Functional improvement after unicompartmental knee replacement: a follow-up study with a performance based knee test

Transcription

1 Knee Surg Sports Traumatol Arthrosc (2007) 15: DOI /s KNEE Functional improvement after unicompartmental knee replacement: a follow-up study with a performance based knee test Lucas L. A. Kleijn Æ Wouter L. W. van Hemert Æ Will G. H. Meijers Æ Arnold D. M. Kester Æ Lukas Lisowski Æ Bernd Grimm Æ Ide C. Heyligers Received: 20 January 2006 / Accepted: 12 April 2007 / Published online: 23 June 2007 Ó Springer-Verlag 2007 Abstract In literature no reports appear on functional recovery of unicompartmental knee replacement using both subjective and objective methods. Functional aspects are especially of importance, since prosthetic replacement is considered more often in younger patients, who require an optimal knee function for activities such as sports. In a prospective study on 38 patients with a mean age of 62.2 years functional improvement was measured. Using Knee Society Score (KSS) as a clinician based score and the Dynaport Ò Knee Test as a functional test measurements were performed at preoperative, 3 and 6 months, 1 and 2 years, after surgery. The Dynaport Ò Knee Test is an accelerometer-based system that objectively measures functional aspects of gait during various tasks of daily life. It consists of four sub scores. The KSS assesses pain and function. Both scores range from 0 to 100. The mean KSS preoperative was 44.0 and improved significantly to 81.7 at 3 months (P < 0.001) and to 87.4 (P = 0.025) at 6 months. No significant differences were noted after 6 months. The mean preoperative Dynaport Ò Knee Test score was 35.8 and at 3 months 43.6 (P < 0.001), 48.6 at 6 months L. L. A. Kleijn W. L. W. van Hemert (&) W. G. H. Meijers B. Grimm I. C. Heyligers Department of Orthopaedic Surgery and Traumatology, Atrium Medical Centre, PO Box 4446, 6401 CX Heerlen, The Netherlands wvanhemert@planet.nl A. D. M. Kester Department of Methodology and Statistics, Maastricht University, Maastricht, The Netherlands L. Lisowski Department of Orthopaedic Surgery, Amsterdam Medical Centre, Amsterdam, The Netherlands (P < 0.001). No significant differences were noted after 6 months follow-up. Of the Dynaport sub scores, the low demanding tasks Lift and Move and Locomotion, cease to improve beyond 6 months. The high demanding task Transfers only improved up to 3 months. However, the other high demanding tasks Rise and Descend showed improvement beyond 1 year after surgery, since the improvement from 6 months to 2 years was significant (P = 0.023). This study has found that functional recovery continues beyond 6 months and even up to 2 years. It seems only more challenging tests can discriminate on improvement beyond a point where questionnaires cease to improve. The use of objective measurement methods is advocated next to the clinician based scores and self reported questionnaires. Keywords Unicompartimental knee replacement Knee osteoarthritis Knee test Knee function Rehabilitation Introduction Various surgical options are available for treating medial compartmental osteoarthritis of the knee, such as high tibial osteotomy, unicompartmental knee replacement, and total knee arthroplasty. Regarding prosthetic replacement, unicompartmental knee replacement can be considered an alternative to total knee replacement since it preserves bone stock and the procedure is less invasive. It also approaches physiologic knee better than does total knee arthroplasty, since it preserves the cruciate ligaments and the considered healthy remaining two compartments [3]. In young patients, unicompartmental knee replacement for medial osteoarthritis has proven to be successful both on short

2 1188 Knee Surg Sports Traumatol Arthrosc (2007) 15: term as well as on long term follow-up [12] and it has been reported to be more cost-effective over total knee replacement in the same kind of patients [19]. Clinical outcome after knee replacement is commonly assessed by function scores such as the clinician-based American Knee Society Score (KSS) [4]. These methods are accepted, routinely used and proven valid [6]. Nevertheless, Ryd et al. [20] question their reliability and claim large bias of subjective knee scores after knee arthroplasty. They report on substantial inter and intra observer reliability and therefore potential under- and over performance of these scores is likely. As unicompartmental knee replacement tends to be performed at a younger age functional aspects of knee replacement gain more attention. Therefore, next to such knee scores, outcome and function should be measured using more objective methods. Only a few reports are available concerning true objective measurements of function after unicompartmental knee replacement. Webster et al. [24] report on 13 patients between 12.2 and 