Appraising the Quality of Drug-Evaluation Research

Abstract

The present paper describes a methodological review procedure for evaluating a body of research focusing on the study of a prophylactic treatment medication. An illustrative application of this procedure to research evaluating lithium prophylaxis for bipolar affective disorders revealed possible biases in respect to subject selection, and general problems with diagnostic reliability and the reliability of the dependent variable. A tendency to ignore the statistical power aspects of determining sample size was also noted, along with possible difficulties relating to infrequent assessments of patients'' clinical conditions. The implications of these matters were discussed in terms of the unique vantage point offered by this methodological review process.