Biostatistical Methods in Clinical Trials

Catalog description

Recommended Preparations: Ph.D. Biostatistics candidates must have completed a Master's degree that includes at least four graduate level Statistics courses. Dr.P.H. candidates must have completed a graduate (e.g., master's) or advanced professional (e.g., MD) degree. Dr.P.H. candidates may also be required to successfully complete PHS 520 , a core Biostatistics course, in their first semester. The background knowledge of these candidates provides sufficient preparation for PHS 582 . An in-depth course on biostatistical methods in the design and analysis of randomized and controlled clinical trials. PHS 582 Biostatistical Methods in Clinical Trials (3) This is an in-depth course on biostatistical methods in the design and analysis of randomized and controlled clinical trials. The course provides foundational information regarding the principles behind randomized and controlled clinical trials and then delves into topics that are specific to clinical trials. Study designs determine how the data are analyzed and how to avoid/minimize clinical biases, so study designs are investigated: (a) Phase I-IV trials, (b) non-inferiority designs, (c) factorial designs, and (d) crossover designs. Next, sample size calculations are investigated to optimize precision, and the biostatistical and logistical aspects of randomization methods are described. Adaptive designs can provide more efficient approaches, especially two-stage approaches. Adaptive designs are investigated for Phase II-III clinical trials. Many clinical trials invoke time-to-event outcomes, so survival analysis methods are covered in detail. Systematic reviews and meta-analyses consist of summarizing and analyzing the data across a set of independent clinical trials. In particular, fixed-effects and random-effects models for meta-analyses are explored. The final topic is medical diagnostic testing, in which clinical researchers try to identify new testing procedures for distinguishing between healthy and diseased individuals.