33.3 months after surgery operated on for medial osteoarthritis with a cementless unicompartmental knee system. The authors performed a one-time measurement using a 3D motion analysis system and electronic walkway and all but one patient showed a gait pattern not significantly differing from healthy controls. Mattsson et al. [9] studied 20 patients before and 1 year after knee replacement with a cemented unicompartmental prosthesis and provided data on walking speeds. They found that patients could increase their walking speed by 28% after surgery compared to preoperative. Weidenhielm et al. [25] reported kinematic data after unicompartmental knee replacement using electrogoniometers. In 26 of 36 patients they found improvement in stance knee flexionextension patterns, although this improvement has not been supported by range of motion measurements. Pre- and postoperative comparisons between stance and swing phases were not significantly different. No information on the course of recovery within 2 years is available. Current objective measurement systems such as electromyography, force platforms, optokinematic systems and 3D motion analysis are available and considered as the gold standard, but are time consuming and require sophisticated laboratories. In view of this problem an accelerometer based, user-friendly system was developed, the Dynaport Ò Knee Test [22]. It objectively assesses functional abilities in a standardized set of tasks, closely related to activities of daily living (ADL). The rationale of the system has been explained, indicating the potential value in evaluating patients functional abilities in kneerelated clinical practice and research [22, 26]. Regarding the post operative functional improvement after unicompartmental knee using a performance based knee test on multiple time points, literature lacks papers reporting on this issue. The goal of this study was to objectify functional recovery after unicompartmental knee replacement using both subjective and objective methods. Materials and methods Study design Between February 2000 and August 2001, 40 patients (41 knees) were operated for medial osteoarthritis of the knee. When eligible according to the Oxford guidelines, a primary Phase 3 Oxford TM Unicompartimental Knee System (Biomet, Dordrecht, The Netherlands) was placed [3]. All patients were operated in a joint care program. When screened eligible for surgery informed consent was acquired and preoperative knee flexion, KSS and Dynaport Ò Knee Test scores were measured within 2 weeks before surgery. After surgery a standard rehabilitation program was followed, supervised by a physical therapist, allowing mobilization and active knee exercises the first day after surgery. Patients were discharged from the hospital within 1 week. The majority followed an outpatient rehabilitation program for 3 months. The time points of follow-up were pre operative, 3 and 6 months, 1 and 2 years after surgery. Two orthopedic surgeons examined the patients at the outpatient clinic and the KSS and knee function was scored. A physical therapist operated the Dynaport Ò Knee Test. Instruments of measurement The KSS score is subdivided into a Knee score that scores the knee joint only and a Function score that rates pain, patient s ability to walk and climb stairs and the need for a support device [4]. It has been proven reliable as described by Kreibich et al. [6]. During the Dynaport Ò Knee Test patients perform various tests related to Activities of Daily Living (ADL), such as walking, stair climbing, getting up and moving objects [22]. The test lasts approximately min and is supervised by a physical therapist. Patients are instructed to perform the test items at their own pace. An item can be skipped if considered too difficult, but then a score of zero is given. The Dynaport Ò Knee Test consists of five small movement sensors that are fixed to the patient s thorax, pelvis, left thigh and beneath both knees. These sensors measure the accelerations related to the orientation and movement patterns of the body and the trunk. Data is captured by a portable recorder worn around the waist as the patient performs a set of 29 test items, such as walking up and down stairs and sitting down. An algorithm programmed

