Patient heterogeneity in clinical trials.
Interpretation of therapeutic results is complicated by variability in response among patients. This paper reviews fundamental statistical principles for the design of clinical trials. These methods seek to evaluate relative therapeutic efficacy in the presence of patient heterogeneity. Statistical science has more to offer therapeutics than significance tests among "comparable" treatment groups. The role of randomization and stratification is reviewed. The importance of study design, including patient eligibility and therapeutic standardization, to the generalization of conclusions is discussed.
Breast Neoplasms↗