Three-year medication prophylaxis in panic disorder: to continue or discontinue? A naturalistic study.

OBJECTIVE

Little is known about maintenance treatment for panic disorder. The purpose of this naturalistic study is to compare outcomes of remitted panic disorder patients continued on versus those successfully discontinued from maintenance medication.

METHODS

After 3 years of sustained remission with medication in a naturalistic setting, 168 patients were continued on, whereas 37 successfully discontinued from medication. Continued and discontinued groups were followed for an additional 4 to 8 years and compared for differences in treatment outcome using chi(2) and Wilcoxon rank sum tests. Times to relapse were analyzed using the Kaplan-Meier product-limit method, and risk factors for relapse were assessed using Cox proportional hazards regression.

RESULTS

The discontinued group was healthier at baseline but had a significantly worse outcome compared with the continued group. Panic-free survival probabilities for the continued group at 1, 2, 3, and 4 years were 0.87, 0.81, 0.71, and 0.64, respectively, and were significantly higher than respective probabilities of 0.53, 0.35, 0.29, and 0.15 for the discontinued group. Median survival time in the continued group was significantly longer, at 5.67 years, than in the discontinued group, at 1.17 years. Cognitive behavioral therapy significantly reduced hazard in the discontinued but not in the continued group. Residual symptoms in either group at time of assignment predicted poorer outcome.

CONCLUSION

Our small study suggests that relapse of panic disorder in routine clinical practice occurs even after long-standing remission on maintenance medication, and that relapse risk appears to be markedly higher after medication discontinuation. Discontinuation may be more successful in candidates who received cognitive behavioral therapy and have minimal residual symptoms.