A correlation study of objective functionality and

Knee Surg, Sports Traumatol, Arthrosc
(2002) 10 : 347–351
KNEE
DOI 10.1007/s00167-002-0302-2
E. Witvrouw
J. Victor
J. Bellemans
B. Rock
R. Van Lummel
R. van der Slikke
R. Verdonk
Received: 11 December 2001
Accepted: 25 April 2002
Published online: 22 June 2002
© Springer-Verlag 2002
E. Witvrouw (✉)
Department of Rehabilitation Sciences
and Physiotherapy, University Hospital,
6K3, De Pintelaan 185,
9000 Ghent, Belgium
e-mail: erik.witvrouw@rug.ac.be,
Tel.: +32-9-2402609, Fax: +32-9-2403811
E. Witvrouw · R. Verdonk
Department of Orthopedic Surgery
and Physical Medicine, Faculty of Medicine,
Ghent University, Ghent, Belgium
J. Victor · B. Rock
Department of Orthopedic Surgery,
Sint Lucas Hospital, Bruges, Belgium
J. Bellemans
Department of Orthopedic Surgery,
Catholic University of Leuven,
Leuven, Belgium
R. Van Lummel · R. van der Slikke
McRoberts, The Hague, The Netherlands
A correlation study
of objective functionality and WOMAC
in total knee arthroplasty
Abstract Function studies after total knee arthroplasty (TKA) generally use either self-reported instruments or objective quantitative measurements of walking activities, but
it is unclear how objective factors influence self-reports. This retrospective multicenter study assessed the
impact of objective variables on subjective self-reported functional outcome after TKA in 128 patients at a
mean of 41 months after TKA. Selfreported disability was measured by
the Western Ontario and McMaster
University (WOMAC) scale, and the
quality of walking and representative
daily activities were determined objectively by the DynaPort knee test.
Objective assessment was found to
be a poor predictor of WOMAC abnormality, with correlations of 0.343,
0.246, and 0.269 with WOMAC
function, stiffness, and pain, respectively. The results of WOMAC are
Introduction
Total knee arthroplasty (TKA) is performed as an endstage procedure for patients with degenerative arthritis to
eliminate pain and improve function [5, 8]. The surgical
techniques and prosthetic engineering developments associated with these elective procedures have been highly refined over the past 20 years and have resulted in high success and survival rates [4, 23]. During the past decade the
technology and design of joint implant systems have progressed considerably, and a great variety of systems are
now on the market. From the patient’s perspective the
goals of knee joint replacement are pain relief and restora-
thus strongly influenced by factors
other than quantitative parameters of
the lower extremity. These findings
provide support for including an objective assessment of activities of
daily living together with a subjective self-report.
Keywords Total knee arthroplasty ·
Western Ontario and McMaster
University scale · Objective
measurement · Functional outcome ·
Dynaport knee test
tion of function and mobility [5]. The surgeon has the
same objectives, and additionally aims to improve range
of motion, stability, walking distance, and stair climbing.
Most of these variables are addressed in the various knee
scoring scales [7, 8]. Therefore most previous studies of
function after TKA have used these self-report instruments to obtain an understanding of the patients’ functional ability. However, the results of these self-report instruments have recently been shown to reflect psychological and constitutional status as well as regional lower
limb abnormalities [22].
To our knowledge, no study has examined the extent to
which objective locomotive factors influence the data obtained by these self-report instruments. Therefore the goal
348
of this study was to assess the correlation between quantitative performance with the results of self-report instruments.
Materials and methods
This retrospective multicenter study examined the results of 128
total knee arthroplasties (94 men, 34 women). Only patients with a
unilateral total knee prosthesis, implanted for primary osteoarthritis at least 1 year previously were included in this study. Patients
with posttraumatic, postinfectious, or rheumatoid arthritis were excluded. All patients were evaluated by the same independent observer using the Western Ontario and McMaster Universities
(WOMAC) function subscale OA index, and the DynaPort knee
test. This WOMAC index is a well-validated scale [1] designed to
reflect the problems experienced by individuals with lower limb
deficiencies. It assesses pain (five items), stiffness (two items), and
physical function (17 items) related to the hip or knee, with higher
scores indicating greater difficulty.
The DynaPort knee test is a short standardized test to assess the
quality of movement of patients with knee dysfunction. It is suitable for clinical applications, and takes only 15–20 min to administer. No further equipment or specialized laboratory is required.
