Survival analyses of randomized clinical trials adjusted for patients who switch treatments

Patients who switch treatment groups in randomized clinical trials can cause problems in the interpretation of the results. Although the intention-to-treat method is recognized as being the most reliable analysis, it may result in an underestimate of the treatment effect if there have been patients who switch treatments. In this paper, an adjusted analysis, based on a Cox model, is proposed which estimates the treatment effect allowing for patients who switch treatments. Because this analysis requires much weaker assumptions than 'as treated' analyses, it is likely to be much more reliable. The method is illustrated by a trial comparing radical radiotherapy with radiotherapy followed by cystectomy in the treatment of invasive bladder cancer.