Methods for analyzing recurrent event times

In many cardiovascular or oncological trials the analysis of rarely occurring events, such as myocardial infarction, stroke, a specific cause of death, or the occurrence of relapses or metastases is of interest. Usually only the first occurring event is thereby considered in the analysis. Hence, it is neglected that a patient might experience more than one non-fatal event. The events that are observed per patient might be of different clinical relevance. To describe such a disease process and the efficacy of a medication adequately and holistically, it is of interest to include all observed events in an analysis. Therefore methods for analyzing recurrent events were evaluated within the project entitled

Methods for planning and analyzing clinical trials with recurrent time to events with different event types

The project was supported by the Research Promotion Fund of the Faculty of Medicine (FFM), University Medical Center Hamburg-Eppendorf.

  • Methods for analyzing recurrent events were already proposed. However, their assumptions are restrictive and usually not met in clinical practice. Hence, the performance of this methods were evaluated in different realistic settings for recurrent correlated events of different event types with the help of a simulation study.

    The corresponding results are used to give a recommendation how to adequately analyze trials with recurrent events in the future and develop new methods that describe the disease process efficiently, unbiased, and as a whole.

  • -

  • -