Pitfalls and prospects in clinical research on antianxiety drugs: benzodiazepines and placebo--a research review.

The research methodology of 78 double-blind studies comparing benzodiazepines and placebo in treating neurotic anxiety is critically reviewed. Many faults are noted in areas of subject selection, assessment of clinical response, study design and data analysis. Although 56.4% of the studies reviewed had results demonstrating a significant difference in clinical response between benzodiazepines and placebo, the frequent methodologic difficulties and inconsistent results from study to study lead the authors to question the efficacy of these drugs as a treatment for anxiety. A need for better designed studies is discussed and some research methodologies proposed.