3 Knee Surg Sports Traumatol Arthrosc (2007) 15: by the manufacturer calculates four cluster scores, Locomotion, Transfer, Lift and Move, Rise and Descend, and automatically relates the cluster scores to a control of healthy subjects. The scores are weighed and combined into one overall Dynaport Ò Knee Test score that ranges from 0 to 100. Mokkink et al. [11] investigated the reproducibility and validity of the Dynaport Ò Knee Test on 92 total knee arthroplasty patients who performed the test twice on the same day and 94 healthy controls performed the Knee Test once. The inter- and intra-observer reliability was found to be excellent. Construct validity was expressed in a correlation coefficient with WOMAC physical functioning (0.55), SF-36 physical functioning (0.62) and Knee Society Score (KSS) function (0.64) [11]. Statistics Data analysis was performed using Statistical Package for the Social Science (SPSS) 12.0 (Chicago, IL, USA). Reported values are mean ± standard deviation (SD). Differences in KSS and knee flexion were compared using the Friedman test, the equivalent of a repeated measures ANOVA for nonparametric data. Post hoc comparisons were checked using Wilcoxon Signed Ranks test. Nonparametric tests were used because of non-normality of these scores in the tested population. Differences in the Dynaport Ò Knee Test score and the subscores were compared using repeated measures ANOVA. Post hoc comparisons were checked using least-significant difference (LSD). Level of significance was set on P < Results Two patients (three knees) were excluded from data analysis. One patient had both knees operated on at two different time points, but measurements were recorded for both knees at the same time. Another patient had incomplete scores and was considered a loss to follow-up and therefore excluded from data analysis. Thirty-eight patients (38 knees) were followed up and included in data analysis. All patients completed the 2 year follow-up, although only 23 of the 38 patients performed the Dynaport Ò Knee Test at 2 years, whereas, all 38 patients were examined with the KSS. Eight males and 30 females were analyzed in the study, with a mean age at time of surgery of 62.2 ± 9.5 years (range 41 78). The mean Body Mass Index was 28.5 ± 3.6 (range 23 39). patients preoperative was 44.0 ± 9.6 and 81.7 ± 13.5 at 3 months follow-up. Post hoc comparisons showed that this was a steep improvement and an expected significant difference (P < 0.001). The patients seen at 6 months scored 87.4 ± 9.8, a less steep but significant improvement compared to 3 months follow-up (P = 0.025). The mean KSS decreased slightly to 86.1 ± 12.8, (P = 0.55) at 1 year and leveled at 86.3 ± 12.6 at 2-year follow-up (P = 0.83). Also the improvement from 6 months to 2 years was not found to be significant (P = 0.90). The time course of the KSS is displayed in Fig. 1. The breakdown of the KSS into function- and knee score is shown in Table 1. Knee flexion (Table 1) showed significant differences among the measured time points (P < 0.001). The mean flexion of all patients preoperative was ± 14.3 degrees. Post hoc comparisons showed that the improvement to ± 12.1 degrees at 3 months was significant (P = 0.025). Of all patients reviewed at 6 months the mean flexion was ± 11.1 degrees, again a mild and significant improvement (P = 0.032) compared to 3 months. No significant differences were noted after 6 months. The knee flexion leveled at 1 year at ± 11.7 (P = 0.835) and ± 12.1 (P = 0.23) at 2 years. The Dynaport Ò Knee Test Significant differences could be found in the Dynaport Ò Knee Test score among the measured time points. The mean preoperative Dynaport Ò Knee Test score was 35.8 ± Patients reviewed at 3 months follow-up scored a mean of 43.6 ± 15.3, which was a significant improvement compared to preoperative (P < 0.001). It continued to improve significantly to 48.6 ± 15.5 (P < 0.001) at 6 months. The Dynaport Ò Knee Test score at 1 year reached to 50.5 ± 14.0, which was not a significant improvement (P = 0.311). At 2 year follow-up the studied patients scored 52.3 ± 16.6, which was not a significant improvement compared to 1 year (P = 0.366). The KSS and knee flexion Significant differences could be found in KSS among the measured time points (P < 0.001). The mean KSS of all Fig. 1 The knee society score (mean ± SD) in time. The result that is marked asterisk is a significant improvement compared to the previous time point

4 1190 Knee Surg Sports Traumatol Arthrosc (2007) 15: Table 1 Knee flexion and breakdown of KSS Time point Knee flexion (mean ± SD, range) Knee score (mean ± SD) Function score (mean ± SD) Pre operative ± 14.3 (90 135) 41.7 ± ± 12.1 Three months ± 12.1 (90 135)* 83.9 ± 12.2* 79.1 ± 16.2* Six months ± 11.1 ( ) 86.8 ± ± 12.3 One year ± 11.7 ( ) 83.3 ± ± 15.6 Two years ± 12.1 ( ) 83.7 ± ± 14.0 * Significant improvements compared to previous time points are marked results and time course of the Dynaport Ò Knee Test score and sub scores are displayed in Fig. 2. Significant differences could be found in the cluster score Lift and Move among the measured time points. The mean pre operative score was 37.4 ± Post hoc comparisons showed a significant improvement to 45.9 ± 16.3 at 3 months (P < 0.001). It improved significantly to 50.5 ± 15.8 at 6 months (P < 0.001). At 1 year it improved, though not significantly, to 53.0 ± 15.2 (P = 0.178). It increased to 54.1 ± 18.1 at 2 years, but not