Description of the groups at trial entry

Of the 2374 participants randomised, 34 were recruited to the comparison assessing unicompartmental knee replacement versus TKR.

Description of data available for those recruited

A description of the group of participants recruited to this comparison is in Table 36.

TABLE 36

Description of groups at trial entry for unicompartmental knee replacement vs. TKR

Outcomes after a median of 10 years post operation

Oxford Knee Score

Table 37 and Figure 55 describe OKS over the 10-year follow-up period by allocated group.

TABLE 37

Descriptive statistics and estimated treatment effects at each follow-up time point for OKS for unicompartmental knee replacement vs. TKR

FIGURE 55

Mean (SD) OKS by group at each follow-up time point for unicompartmental knee replacement vs. TKR.

EuroQol 5D

Table 38 and Figure 56 describe EQ-5D over the 10-year follow-up period by allocated group.

TABLE 38

Descriptive statistics and estimated treatment effects at each follow-up time point for EQ-5D for unicompartmental knee replacement vs. TKR

FIGURE 56

Mean (SD) EQ-5D utility by group at each follow-up time point for unicompartmental knee replacement vs. TKR.

Short Form 12

Table 39 and Figure 57 describe the SF-12 PCS over the 10-year follow-up period by allocated group.

TABLE 39

Descriptive statistics and estimated treatment effects at each follow-up time point for SF-12 PCS for unicompartmental knee replacement vs. TKR

FIGURE 57

Mean (SD) SF-12 PCS by group at each follow-up time point for unicompartmental knee replacement vs. TKR.

Table 40 and Figure 58 describe SF-12 MCS over the 10-year follow-up period by allocated group.

TABLE 40

Descriptive statistics and estimated treatment effects at each follow-up time point for the SF-12 MCS for unicompartmental knee replacement vs. TKR

FIGURE 58

Mean (SD) SF-12 MCS by group at each follow-up time point for unicompartmental knee replacement vs. TKR.

Discussion

Recruitment to this arm of the trial was very slow and was therefore terminated early. Prior to stopping, 34 patients had been recruited. As there has been only one other randomised trial of unicompartmental knee replacement versus TKR, it was felt that the clinical scores should be described.⁹² No difference was found, as would be expected with small numbers. Complications, reoperations and revisions were not analysed, as it was felt the numbers were too small for this analysis to be of any value. The data from KAT has therefore not contributed significantly to the debate about whether unicompartmental knee replacement should or should not routinely be used. The experience gained from KAT has, however, been very useful in the planning of another randomised study of unicompartmental knee replacement and TKR – TOPKAT (Total Or Partial Knee Arthroplasty Trial).⁹³

During the planning and application for funding stages of KAT, unicompartmental replacements were implanted through the standard approach used for TKR and many surgeons had equipoise about the two types of replacement. We should, therefore, have been able to recruit, using the standard KAT methodology, an appropriate number of patients for the trial. However, prior to starting the recruitment, a new, minimally invasive technique for implanting unicompartmental replacement was introduced. This has many advantages over the standard approach, including a faster recovery, lower morbidity and improved function. As a result, many surgeons who would have recruited to the trial instead learnt the minimally invasive technique. In addition, some surgeons lost their equipoise. As a result, the recruitment rate was very much lower than predicted. The new trial, TOPKAT, was therefore designed differently from KAT in that it has two options. Surgeons with equipoise are able to randomise in a standard fashion, whereas surgeons who do not have equipoise can use an expertise-based randomisation. As most surgeons have now learnt the minimally invasive technique, many more are now willing to be involved in the standard randomisation arm. In the expertise-based option of the trial, patients who are appropriate for the study are randomised and then either have a unicompartmental replacement implanted by a surgeon who believes in unicompartmental replacement or a total replacement implanted by a surgeon who believes in total replacement. The TOPKAT study, which has been funded by the NIHR HTA board, finished its recruitment in September 2013 (HTA project reference number 08/14/08).