Evaluation of current clinical trial methodology in analgesimetry based on experts' opinions and analysis of several analgesic studies.

There is general agreement that the controlled clinical trial is the best method for analgesic evaluation. Current practice in clinical analgesimetry, however, varies considerably with regard to design, measurement of pain, and statistical analysis. We attempted to assess the degree of this diversity by a questionnaire survey of a sample of investigators who are currently conducting clinical pain studies. To examine how various statistical methods perform in practice, we reanalyzed data from four of our published analgesic studies using a number of standard methods.