Commentary: Unmasked and Uncontrolled Medication Trials in Child and Adolescent Psychiatry

Abstract

The four open and uncontrolled pharmacological trials published in this issue of the Journal are timely and of interest to the psychiatrist who treats children and adolescents. These articles include important informa­ tion on dosage, tolerability, and usefulness of four dif­ ferent medication treatments used in children and adolescents. Each of these clinical trials is unmasked, i.e., unblinded (Meinert, 1986), and none of the studies uses a concurrent control group, i.e., a placebo or com­ parison treatment group. The subjects in each trial were given the medication, and the treating physician, the rater of symptoms, the subject, and family were all aware that the patient was taking active medication. The problem with unmasked and uncontrolled trials is the potential for introducing bias that can influence the results of the study. Of most concern is bias that results in inflated estimates of efficacy and unrealistic reporting of adverse events. We do not suggest that authors of open trials intend to bias the outcome of a study. Rather, in uncontrolled trials, nonspecific treat­ ment effects and the natural course of illness can often account for the observed clinical improvement and be falsely attributed to the active treatment. In addition, the publication of unmasked and uncontrolled trials can create the impression that a medication is effective or lead uniformed readers to conclude that because the trial was published, the results must be true. The authors must therefore be very clear about the limitation of interpreting the results, and the reader must be skeptical of claims made on the basis of unmasked and uncontrolled trials (Meinert, 1986). Why are these studies published, if there is some chance that bias affected the outcome of the study? The