Does algorithm-guided treatment modify patients’ outcomes in the longer run? Data from the German Algorithm Project (GAP3)

M. Adli; K. Wiethoff; T. Baghai; T. Stamm; R. Fisher; M. Jäger; R. Ricken; P. Brieger; G. Laux; H. J. Möller; M. Bauer

doi:10.1055/s-2007-1002798

Does algorithm-guided treatment modify patients’ outcomes in the longer run? Data from the German Algorithm Project (GAP3)

M Adli ¹, K Wiethoff ¹, T Baghai ², T Stamm ¹, R Fisher ¹, M Jäger ², R Ricken ¹, P Brieger ³, G Laux ⁴, HJ Möller ², M Bauer ⁵

¹Charité-Universitätsmedizin Berlin, Department of Psychiatry and Psychotherapy, Campus Charité Mitte, Berlin, Germany
²Department of Psychiatry and Psychotherapy, Ludwig-Maximilians-Universität Munich, Germany
³Department of Psychiatry and Psychotherapy, University of Halle, Germany
⁴Department of Psychiatry and Psychotherapy, District Hospital Gabersee, Wasserburg am Inn, Germany
⁵Department of Psychiatry and Psychotherapy, Carl Gustav Carus Universität Dresden, Germany

Abstract

The German Algorithm Project (GAP) demonstrated that a standardized stepwise drug treatment regimen (SSTR) results in superior treatment outcomes compared to treatment as usual in inpatients with depression. The modifying effects of algorithm-guided acute treatments of inpatients on longer term outcomes have not yet been investigated. 429 GAP participants were studied with regard to overall outcomes at hospital discharge after treatment of the index episode and clinical status after 1 year. Remission was defined as a Hamilton-Depression Rating Score (HAMD-21)=9. GAP3 was a multicenter randomized clinical trial comparing two different algorithms (I: SSTR, II: computerized documentation and expert system, CDES) with treatment as usual (TAU) within the German Research Network on Depression. SSTR more likely resulted in remission at discharge than CDES or TAU (OR=2,641;p=.003), but no group differences were found after 1 year. Patients with remission at discharge were more likely to be remitted after 1 year than patients who were discharged without full remission(OR=2,897;p=.045). Patients with comorbid axis I (OR=0,404;p=.049) or axis II conditions (OR=0,226;p=.078) were less likely to show remission after 1 year. SSTR but not CDES increased the probability of being remitted at discharge independent from study completion or dropout. Full remission from index episode decreases the risk of relapse, comorbid axis I or II conditions increase the risk of relapse after 1 year.

This study was supported by BMBF, Pfizer, Pharmacia, Wyeth Pharma, Lilly Deutschland, Janssen-Cilag