Treatment of depressive disorders with and without medication - a naturalistic study.

OBJECTIVE Although randomized clinical trials are the standard method for comparing the efficacy of various depression treatments, the external validity and generalizability of findings obtained by this approach can be questioned for several reasons. In this naturalistic study, we compared the effectiveness of treatments conducted by psychiatrists and clinical psychologists without prescription of drugs to treatments by psychiatrists and physicians using antidepressant agents in patients with depressive disorders in a representative sample of the normal population. Our assumption was that this sample is more representative for subjects treated for depressive disorders than subjects included in controlled trials. METHODS In a post hoc analysis, the health status of depressed patients under treatment for major, minor or recurrent brief depression with medication (20 patients) and without medication (30 patients), and untreated depressed subjects were compared over a one-year period using SCL-90-R depression scale scores. RESULTS At baseline, the two treatment groups were comparable in terms of diagnosis and severity of symptoms. Treatment effects were relatively small; patients treated with antidepressants tended to improve, whereas patients treated without drugs deteriorated slightly over one year. Seven years later, treated and untreated subjects no longer differed. CONCLUSION The results are consistent with previous studies on usual care of depressed patients showing low response rates for non-standardized treatments outside of research settings. Despite a number of methodological shortcomings (small sample size, heterogeneity of subjects and treatments, unusual rating instruments), the results presented here are unique, and provide relevant insight into usual care of depressed patients. This study underlines the importance of standardized psychotherapy and pharmacotherapy in routine practice, of studies investigating the transferability of results of outcome studies across clinical populations, and of quality assurance in primary care psychiatry.