Application of Decision Science to Mental Health Policy

Not only are resources for mental health scarce, but they are also distributed inequitably
Matt Miller



Evaluating the cost-effectiveness of mental health treatment strategies

This initiative encompasses several individual projects which employ decision analytic methods to comparatively assess different treatment strategies for specific disorders. We are constructing several different computer-based models to synthesize the best available data and simulate the natural history of several mental health disorders. These models are then used to project the health and economic consequences of alternative strategies.

Clinical areas of application have thus far included depression, panic disorder, and cluster B and C personality disorders. Target populations have ranged from children, where we are assessing the risks and benefits associated with antidepressant use, to adults where we are evaluating different intervention strategies for panic disorder including cognitive behavioral therapy, pharmacotherapy, and combination therapies. Settings of investigation include the United States, the Netherlands, Chile and Uganda.

Technical modeling approaches have ranged from a time varying discrete-event simulation model (depression) to first order Monte Carlo simulation (panic disorder). In addition to applied analyses assessing the comparative effectiveness (health and economic outcomes) of different strategies for personality disorders, we also use model-based methods to assess the value of information of further research on psychotherapy interventions. This project draws on primary economic and clinical data from the largest existing treatment study on personality disorders.