In the treatment of obsessive-compulsive disorder (OCD) with antidepressant medication, the earliest reliable indication of treatment failure remains uncertain. We investigated if non-improvement following 4 weeks of treatment predicts nonresponse at the end of the trial.

We conducted a random-effects bivariate diagnostic accuracy study using individual patient data from industry-sponsored short-term trials of adults with OCD receiving selective serotonin reuptake inhibitors or clomipramine, submitted for marketing approval. The primary outcome was accuracy of non-improvement (<25% reduction on the Yale–Brown Obsessive Compulsive Scale [YBOCS] after 4 weeks) in predicting nonresponse (<35% YBOCS reduction at trial endpoint [10–13 weeks]). Secondary outcomes were accuracy of non-improvement after 6 weeks, nonresponse after 8 weeks, and inclusion of Clinical Global Impression Scale – Improvement in definitions of improvement and response. We performed meta-regressions for sex, age, severity, trial duration, dosing regimen, and compound.

In 11 studies totaling 1,753 patients, non-improvement at week 4 predicted subsequent nonresponse (positive predictive value, PPV) in 86% of cases (95% confidence interval [CI] = 83–88%). Sensitivity was 78%, specificity was 70%, and the negative predictive value was 60%. Secondary outcomes showed similar PPV after 6 weeks and a PPV of 93% for nonresponse after 8 weeks. Predictive accuracy was significantly higher in men relative to women (β = −0.64, 95% CI = −1.12 to −0.16, p = 0.0089).

Patients with OCD who do not improve after 4 weeks of antidepressants will likely not respond to short-term treatment. Thus, a change in strategy should be considered after 4 weeks without treatment benefits.

Electroconvulsive therapy (ECT) presents itself as a highly effective therapeutic approach in various psychiatric conditions, especially affective disorders and catatonia. Although obsessive-compulsive disorder (OCD) is not an established indication for ECT, there are several positive results that have been replicated, giving us an account of its potential applicability.

To emphasize the importance of defining predictors of response to ECT in OCD.

The authors’ clinical experience is combined with the review of clinical cases, available in the literature, related to the application of ECT in OCD.

Personal or family history of affective pathology and obsessions of sexual content were identified as potential predictors of response to ECT in patients with obsessive and compulsive symptoms.

Although preliminary and based solely on case reports, the replicability of results should promote special attention to situations in which OCD is marked by particular characteristics that favor the response to ECT. In this way, it would be possible to prevent the dragged consumption of health resources and minimize the expected chronicity associated with this clinical condition.