Statistical methods for evaluating multiple sclerosis therapies.

Abstract

This article will discuss basic concepts and simple but commonly used methods of statistical analysis which are relevant to the evaluation of the results of randomized controlled clinical trials in multiple sclerosis. The focus throughout will be on an expository discussion to facilitate understanding of the logic, objectives and implementation of these methods. The context for most of the discussion is that of a two-armed clinical trial, involving a placebo and an active treatment arm. Clinical trials are carried out to allow conclusions concerning the efficacy and effectiveness of therapies, so the discussion will focus on aspects of inferential statistics. Simple methods for continuous and count responses, as well some specifically developed for use with categorical and time-to-event data are discussed. A very brief discussion of some of the more sophisticated methods that are often essential for a comprehensive analysis of the data collected in a clinical trial is also provided.