Estimation of heterogeneity of treatment effect using EM algorithm

Evgeniya Reshetnyak, Fordham University


Heterogeneity of treatment effect in randomized clinical trials is referred to the problem of high variability in treatment responses across patients. Although a study may suggest positive average treatment effect for a trial population, for subgroups of individuals a treatment may be not beneficial. Despite the existence of a wide array of ad-hoc and post-hoc statistical procedures, which are designed to discern or control for such variations of responses, each of them has certain limitations. The purpose of the current study is to propose the statistical model of estimation of Heterogeneity of Treatment Effect (HTE) in randomized clinical trials using EM algorithm, the approach that so far has been overlooked in the clinical research literature.

Subject Area

Statistics|Clinical psychology|Quantitative psychology

Recommended Citation

Reshetnyak, Evgeniya, "Estimation of heterogeneity of treatment effect using EM algorithm" (2015). ETD Collection for Fordham University. AAI1585672.