Methods for analyzing composite endpoints

In many clinical studies the interest lies in the comparison between two treatment groups regarding a rarely occurring event. To show a relevant treatment effect with sufficient power many patients have to be included and observed over a long time period. A composite endpoint was proposed to avoid this. A composite endpoint combines all events/endpoints of interest. Usually time-to-first-event analysis is used to analyze a composite endpoint. Hence, it is neglected that more than one event might be observed per individual. Furthermore, the different clinical relevance of the combined events is not considered. Therefore, different approaches to analyze composite endpoints were evaluated within the DFG-project entitled

Methods for planning and analyzing clinical studies with composite endpoints

  • Within the project methods for analyzing recurrent events were compared with the help of a Monte-Carlo simulation study and applied to composite endpoints.

    Moreover, the so-called weighted all-cause hazard ratio was proposed. The weighted all-cause hazard ratio uses pre-defined weights to capture the different clinical relevance of the components of a composite endpoint.

    • Prof. Dr. Geraldine Rauch (Charité - Universitätsmedizin Berlin)

  • -

  • Reevaluation of risk factors for time to subsequent events after first stroke occurrence using a new weighted all-cause effect measure
    Ozga A, Rauch B, Palm F, Urbanek C, Grau A, Becher H, Rauch G
    BMC PUBLIC HEALTH. 2020;20(1):817.

    Introducing a new estimator and test for the weighted all-cause hazard ratio
    Ozga A, Rauch G BMC MED RES METHODOL. 2019;19(1):118.

    A systematic comparison of recurrent event models for application to composite endpoints
    Ozga A, Kieser M, Rauch G
    BMC MED RES METHODOL. 2018;18(1):2.

    A weighted combined effect measure for the analysis of a composite time-to-first-event endpoint with components of different clinical relevance Rauch G, Kunzmann K, Kieser M, Wegscheider K, König J, Eulenburg C STAT MED. 2018;37(5):749-767.