An Analysis of a Set of Medical Data Using the Bootstrap Procedure

Abstract

The efficacies of two anti-inflammatory drugs in ankylosing spondylitis and related complaints were studied at a single medical clinic over a period of twenty-eight weeks. The purposes of this project were: (1) -To determine .any significant differences within and between the two drug groups using well-known nonparametric procedures, and (2) To illustrate the use of the bootstrap method and determine whether it is appropriate and useful for this data set. Some statistically significant changes indicative of improvement occurred among both groups of patients for primary efficacy variables. No definite trend was found for most of the laboratory variables. Both drugs demonstrated effective pain relief. Regarding the variables of day and night pain relief as well as pulse, the Experimental Drug proved to be clinically but not statistically superior to the other commonly used drug. Analyses of safety data indicated some statistically significant changes in both drug groups. There was a statistically significant difference between drug groups at baseline.