Meta-Analytic Predictive Modeling

Article 3 is entitled “Predicting Effects of a Digital Stress Intervention for Patients with Depressive Symptoms: Development and Validation of Meta-Analytic Prognostic Models Using Individual Participant Data”. This paper describes the development and validation of meta-analytic prognostic models, which are used to predict individual treatment benefits of a digital stress intervention among patients with clinically relevant depressive symptoms. We employ both an “effect” and “risk modelling” strategy, and maximized models using novel multilevel machine learning algorithms. Optimal decision rules were examined using a framework drawing on the Neyman-Rubin causal model. The developed models may either be used to “screen out” the small proportion of patients for whom no relevant effects are predicted; or to provide the intervention to only those patients with very high expected benefits.

Article 3 has been published in the “Journal of Consulting and Clinical Psychology” (Harrer, Baumeister, et al., 2023).