Clinical trials symposium. Highlights in development of randomized clinical trials

Abstract

The randomized clinical trial is accepted as the most reliable method of determining the relative merits of different therapies. At best it is a scientific experiment in the clinical setting, and demands all the rigors of experimental methods to produce valid data from which inference may be drawn. Randomization ensures the highest probability that the only difference between treated patient groups is the treatment given, and it is the component of clinical trials that produces the greatest ethical problems. It has three essential features. First, it eliminates investigator bias; second, it tends to balance known and unknown prognostic parameters in treatment groups; and, third, it forms the basis for the validity of statistical tests of significance.