUA in only one knee led them to conclude that
the procedure did not increase the ultimate range of flexion
Since our patients were not selected at random, group 2
cannot be considered to be a control group. Arthroplasty
resulted in a reduction of flexion in these patients. Having
had poor flexion before surgery, they may simply have been
willing to accept a restricted ROM in order to gain freedom
from pain. Alternatively, their poor results may reflect their
levels of motivation. Without carrying out a prospective
randomised trial, we cannot be sure that improvement
would not have occurred without manipulation.
Several studies have shown that the use of a CPM
machine immediately after arthroplasty does not affect the
ultimate ROM but does increase the rate at which flexion is
regained.7-9This accelerated return of flexion may reduce
the need for manipulation.7,10To minimise the risk of
hypoxia of the wound edge and subsequent infection, John-
son11and Goletz and Henry12recommend limiting CPM to
less than 40° in the early postoperative period.
Our primary aim at MUA was to overcome intra-articu-
lar adhesions. We stopped manipulating as soon as we felt
that we had achieved this. We believe that manipulation
beyond this point could increase the risk of the complica-
tions which have given MUA a poor reputation.
Daluga et al13found no difference in the results of MUA
carried out up to 21 days after TKR, or at between 22 and
90 days. Because definite and sustained gains in flexion can
be achieved even when MUA is performed several months
after TKR, there is no reason to manipulate when the
wound is at its weakest.
We found that one year after MUA flexion in 15 of 47
knees had increased beyond that achieved then. After TKR,
MUA can lead to sustained increases in flexion in knees
that are no longer improving, despite physiotherapy. We
believe, however, that the improvements in flexion are
unlikely to be sustained unless patients have access to
adequate physiotherapy afterwards.
No benefits in any form have been received or will be received from a
commercial party related directly or indirectly to the subject of this
1. Laubenthal KN, Smidt GL, Kettelkamp DB. A quantitative analysis
of knee motion during activities of daily living. Phys Ther 1972;52:
2. Ritter MA, Stringer EA. Predictive range of motion after total knee
replacement. Clin Orthop 1979;143:115-9.
3. Schurman DJ, Parker JN, Ornstein D. Total condylar knee replace-
ment: a study of factors influencing range of motion as late as two
years after arthroplasty. J Bone Joint Surg [Am] 1985;67-A:1006-14.
4. Parsley BS, Engh GA, Dwyer KA. Preoperative flexion: does it
influence postoperative flexion after posterior-cruciate-retaining total
knee arthroplasty? Clin Orthop 1992;275:204-10.
5. Fox JL, Poss R. The role of manipulation following total knee
replacement. J Bone Joint Surg [Am] 1981;63-A:357-62.
6. Insall J, Scott WN, Ranawat CS. The total condylar knee prosthesis:
a report of two hundred and twenty cases. J Bone Joint Surg [Am]
7. Romness DW, Rand JA. The role of continuous passive motion
following total knee replacement. Clin Orthop 1988;266:34-7.
8. Colwell CW Jr, Morris BA. The influence of continuous passive
motion on the results of total knee arthroplasty. Clin Orthop 1992;
9. Pope RO, Corcoran S, McCaul K, Howie DW. Continuous passive
motion after primary total knee arthroplasty: does it offer any benefits?
J Bone Joint Surg [Br] 1997;79-B:914-7.
10. Coutts RD, Sharp DN, Borden LS, et al. The effect of continuous
passive motion on total knee rehabilitation. Orthop Trans 1983;7:
11. Johnson DP. The effect of continuous passive motion on wound-
healing and joint mobility after knee arthroplasty. J Bone Joint Surg
12. Goletz TH, Henry JH. Continuous passive motion after total knee
arthroplasty. South Med J 1986;79:1116-20.
13. Daluga D, Lombardi AV Jr, Mallory TH, Vaughan BK. Knee
manipulation following total knee arthroplasty. J Arthroplasty 1991;
29MANIPULATION OF TOTAL KNEE REPLACEMENTS
VOL. 81-B, NO. 1, JANUARY 1999