Reducing unnecessary measurements in clinical trials with multiple primary endpoints

Clinical trials often involve two or more primary endpoints. However, observing or measuring high-cost endpoints often reduces the efficiency of the study because of high medical costs, highly invasive measurements, or long-term follow-up. Further, the individual powers to demonstrate the overall efficacy of a new intervention for the multiple endpoints often differ under a given sample size. We propose an efficient clinical trial design in which the sample size for each of the endpoints is individually determined, taking into consideration both the cost and the individual power for each endpoint. We compared the efficiency of the proposed design with that of the conventional design using three variables: (1) the number of participants in the study, (2) the total number of measurements for all endpoints, and (3) the cost of enrolling the participants and obtaining the measurements for all endpoints. We extended the proposed design to a group-sequential design. Numerical examples show that the proposed design can reduce unnecessary measurements and adjust the individual powers for the endpoints, especially when the individual power for one endpoint is relatively higher than that for other endpoints in a study with multiple co-primary endpoints.