Computational Psychiatry & Decision-making

Other Research Topics



The documents distributed here have been provided as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.


  • doi pdf Low predictive power of clinical features for relapse prediction after antidepressant discontinuation in a naturalistic setting
  • Berwian IM, Wenzel J, Kuehn L, Schnuerer I, Seifritz E, Stephan KE, Walter H, Huys QJM
  • bioRxiv
  • Background: The risk of relapse after antidepressant medication (ADM) discontinuation is high. Predictors of relapse could guide clinical decision-making, but are yet to be established. Method: We assessed demographic and clinical variables in a longitudinal observational study before antidepressant discontinuation. State-dependent variables were re-assessed either after discontinuation or before discontinuation after a waiting period. Relapse was assessed during six months after discontinuation. We applied logistic general linear models in combination with least absolute shrinkage and selection operator and elastic nets to avoid overfitting in order to identify predictors of relapse and estimated their generalisability using cross-validation. Results: The final sample included 104 patients (age: 34.86 (11.1), 77% female) and 57 healthy controls (age: 34.12 (10.6), 70% female). 36% of the patients experienced a relapse. Treatment by a general practitioner increased the risk of relapse. Although within-sample statistical analyses suggested reasonable sensitivity and specificity, out-of-sample prediction of relapse was at chance level. Residual symptoms increased with discontinuation, but did not relate to relapse. Conclusion: and Relevance Demographic and standard clinical variables appear to carry little predictive power and therefore are of limited use for patients and clinicians in guiding clinical decision-making.