Assessment of the integrity of study blindness in a pediatric clinical trial of risperidone.

OBJECTIVE Controlled clinical trials in pediatric psychopharmacology rely on blinded parents and clinical evaluators for outcome data, but little is known about the success of the masking procedures. The blindness of clinical evaluators and parents was examined in a clinical trial of risperidone in autism. METHODS Clinical evaluators and parents were asked to guess individual treatment assignments at the end of an 8-week placebo-controlled trial of risperidone in children (aged 5-17 years) with autism. Clinical evaluators did not have access to adverse event data. RESULTS The rates of correctly guessing individual treatment assignment (risperidone or placebo) were significantly greater than chance for both clinical evaluators and parents (P < 0.001). Clinical evaluators associated improvement with attribution to risperidone, and lack of improvement with attribution to placebo, in both the risperidone and placebo treatment arms. Parents associated improvement with attribution to risperidone only in the placebo treatment arm. Parents reported that adverse events influenced their guesses, but presence of adverse events was not associated with correctness of guess. CONCLUSION Improvement was associated with attribution to active treatment regardless of actual treatment assignment, and adverse events did not appear to be a threat to study blindness.