Dr. Wagner and Colleagues Reply

To the Editor: Dr. Mathews and colleagues request further information about the randomized, placebo-controlled trial of citalopram for treatment of depression in children and adolescents. Randomization was on a 1:1 basis and was stratified by age group. The random assignment list was concealed from the investigators, which is fundamental to the claim that the study was performed under double-blind conditions. The protocol-specified population for all efficacy analyses, defined as the “intent-to-treat” population, included all patients who received at least one dose of double-blind study medication and had at least one postbaseline efficacy assessment. The analyses we presented in the manuscript were not only conventional in nature; they were, in fact, defined a priori. The justification for defining this population for the efficacy analyses is that the primary analysis was the change from baseline, therefore requiring a postbaseline assessment.

Although recently a mixed-model approach has gained some currency for the analysis of efficacy in antidepressant trials, the last-observation-carried-forward method of analysis has always been conventionally considered the most conservative method of analysis. Certainly this was the case when the study protocol was being developed. In escitalopram trials in adult patients, last-observation-carried-forward analyses minimize the treatment effect that is demonstrated by observed-cases analyses of the patients who actually remain in treatment (1, 2). These analyses are considered more conservative than observed-cases analyses for acute treatment antidepressant studies because the onset of antidepressant effect is typically delayed for up to several weeks. Therefore, the last observation of patients who discontinue active treatment prematurely is not likely to capture the full potential antidepressant effect.

Regarding suicidality, it is helpful to note that the manuscript states clearly that no serious adverse events were observed in the trial for citalopram-treated patients. At the time the manuscript was developed, reviewed, and revised, it was not considered necessary to comment further on this topic.

Dr. Martin and colleagues inquire about the value of 2.9, which was calculated as the quotient of the least square mean, divided by the common standard error of the mean for each treatment group. With Cohen’s method, the effect size was 0.32.

In response to Dr. Barbe’s questions about the methods of this randomized clinical trial for the treatment of depressed children and adolescents, there were 75 subjects who were screened but not randomly assigned. The method for elicitation of adverse events was chosen because it was the accepted standard at the time the study was designed for multicenter, industry-sponsored clinical trials in juvenile depression.

It may be considered premature to compare the results of this trial with unpublished data from the results of a study that has not undergone the peer-review process. Once the investigators involved in the European citalopram adolescent depression study publish the results in a peer-reviewed journal, it will be possible to compare their study population, methods, and results with our study with appropriate scientific rigor.

We believe that the results of our study, which demonstrated a significant difference between citalopram and placebo beginning at week 1, is clinically meaningful, particularly at a time when there have been so few antidepressants shown to have superiority to placebo for depressed children.