Scores reflect the level of disability in daily living; a higher score
indicates a better function. The test uses sensors which are fixed to
the patient’s body, while the patient performs predetermined tasks
(Fig. 1). The patient is asked to perform 14 selected activities, using 29 easy-to-measure test items that are often problematic for the
patients. These tasks can be categorized under locomotion (walking), rising and descending (stairs, slopes, and wooden blocks),
lifting and moving (carrying a tray or a bag, picking up a weight,
and walking with a shopping trolley), and transfers (going to sit or
lie down and then standing up again, as well as bending forwards
to pick up a weight and returning to the upright position). To standardize the test a standard package with all equipment is included
in the test material (such as wooden blocks, stairs of three steps,
Fig. 1 The Dynaport knee test
measuring device
and a slope). The sensors are sensitive to both slow and fast
changes in acceleration; they are small and of low weight. The sensors react to gravity and movement acceleration, thus giving information about posture and movement of the test person [21].
A portable recorder worn around the waist allows easy and stable
data acquisition. The whole system is fixed to the patient with elastic straps over his daily clothing (Fig. 1). In total six sensors are
fixed to the chest, the pelvis (two), the left thigh, and the left and
right shanks. The raw signals are stored in a portable recorder and
are analyzed with special software which extracts movement features. All signal properties are derived from the original accelerometer signals, while several movement features are calculated in terms of angle, duration (time), frequency, or a dimensionless number [6].
Before the Dynaport knee test was used in the present investigation it was deemed necessary to obtain preliminary information
on its reliability and validity (responsiveness). In a test-retest study
with 37 healthy subjects, measured twice at an interval of 1 week
the reliability of the device was found to be high, with an overall
intraclass correlation of 0.81 (95% confidence interval 0.69–0.93)
and values for the cluster ranging from 0.73 (95% CI 0.58–0.89) to
0.84 (95% CI 0.73–0.94). To test the validity of the Dynaport knee
test its responsiveness was tested on 244 TKA patients. These patients were tested before the procedure and 3, 6, 12, and 24 months
thereafter. This study [6] showed a clear trend to improvement,
with an average extent of effect after 24 months (mean improvement) of 0.7 (when expressed in preoperative SD of TKA patients)
or 1.5 (SD of the healthy group). Such responsiveness is satisfactory.
In addition to the reliability and validity (responsiveness) of the
Dynaport Knee Test, it is imperative to know whether it discriminates between injured and uninjured persons. In a previous study
[6] 140 persons with knee problems (most of them diagnosed with
osteoarthritis of the knee while waiting for TKA) and 32 matched
persons without knee pain underwent a Dynaport knee test. The results of this study showed that the majority of item X movement
feature combination discriminated significantly between the
healthy subjects and TKA patients. The number of significant differences per item ranged from 10 to 24 (out of 30) and per movement feature from 1 to 29 (out of 29).
The mean follow-up period of the patients in this study was
41 months (SD 22.6). All patients in our study signed an informed
consent statement. All the ethics committees of the various cooperating hospitals approved this study.
Data were analyzed using the SPSS software version 10.0. Correlations between the different assessment scores were evaluated
in bivariate analyses with Spearman’s rank order correlation. To
explore multivariate relationships we performed multiple regression with WOMAC function, pain, and stiffness as the dependent
variables and Dynaport score as the independent variable. Statistical significance was accepted at the level of P<0.05.
Results
Mean and standard deviations of the anthropometric variables of the 128 patients are given in Table 1. Table 2 illustrates the results of the Dynaport knee test and the
WOMAC score. The relationship between the Dynaport
knee test and the WOMAC scores (Fig. 2) shows that the
two are inversely: the higher the Dynaport score, the less
the disability. However, the correlation between Dynaport
score and WOMAC pain score was poor (r=0.359; Fig. 2).
Similarly, there was a poor correlation between Dynaport
score and the results of the WOMAC function and stiffness scores (r=0.342 and r=0.216, respectively). The R2
349
Table 1 Anthropometric data on the patient group (n=128)
Age (years)
Weight (kg)
Height (cm)
Body mass index
Mean ±SD
Range
62.5±6.3
84.0±12.4
169.3±7.6
29.3±4.0
41–77
58–120
148–191
20.8–40.1
Table 2 Dynaport knee test and WOMAC scores for pain, function, and stiffness (n=128)
Dynaport
WOMAC pain
WOMAC function
WOMAC stiffness
Mean ±SD
Range
53.6±10.8
7.2±8.2
8.3±8.5
10.8±12.9
22.8–79.7
64–100
48–100
30–100
Fig. 2 Correlation between WOMAC pain score (100=max) and
the result of the Dynaport knee test (100=max; n=128, r=0.359)
with WOMAC pain, function, and stiffness scores as dependent variables was 0.14, 0.12, and 0.05, respectively.
These data indicate that during daily living in bivariate
analyses WOMAC scores are poorly associated with
quantitative measures of activities. To identify the impact
of the three WOMAC scores on the Dynaport score, a
multiple regression analysis was performed with the Dynaport score as the independent variable and with
WOMAC function, pain, and stiffness scores as the dependent variables. Statistical analysis failed to identify
any of the three WOMAC scores as important predictors
of the Dynaport score (P>0.05).
