OBJECTIVES: Psychotic disorders are burdensome, episodic, and commonly persist over the life course. Current Health Economic Decision Models (HEDM) for psychotic disorders have mostly been developed for pharmacotherapies, and often use trial input data. For this study, a novel patient-level continuous-time state transition model was developed using available observational data obtained from clinical routine practice, simulating direct healthcare use costs and health related outcomes for psychotic patients over the life course.

METHODS: Registry data from the Netherlands (N = 12,835) collected between 2000 and 2019 was available for model development. In- and out-of-episode status were represented by two model states based on structural changes in healthcare use. Death was the absorbing model state. Mortality was approximated by Gompertz distributions fitted conditionally on age and sex using observed data. Sojourn time for the two alive states was modeled using two parametric Generalized Gamma distributions. Mean healthcare use costs were assigned to alive states based on regression prediction models. Quality of life weights were extracted from literature. Internal validation was performed at population level by comparing distribution quantiles. Additional validation tests were performed using the AdViSHE tool.

RESULTS: Between 2010 and 2019, the observed direct healthcare use costs were €377 million, and more than a total of 12,800 patient life years were spent in-episode. The simulation model predicted €426 million (13% overprediction) and more than 16,200 in-episode patient life years (27% overprediction) for the same period.

CONCLUSIONS: The study introduces a micro-simulation model for psychosis patients over their life course. It differs from existing health economic decision models since it was developed using long-term observational data, and since it uses an individual and relative definition for episodes. The model is suitable for assessing interventions with treatment effects in terms of episode occurrence, episode duration, or relapse for any recurrent disorder.