significantly so (P = 0.620). No significant differences were noted after 6 months. Significant differences could be found in the cluster score Locomotion among the measured time points. The mean score pre operative was 37.4 ± It improved significantly to 45.7 ± 15.4 at 3 months (P < 0.001) and to 52.5 ± 15.1 at 6 months (P = 0.002). Neither the score of 54.1 ± 13.4 was a significant improvement at 1 year (P = 0.402), nor was the score of 55.1 ± 16.5 at 2 years (P = 0.702). No significant differences were noted after 6 months. The mean cluster score Transfers pre operative was 28.9 ± It improved significantly to 37.7 ± 15.1 at 3 months (P < 0.001) and to 42.7 ± 18.2 at 6 months, but not significantly so (P = 0.122). It improved not significantly to 45.1 ± 16.1 at 1 year, (P = 0.464) and to 45.7 ± 18.2 at 2 years (P = 0.75). Rise and Descend showed significant differences among the measured time points. The mean score pre operative was 32.3 ± Post hoc comparisons showed a significant improvement to 41.4 at 3 months (P = 0.002) and the score improved significantly to 48.8 ± 17.1 at 6 months (P = 0.006). The improvement to 50.6 ± 15.8 at 1 year was not significant (P = 0.587), as well as the improvement from 1 year to 54.5 ± 16.9 at 2 years (P = 0.71). However, when considering the improvement from 6 months to 2 years the found improvement was significant (P = 0.023). Discussion To our knowledge this is the first study to report on repetitive performance based knee test measurements using both the Dynaport Ò Knee Test and KSS during a 2 year Fig. 2 The Dynaport Ò Knee Test and its subscores in time (mean ± SD). The result that is marked asterisk is a significant improvement compared to the previous time point

5 Knee Surg Sports Traumatol Arthrosc (2007) 15: follow-up of patients after unicompartmental knee replacement. We were able to measure further improvement beyond 6 months according to the Dynaport Ò Knee Test. Looking closely at the subscores it can be seen that this improvement was less in the considered low demanding tasks, but more substantial in the considered high demanding tasks. The cluster score Rise and Descend can be considered as the most difficult task. It resembles climbing stairs of different levels. Active extension during stair climbing is dependent on both the amount on quadriceps contracture as the amount of knee flexion. At this point and the patellofemoral contact forces are increased and the patello-femoral lever arm is the largest at 45 of flexion [2], which typically occurs in the Rise and Descend task. The quadriceps muscles are already significantly weakened due to osteoarthritis, but needed substantially during recovery [7, 10]. It may require a longer time for the quadriceps muscles to regain strength and this could explain why a larger interval was needed for significance than that between 1 and 2 years. The task Locomotion resembles walking distances of different lengths and this score levels at 6 months. Walking on a flat surface can be considered a low demanding task that recovers quickly after surgery as expected. Lift and Move combines scores that are acquired during lifting certain objects from different heights while moving it elsewhere. This is expected to be a more demanding task, since bending in order to reach objects is also done from the hip and spine. The cluster score Transfers contain items such as getting up and down. It can be considered a difficult task, since patients have a considerable influence of their body weight and impairment of other joints. This could explain the repeated lower scores compared to the other three sub scores. Shortly after surgery the patient rather than the knee itself is influenced and therefore it can be expected that this task is completed with more effort. It shows similar improvement compared to the other three subscores, because of the expected improvement due to surgery. It can be expected that the continuing lower results compared to the other three subscores can be attributed to its lower starting value. In our study the pre- and post-operative KSS was similar to the KSS of patients described in literature [5, 13, 18]. Therefore it can be expected that the study was conducted on a representative group of patients and the acquired data with The Dynaport Ò Knee Test can be considered representative. In this study the KSS showed its usual, very steep, improvement in patients after knee replacement [5, 13, 18], but contrary to the Dynaport Ò Knee Test, it ceased to improve after 6 months and even decreased slightly beyond. It can be debated why such differences between KSS and the Dynaport Ò Knee Test have been found. The KSS does not account for knee flexion beyond the maximum value of 120. This creates a premature ceiling effect in the results curve, since patients individually in our study and other studies