Discussion
The literature indicates that TKA is highly successful, and
that it contributes significantly to improving an individual’s functional status and well-being, with a considerable
individual variation being shown in treatment response
[13, 16, 19]. Therefore it is important to understand clearly
the factors that contribute to the functional outcome after
TKA [20]. This study showed that the quantitative variables measured by the Dynaport knee test are poorly correlated with self-reported disability as obtained by the
WOMAC questionnaire. The correlation coefficient r measures the strength of the linear association between two
variables. An interpretation of this strength is found by
the R2 statistic, which represents the proportion of the
variance in one of the variables that can be explained by
the values of the other. The data found in this study show
that the quantitative measurement accounts for only
5–14% of the subjective assessment. This demonstrates
the variability and discrepancy between these two evaluation methods.
Our results are in accordance with those of Lamb et al.
[9] who found that pain was associated with limited mobility (obtained by measuring ability to rise from a chair
and walking ability) only when severe. Consistent with
these findings, Madsen et al. [11] found no correlation between walking or stair climbing and pain level in patients
with gonarthrosis. Based on their results these authors
concluded that walking time and stair climbing seem to
have little clinical value in predicting self-reported pain
and pain disability. This may not be very striking since
self-reported instruments appear to indicate more than
merely regional pain [22]. These functional self-reported
scales are sensitive to psychological and nondisease factors such as fatigue and depression. This can be considered as a merit of these scales since such factors contribute significantly to the patient’s actual pain and dysfunction. Previous research has shown that depression is
associated with and predictive of physical disability [18]
and is strongly related to pain [12].
On the basis of these findings conclude that the evaluated outcome measurements in this study (WOMAC and
Dynaport knee test) include numerous dimensions such as
quality of life, general health, and functional impairment,
which reflect the different viewpoints and relative importance to the patient, physician, and society. Regarding the
results of this study, the clear discrepancy between self-report questionnaire and objective functionality suggests
that both assessments should be used to provide the most
complete evaluation. In addition, research has revealed the
importance of using both a generic quality-of-life measuring instrument and a disease-specific instrument [2, 10,
15]. We believe that when evaluating patients with TKA
both a self-report questionnaire and an objective evaluation of function are essential because, as shown in this
study, they provide distinct but complementary aspects.
350
However, our results contrast with those results of previous research [14, 17] which found an association between quadriceps strength and disability and knee pain in
patients with knee OA. Likewise, Boardman et al. [2]
found a sound statistical relationship between walking
ability and the functional aspects of the WOMAC in patients with total hip arthroplasty. The discrepancies between our results and these findings can be explained by
the fact that a different patient population and different
parameters were studied. In our study we did not measure
quadriceps strength, although this has been shown by several authors [3, 11, 14, 15, 17] to be correlated with selfreported disability. This can be considered an important
limitation of this study. However, the strength of this
study is that all objective functional tests were performed
by the same independent person who was blinded to the
results of the self-report questionnaire.
Conclusions
An important principle in evaluating the outcome of a surgical procedure is that tools used to measure the results
must be relevant to the clinical situation. The purpose of
this study was to obtain a better understanding of how the
subjective scores derived from the self-reported instrument (WOMAC) are correlated with objective functional
outcome data measured by the Dynaport knee test. We
found that the relationship between the scores of quality
of movement during daily activities and the WOMAC is
poor. The objective functional measurements predicted
only 5–14% of the WOMAC score. The results of this
study show that patient’ subjective assessment is poorly
related to the extent of movement, and that the result is influenced (85–95%) by many other factors. We believe that
our results support the continued use of a self-reported instrument (WOMAC) combined with an objective measurement of the quality of movement. Since the results
suggest that the outcome measurements evaluated in this
study (WOMAC and Dynaport knee test) cover different
dimensions of the functional outcome, their combined use
provide the clinician with a more complete evaluation of
his patients and allows more appropriate decision for rehabilitation strategy.
Acknowledgements The authors thank the Belgian Society of
Orthopaedics and Traumatology (BVOT) for a research grant, and
the members of the Belgian Knee Group for their cooperation in
this study. In addition we thank A.A. Amis, N. Friederich, and P.
Neyret for their important assistance in this study.
References
1. Bellamy N, Buchanan WW, Goldsmith
CH, Campbell J, Stitt L (1988) Validation study of the WOMAC: a health
status instrument for measuring clinically-important patient relevant outcomes following total hip- or knee
arthroplasty in OA. J Orthop Rheum 1:
95–108
2. Boardman D, Dorey LF, Thomas BJ,
Lieberman JR (2000) The accuracy of
assessing total hip arthroplasty outcomes: a prospective correlation study
of walking ability and 2 validated measurement devices. J Arthroplasty 15:
200–204
3. Bolanos A, Colizza A, McCann WA,
Gotlin PD, Wootten RS, Kahn ME,
Insall, JN (1998) A comparison of isokinetic strength testing and gait analysis in patients with posterior cruciateretaining and substituting knee arthroplasties. J Arthroplasty 13:906–915
4. Calder JD, Ashwood N, Hollingdale JP
(1999) Survivorship analysis of the
“Performance” total knee replacement7-year follow-up. Int Orthop 23:100–
103
5. Creamer P, Lethbridge Cejku M,
Hochberg MC (2000) Factors associated with functional impairment in
symptomatic knee osteoarthritis.
