German Journal of Psychiatry    ISSN 1455-1033

2000, Vol. 3, Supplement 1, S4

Predictors for Slow Response to Treatment in Panic Disorder

P. Berger, G. Sachs, M. Amering, A. Holzinger, D. Doby, B. Bankier, H. Katschnig

Department of Psychiatry, Division of Social Psychiatry, University of Vienna, Austria

Background: In the treatment of panic disorder improvement starts usually within the first two months, but some patients may improve after prolonged treatment. At present it is not known what predicts early or late response. The aim of the study was to analyzed the impact of pretreatment characteristics and personality disorders on the onset of remission in the treatment of panic disorder.

Methods: Data of seventy-three out-patients suffering from panic disorder with or without agoraphobia and who completed at least 6 weeks of a randomized trial of 24 weeks of either paroxetine only or paroxetine combined with cognitive group-therapy were analyzed in a Cox-proportional hazards model.  Structured diagnostic interviews (SCID and IPDE) were used to assess Axis I and Axis II disorders.

Results: The likelihood to respond to treatment (defined by absence of panic attacks and a clinical rating of improvement) at any point of time for patients without a personality disorder was 2.8 times greater than for patients with a personality disorder, and twice as high for patients without agoraphobia than with agoraphobia. Median time to response in patients with and without a personality disorder was 20 weeks and 8 weeks, respectively, and in patients with and without agoraphobia 10 and 8 weeks, respectively.

Conclusion: In previous studies personality disorders and phobic avoidance have been found as predictors for non-response in panic disorder. Our study shows that patients with these characteristics have a chance to profit from treatment, but they may need a substantially longer time. This has to be considered before classifying a patient as non-responder and switch to another treatment.