Unicompartmental Arthroplasty with the Oxford Knee. Buy the book here.
This chapter is available from Goodfellow Publishers as a PDF.
The results of UKA can be gathered from three main sources: the reports of the national registers, observational studies (both comparative and case series), and randomised controlled trials. In this chapter, we attempt an overview of the clinical results of UKA in general and OUKA in particular. It is important to note that the ‘result’ is of the whole arthroplasty which includes the indications, the technique as well as the implant.
National registers
In Joint Registries, longitudinal data is collected from large numbers of participating institutions before being assembled centrally. In most cases, a report is issued annually and raw data are released on request for research studies. The principal aim of joint registries is to facilitate the identification of poorly-performing implants at the earliest possible stage, allowing modification or abandonment of such implants before large numbers are implanted. Registers collect data when a joint replacement implant is inserted, ideally from the surgeons. Although there is some variation between registers, a revision operation is usually considered to have occurred the second time an implant is inserted in a particular joint. Using this information, cumulative revision rates (CRR) can be calculated. They are our best source of information on the epidemiology and demography of arthroplasty.
The first national joint registry was established in Sweden in 1975. Since then, national joint registers have been established in Finland (1980), Norway (1987), Denmark (1995), Australia, and New Zealand (both 1998), amongst others. The National Joint Registry for England and Wales (NJR) was established in 2003 and is currently the largest database of joint replacements in the world.
The principal advantage of national joint registers is the large number of cases they report. In some cases, participation is near-universal (and in some cases is obligatory) which minimises the problems of reporting and publication bias. The fact that national joint registers study the population as a whole allows great diversity within cases studied in terms of implant type, surgical technique and experience, patient selection and postoperative regimen. The large overall number of cases allows the study of these subgroups with acceptable power. National registers have additional benefits in allowing surgeons to compare their results with their peers’, and allowing easy identification of implants in need of recall.
However, national registers remain imperfect tools to measure outcome. The large number of cases reported, and the reliance on operating units to report their cases, limits the quantity of data that can be gathered on each patient. In all national registers, the primary measure of outcome is the rate of revision surgery; whilst this has the benefit of being objective and easy to measure, it has several deficiencies. When a revision occurs, the implant is considered to have failed. If it has not been revised, it is considered to have survived and be a success even if it is painful and has poor function. Implant survival is a solid end-point and has been described as the point at which both the surgeon and the patient agree that revision is preferable to continuing with the prosthesis in situ. As a result of the way data is collected by the registers, revision is considered to occur if a new implant is inserted. The commonest revision is therefore removal of a joint replacement and replacement with a new one. The addition of an extra component, such as secondary resurfacing (after TKA), the addition of a lateral or patellofemoral replacement (to a medial UKA with osteoarthritis progression) or exchange of a bearing (for a dislocation or a washout) are therefore also considered to be a revision whereas replacement of the original bearing after a dislocation is not. Using the same definition, an amputation, a knee fusion or death resulting from surgery would not be considered a “revision” and so that the knee arthroplasty would be considered a success.
It is therefore important to consider a whole series of different end points other than just revision to assess success or failure of a joint replacement. These could include all adverse events such as reoperations, complications, mortality and morbidity and patients with poor outcome scores and/or dissatisfaction.
