Statistical analysts engaged in typical clinical trials often have to confront a tight schedule to finish massive statistical analyses specified in a Standard Operation Procedure (SOP). Thus, statisticians or not, most analysts would want to reuse or slightly modify existing programs. Since even a slight misapplication of statistical methods or techniques can easily drive a whole conclusion to a wrong direction, analysts should arm themselves with well organized statistical concepts in advance. This paper will review basic statistical concepts related to typical clinical trials. The number of variables and their measurement scales determine an appropriate method. Since most of the explanatory variables in clinical trials are designed beforehand, the main statistics we review for clinical trials include univariate data analysis, design of experiments, and categorical data analysis. Especially, if the response variable is binary or observations collected from a subject are correlated, the analysts should pay special attention to selecting an appropriate method. McNemar's test and multiple McNemar's test are respectively recommended for comparisons of proportions between correlated two samples or proportions among correlated multi-samples.
Chen HL, Yang A. Common Pitfalls in SAS Statistical Analysis Macros in a Mass Production Environment. NESUG 2007; Statsitics and Data Analysis, http://www. nesug.org/proceedings/nesug07/sa/sa05.pdf [Online] (last visited on 8 May 2012).
Montgomery DC. Design and Analysis of Experiments. 8th ed, John Wiley & Sons, 2012.
Meyers LS, Gamst G, Guarino AJ. Applied multivariate research: Design and interpretation. Sage publications, 2006.