Date created: | Last Updated:

Category: Project

Description: Background With estimated rates of about 25%, dropout in psychotherapies is a major concern for individuals, clinicians, and the healthcare system at large. Aims It is essential to predict and at best counteract dropout in psychotherapy. This study aims to compare the accuracy of different machine learning algorithms for the prediction of therapy dropout, to provide an overview of important predictor variables and to gauge the generalizability of the prediction models. Method Logistic regression models were compared with two machine learning algorithms (elastic net regressions and gradient boosting machines) in the prediction of therapy dropout in two large inpatient samples (N = 1,691 and N = 12,473) using patient- and therapist-reported variables (e.g., demographics, diagnoses, and symptoms) collected at the time of admission to the clinic. Results Predictive accuracies of the two machine learning algorithms were similar and higher than for logistic regressions: Therapy dropout could be predicted with an AUC of .73 and .83 for Sample 1 and 2, respectively. The initial evaluation of patients’ motivation and the therapeutic alliance rated by the respective therapist were the most important predictors of dropout across algorithms. Predictive performance highly depended on sample size and events fraction. Conclusions Therapy dropout in naturalistic inpatient settings can be predicted to a considerable degree by using baseline indicators. Feature selection via regularization leads to higher predictive performances whereas non-linear or interaction effects or interactions are dispensable. The most promising point of intervention to reduce therapy dropouts seems to be patients’ motivation and the therapeutic alliance. Keywords: therapy dropout; predictive modeling; machine learning; inpatients; helping alliance