Using smartphone surveys to personalize interventions: development and piloting of the Bayesian Statistical Individualization Algorithm (BaSICA)

Abstract

This article provides details on the development of a statistical learning algorithm developed for constructing personalized treatment plans for psychotherapy. The algorithm takes data collected via Ecological Momentary Assessment (EMA) as an input. From this, it constructs an idiographic disorder model that reflects the latent dimensions of this patient’s psychopathology and their temporal interrelations. The priority of individual problems is derived from this statistical model. Based on this, treatment modules from cognitive-behavioral therapy are ranked so that the problems with the highest priority are dealt with first. A case study is used to illustrate the different analysis steps of the algorithm from data collection to the treatment plan.