Control-theoretic scheduling of psychotherapy and pharmacotherapy for the treatment of post-traumatic stress disorder

Control technology has found widespread use in medical applications, including pharmacology and drug dosing. Although mental and behavioural health applications have much in common with these systems, only a few seminal works have applied control technology to inform treatment with non-pharmacological interventions. Herein, a model-based, predictive control strategy is proposed to optimise treatment for post-traumatic stress disorder (PTSD). A mathematical model is developed that captures the dynamic relationship between five PTSD treatments and three commonly-employed PTSD assessment scales. The model supports an adaptive model predictive control strategy in which treatments are scheduled to achieve a therapeutic objective while recurrent assessment scale data are used to adaptively identify the patient model parameters. The model is corroborated against an independent data set, and the efficacy of the adaptive control strategy is demonstrated through Monte-Carlo simulations. While this work represents only a preliminary step towards a quantitative tool for clinical use, the simulated treatment results indicate that the proposed model and control approach could provide valuable insight towards designing personalised intervention schedules that increase treatment adherence, lower treatment costs and provide PTSD patients with more effective care.