can sometimes reach knee flexion beyond this limit (Table 1). Thus in our study the KSS seems to underrate the performance of a large number of functionally well performing patients. However, even if a modified KSS score was adjusted to account for higher flexion, it would only have resulted in a higher ceiling in the curve, but not in an increased steepness of recovery. Furthermore it can be discussed whether the found improvement with the Dynaport Ò Knee Test is clinically relevant. The younger, more active and demanding patient, sometimes referred to as the millennium patient, in modern orthopaedics, is more active, more demanding and values functional improvement more than clinician based scores were developed for. Besides pain relief, he demands a high knee flexion and expects post-operative improvements to be long lasting. Modern types of prostheses, such as unicompartmental knee replacements, can provide such results. The patients studied in this paper had an average age of 62.2 years and belong to this more demanding patient group. The acquired results indicate that physiological changes in recovery and functional improvement still take place beyond a time point where KSS does not register it in our study. Thus the continuing functional improvement measured with The Dynaport Ò Knee Test may be beneficial to these subjects. Lastly, the found differences can be attributed to measuring different aspects of recovery. Maly et al. [8] has shown that clinician based scores, but more self-report measures are strongly related to pain, whereas physical performance measures are strongly related to self-efficacy. In this paper the continuing functional improvement as measured by the objectiveness of The Dynaport Ò Knee differs from the more subjective, pain related KSS. Several papers have appeared using different assessment scores in order to quantify recovery, but track outcome rather than short-term post-operative functional improvement. Pennington et al. [15] report on functional aspect after unicompartmental knee replacement using The University of California at Los Angeles activity assessment score and evaluate long-term survival after unicompartmental knee replacement. However, they do not address any differences between activity levels in recovery, since a pre operative score lacked. To evaluate prosthetic knee replacement Weale et al. [23] use the Oxford questionnaire, stating that this introduces more objectiveness, yet it still does not reflect true objectiveness, since potential bias of the clinician or patient can be of influence. Schai et al. [21] report on follow-up after unicompartmental knee replacement using Tegner and Lysholm score, but could not provide results within 2 years due to limited time points

6 1192 Knee Surg Sports Traumatol Arthrosc (2007) 15: included. Many authors have tried to convey more responsiveness over clinician-based scores, since knee scores are claimed to be exceedingly unreliable [20]. Whereas patient perceptions after knee replacement become increasingly important, self-reported questionnaires still include certain subjectiveness. In agreement with Maly et al. [8] and based on our findings we advocate the use of performance based tests next to clinician based scores and self reported questionnaires when evaluating functional recovery during follow-up. Our study holds a few limitations. We were not able to acquire the 2-year follow-up scores of Dynaport Ò Knee Test of all patients. Although these patients did visit the outpatient where KSS could be scored, apparently some considered the Dynaport Ò Knee Test still too time consuming. Next the number of males in our study is skewed, which can be attributed to the consecutive enrollment in this study. A difference may exist in performance between males and females, which may result in a performance bias. Although it was not the aim of this investigation, it would be of interest to compare the results as measured with the Dynaport Ò Knee Test to those of total knee arthroplasty patients. In literature van den Dikkenberg et al. [22] report to have studied 244 patients during the same time scale as in the present study, however, their intention was to use the data for validation of the Dynaport Ò Knee Test. They refrain from detailed information on longevity of recovery of patients. Other reports on The Dynaport Ò Knee Test after total knee arthroplasty patients are not available to our knowledge. Considering post-operative recovery measured with KSS, the curve of recovery shows a similar, although all values for unicompartmental knee replacement are slightly higher [1, 14]. This can be due to a better functioning knee pre operative. Unicompartmental knee replacement in the treatment of medial osteoarthritis of the knee is a successful technique with excellent long-term results [12, 16, 17]. This study provides