Rheumatology (Oxf) 39:490–496
6. Dikkenberg N van den, Meijer OG,
van der Slikke RMA, van Lummel RC,
van Dieën JH, Pijls B, Bennink RJ,
Wuisman PM (2002) Measuring functional abilities of patients with knee
problems: rationale and construction of
the DynaPort knee test. Knee Surg
Sports Traumatol Arthrosc 10, DOI
http://dx.doi.org/10.1007/s00167-0020279-x
7. Drake BG, Callahan CM, Dittus RS
(1994) Global rating systems used in
assessing knee arthroplasty outcomes.
J Arthroplasty 9:409–412
8. König A, Walther M, Kirschner S,
Gohlke F (2000) Balance sheets of
knee and functional scores 5 years after
knee arthroplasty for osteoarthritis.
J Arthroplasty 15:289–294
9. Lamb SE, Guralnik JM, Buchner DM,
Ferrucci LM, Hochberg MC, Simonsick
EM, Fried LP (2000) Factors that modify the association between knee pain
and mobility limitation in older women:
the Women’s Health and Aging Study.
Ann Rheum Dis 59:331–337
10. Lieberman JR, Shekelle P (1997) Outcome after total hip arthroplasty: comparison of traditional disease-specific
and quality-of-life measurement of outcome. J Arthroplasty 12:639–643
11. Madsen OR, Bliddal H, Egsmose C,
Sylvest J (1995) Isometric and isokinetic quadriceps strength in gonarthrosis; inter-relations between quadriceps
strength, walking ability, radiology,
subchondral bone density and pain.
Clin Rheumatol 14:308–314
12. Magni G, Caldieron C, Rigatti-Luchini
S, Merskey H (1990) Chronic muskuloskeletal pain and depressive symptoms in the general population. An
analysis of the National Health and
Nutrition Examination Survey data.
Pain 43:299–307
13. Marcacci M, Iacono F, Zaffagnini S,
Visani A, Loreti I, Petitto A, Neri MP,
Kon E (1997) Total knee arthroplasty
without patellar resurfacing in active
and overweight patients. Knee Surg
Sports Traumatol Arthrosc 5:258–261
14. McAlindon TE, Cooper C, Kirwan JR,
Dieppe PA (1993) Determinants of disability in osteoarthritis of the knee.
Ann Rheum Dis 52:258–262
351
15. McGuigan FX, Hozack WJ, Moriarty
L, Eng K, Rothman RH (1995) Predicting quality-of-life outcomes following
total joint arthroplasty. Limitations of
the SF-36 Health Status Questionnaire.
J Arthroplasty 10:742–747
16. Murray DW, Frost SJ (1998) Pain in
the assessment of total knee replacement. J Bone Joint Surg Br 80:426–
431
17. O’Reilly S, Jones A, Muir KR, Doherty
M (1998) Quadriceps weakness in knee
osteoarthrosis: the effect on pain and
disability. Ann Rheum Dis 57:588–594
18. Penninx BW, Guralnik JM, Ferrucci L,
Simonsick EM, Deeg DJ, Wallace RB
(1998) Depressive symptoms physical
decline in community-dwelling older
persons. JAMA 279:1720–1726
19. Ritter MA, Albohm MJ, Keating EM,
Faris PM, Meding JB (1995) Comparative outcomes of total joint arthroplasty. J Arthroplasty 10:737–741
20. Steultjens MP, Dekker J, Van Baar E,
Oostendorp R, Bijlsma JW (2000)
Range of motion and disability in patients with osteoarthritis of the knee or
hip. Rheumatology (Oxf) 39:955–961
21. Veltink PH, Bussmann HJ, De Vries
W, Martens WL, van Lummel RC
(1996) Detection of static and dynamic
activities using uniaxial accelerometers. IEEE Trans Rehabil Eng 4:375–
385
22. Wolfe F (1999) Determinants of
WOMAC function, pain and stiffness
scores: evidence for the role of low
back pain, symptoms counts, fatigue
and depression in OA, rheumatoid
arthritis and fibromyalgia. Rheumatology (Oxf) 38:355–361
23. Zahiri CA, Schmalzried TP,
Szuszczewicz ES, Amstutz HC (1998)
Assessing activity in joint replacement
patients. J Arthroplasty 13:890–895