Survivorship analysis in the evaluation of joint replacement.

A major problem in the evaluation of patients receiving joint replacements is the lack of complete information on all patients. That is, at the time of analysis patients will not have been followed long enough to know what their status will be at the time point of interest (5 years, for example). The statistical technique of survivorship analysis is the appropriate method for evaluating this type of data. The authors illustrate the technique of survivorship analysis as it applies to total joint arthroplasty, using examples from the UCLA total hip arthroplasty data base. In addition, the benefits of using this technique over the usual "failure rate" analysis are discussed, as well as other issues concerning the analysis of joint replacement data.