In reply: We thank Jureidini and Tonkin for their comments. In relation to their criticism of the study by Emslie et al,1 intention-to-treat analysis is the accepted standard. In relation to the study by Keller et al,2 their criticism about the criteria for response has already been answered elsewhere. (Criteria were defined in the report as a final Hamilton Rating Scale for Depression [HAM-D] score that was ≤ 8 or a reduction from baseline of ≥ 50%. Dual criteria were selected because the scores at entry could range from a minimum of 12 [set by protocol] to a maximum of 53 [highest scores for the 17-item HAM-D]. Limiting response to either a 50% reduction or a specified cut-off point would impede patients at the lower end of the ranges from meeting the criterion.3) The concern about the absence of differences in change scores on the HAM-D cannot be resolved, as mean change in scores and standard errors of the means were not reported. However, 63.3% (57/90) of subjects taking paroxetine (P = 0.02 versus placebo) achieved a HAM-D total score of ≤ 8 at endpoint. With respect to the size of difference in response rate to active treatment versus placebo, the results of trials involving selective serotonin reuptake inhibitors (SSRIs) in children and adolescents are similar to those reported in adults (ie, SSRIs achieve a response in about 20% more participants than placebo,4 which is similar to the 26% difference between cognitive behavioural therapy and various control conditions5). Consistent with these results, the US Food and Drug Administration has recently approved fluoxetine to treat children and adolescents aged seven to 17 years for major depression. We believe that there are sufficient new treatment data (including a study,6 published since the submission of our article, showing fluoxetine's superiority over placebo in symptom improvement and in remission rates) to warrant revision of the 1997 guidelines.
Publication of your online response is subject to the Medical Journal of Australia's editorial discretion. You will be notified by email within five working days should your response be accepted.