One standard request in larger clinical trials is to assess whether the observed treatment effect in the primary endpoint is consistent across subgroups. Analyses on subgroups is a widely discussed topic: searching the internet for “subgroup analysis in clinical trials” gives over 700,000 results. The discussion in the majority of these articles revolves around potential pitfalls of subgroup analysis. The idea of subgroup analyses is to test for homogeneity similarity or heterogeneity across the subgroups.
However, there is another type of “subgroup analysis” - where the interest lies in targeting certain patient populations that would benefit from a particular treatment. This is one goal of precision medicine; to effectively screen patients to ensure they are given the most appropriate treatment, based on genetic data (for example, pre-screening patients to identify those that are “biomarker positive” prior to treatment, or to enrol them onto a clinical trial).