Send to

Choose Destination
Am J Psychiatry. 2003 Nov;160(11):1919-28.

Which SSRI? A meta-analysis of pharmacotherapy trials in pediatric obsessive-compulsive disorder.

Author information

Pediatric OCD Problem, WACC 725, Massachusetts General Hospital, 15 Parkman Street, Boston, MA 02114, USA.



The authors conducted a meta-analysis of published randomized, controlled medication trials in children and adolescents with obsessive-compulsive disorder (OCD) to assess evidence for differential efficacy based on type of drug, study design, and outcome measure.


A systematic literature search was performed for articles pertaining to the pharmacological treatment of pediatric and/or adolescent OCD. All baseline, posttreatment, and change scores with standard deviations reported in each study were included in the analyses. Effect sizes for dependent measures were expressed as standardized mean differences. The analysis included data on efficacy for four selective serotonin reuptake inhibitors (SSRIs) (paroxetine, fluoxetine, fluvoxamine, and sertraline) and clomipramine, four study designs, four dependent outcome measures, and two types of outcome scores (change and posttreatment scores). Multivariate regression was performed to assess the degree to which the effect sizes varied with the methodological features of each study.


Twelve studies with a total of 1,044 participants met all inclusion criteria for the analysis. The pooled standardized mean difference for the results of all studies was 0.46 and showed a highly significant difference between drug and placebo treatment. Only one of the four outcome measures evaluated was not sensitive to change with treatment. A multivariate regression analysis of drug effect with other variables controlled showed that clomipramine was significantly superior to each of the SSRIs but that the SSRIs were comparably effective.


Although highly significant, the overall effect sizes for medication were modest. Similarities and differences between the variables studied that emerged in the meta-analysis may have implications for both clinical care and future research.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Atypon
Loading ...
Support Center