Online Program

Return to main conference page

All Times EDT

Thursday, September 24
Thu, Sep 24, 1:30 PM - 2:45 PM
Virtual
Use of Real-World Data as External Control in Clinical Trials: From Design to Implementation

Use of Real-World Data as External Control in Clinical Trials: From Design to Implementation (301183)

View Presentation

*Ming Tony Tan, Georgetown University Medical Center 

Keywords: RWD, external control, causal inference

We propose a clinical design and analysis procedure for comparing a new therapy with an external control which are obtained using real world data such as historical database, disease registry or electronic health record. In particular we focus on trials for diseases where we generally lack knowledge about disease etiology, natural history, pathophysiology or a small population of patients who may be more geographically dispersed and a randomized control trial is not feasible. The method is motivated by a trial to compare relapse-free survival (RFS) rate at 3 years between locally treated high-risk ocular melanoma patients on adjuvant combination immunotherapy versus a matched contemporaneous control population. The fixed time landmark survival comparison is not only most relevant for adjuvant immunotherapy but also enhances the statistical power of the design. The control is obtained using real world data from a national collaborative registry Ocular Melanoma Registry which is being established with a contemporaneous population collected from centers not participating in the trial. Furthermore, we extend the double robust causal inference procedure to obtain estimate of treatment effect that is robust with respect to misspecification of propensity score and regression model. This work is in collaboration with the Hoosier Cancer Research Network Melanoma Clinical Trials Group and the Cure Ocular Melanoma Foundation (for the control population)