Biostatistics in Clinical Trials

Abstract

Biostatistics plays an important role in the design, conduct, and reporting of clinical trials. In recent years, the recognition and appreciation of biostatistics have greatly improved. Biostatisticians are not only responsible for the quantitative properties of clinical trial designs and for analyzing data from clinical trials, they are also respected partners of the clinical teams contributing to strategic discussions and helping solve day-to-day issues on studies and clinical programs. Biostatisticians also drive the development and implementation of innovative statistical methodology. This chapter provides a brief overview of basic biostatistical principles and analysis methods used in clinical trials. In the Introduction, we cover basic biostatistical principles of clinical trials including population and sample, errors and bias, choice of a control arm, randomization, allocation, and blinding. Phases of clinical development and commonly used statistical designs are described in Sect. 1.2. A brief overview of the most commonly used probability models and basic principles of estimation and inference with an emphasis on modeling techniques are given in Sect. 1.3. Where appropriate, references to available software procedures in R and/or SAS are provided. Finally, we highlight some important considerations for clinical trials, such as defining appropriate estimands, handling outcomes with missing data, applying multiplicity adjustments, analyzing subgroups, planning multiregional clinical trials, and evaluating drug safety.