further insight in the course of functional recovery after unicompartmental knee replacement. It appears to continue beyond 6 months and even up to 2 years. Thus more attention should be paid to this extended recovery period and the use of objective measurement methods are advocated next to the clinician based scores and self reported questionnaires. Acknowledgments No financial benefits were received for this study. The authors thank Alita Hidding, PhD for assistance. References 1. Amin AK, Patton JT, Cook RE, Gaston M, Brenkel IJ (2006) Unicompartmental or total knee arthroplasty? Results from a matched study. Clin Orthop Relat Res 451: Buckwalter JA, Einhorn TA, Simon SR (2000) Orthopaedic basic science, 2nd edn. AAOS, Illinois, USA 3. Goodfellow J, O Connor J (1978) The mechanics of the knee and prosthesis design. J Bone Joint Surg Br 60-B: Insall JN, Dorr LD, Scott RD, Scott WN (1989) Rationale of the knee society clinical rating system. Clin Orthop Relat Res 248: Keys GW, Ul-Abiddin Z, Toh EM (2004) Analysis of first forty Oxford medial unicompartmental knee replacement from a small district hospital in UK. Knee 11: Kreibich DN et al (1996) What is the best way of assessing outcome after total knee replacement? Clin Orthop Relat Res 331: Lewek MD, Rudolph KS, Snyder-Mackler L (2004) Quadriceps femoris muscle weakness and activation failure in patients with symptomatic knee osteoarthritis. J Orthop Res 22: Maly MR, Costigan PA, Olney SJ (2006) Determinants of selfreport outcome measures in people with knee osteoarthritis. Arch Phys Med Rehabil 87: Mattsson E, Olsson E, Brostrom LA (1990) Assessment of walking before and after unicompartmental knee arthroplasty. A comparison of different methods. Scand J Rehabil Med 22: Mizner RL, Stevens JE, Snyder-Mackler L (2003) Voluntary activation and decreased force production of the quadriceps femoris muscle after total knee arthroplasty. Phys Ther 83: Mokkink LB et al (2005) Reproducibility and validity of the Dyna Port Knee Test. Arthritis Rheum 53: Murray DW, Goodfellow JW, O Connor JJ (1998) The Oxford medial unicompartmental arthroplasty: a ten-year survival study. J Bone Joint Surg Br 80: Naudie D, Guerin J, Parker DA, Bourne RB, Rorabeck CH (2004) Medial unicompartmental knee arthroplasty with the Miller Galante prosthesis. J Bone Joint Surg Am 86-A: Newman JH, Ackroyd CE, Shah NA (1998) Unicompartmental or total knee replacement? Five-year results of a prospective, randomised trial of 102 osteoarthritic knees with unicompartmental arthritis. J Bone Joint Surg Br 80: Pennington DW, Swienckowski JJ, Lutes WB, Drake GN (2003) Unicompartmental knee arthroplasty in patients sixty years of age or younger. J Bone Joint Surg Am 85-A: Price AJ, Waite JC, Svard U (2005) Long-term clinical results of the medial Oxford unicompartmental knee arthroplasty. Clin Orthop Relat Res 435: Rajasekhar C, Das S, Smith A (2004) Unicompartmental knee arthroplasty. 2- to 12-year results in a community hospital. J Bone Joint Surg Br 86: Rees JL, Price AJ, Beard DJ, Dodd CA, Murray DW (2004) Minimally invasive Oxford unicompartmental knee arthroplasty: functional results at 1 year and the effect of surgical inexperience. Knee 11: Robertsson O, Borgquist L, Knutson K, Lewold S, Lidgren L (1999) Use of unicompartmental instead of tricompartmental prostheses for unicompartmental arthrosis in the knee is a costeffective alternative. 15,437 primary tricompartmental prostheses were compared with 10,624 primary medial or lateral unicompartmental prostheses. Acta Orthop Scand 70: Ryd L, Karrholm J, Ahlvin P (1997) Knee scoring systems in gonarthrosis. Evaluation of interobserver variability and the envelope of bias. Score Assessment Group. Acta Orthop Scand 68: Schai PA, Suh JT, Thornhill TS, Scott RD (1998) Unicompartmental knee arthroplasty in middle-aged patients: a 2- to 6-year follow-up evaluation. J Arthroplasty 13: van den Dikkenberg N et al (2002) Measuring functional abilities of patients with knee problems: rationale and construction of the

7 Knee Surg Sports Traumatol Arthrosc (2007) 15: DynaPort knee test. Knee Surg Sports Traumatol Arthrosc 10: Weale AE, Halabi OA, Jones PW, White SH (2001) Perceptions of outcomes after unicompartmental and total knee replacements. Clin Orthop Relat Res 382: Webster KE, Wittwer JE, Feller JA (2003) Quantitative gait analysis after medial unicompartmental knee arthroplasty for osteoarthritis. J Arthroplasty 18: Weidenhielm L, Olsson E, Brostrom LA, Borjesson-Hederstrom M, Mattsson E (1993) Improvement in gait one year after surgery for knee osteoarthrosis: a comparison between high tibial osteotomy and prosthetic replacement in a prospective randomized study. Scand J Rehabil Med 25: Witvrouw E et al (2002) A correlation study of objective functionality and WOMAC in total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 10: