Subject-own-control design in evaluating clinical antidepressant effects

This study tested the use of a subject-own-control, multiple crossover design for evaluating the clinical effects of an established tricyclic antidepressant drug, imipramine. Although significant physiological and cognitive performance effects were demonstrated, only one clinical measure, target symptom change rated by the patients, showed a statistically significant drug-placebo difference. This result is in considerable contrast to the sensitivity of similar designs in detecting the clinical effects of antianxiety drugs in studies employing less than 20 patients. Crossover designs appear to be most successful when the treated condition is continuous, rather than episodic, and the treatment effects have a rapid onset and offset.