To summarize the methodological approaches used in published decision-analytic models evaluating interventions for acute stroke treatment, to highlight key components of decision-analytic models of stroke treatment, and to discuss challenges for developing stroke decision models.
A review of the published literature was performed using Medline, to identify studies involving mathematical decision models to evaluate interventions for acute stroke treatment. Articles were analyzed to determine key components of a stroke model and to note areas in which data are lacking.
We identified 13 published models of acute stroke treatment. These models typically possessed a short-term treatment module and a long-term post-treatment module. The following aspects of economic modeling were found to be relevant for developing a stroke model: modeling approach and health state; health state transition probabilities; estimation of short-term, long-term, and indirect costs; health state utilities; poststroke mortality; time horizon; model validation; and estimation of parameter uncertainty.
Data gaps have limited the development of economic models in stroke to date. In order to more accurately assess the long-term incremental impact of a new treatment of stroke, future research is needed to address these data gaps. We recommend that the complexity of models for examining the cost-effectiveness of an acute stroke treatment be kept to a minimum such that it can incorporate the currently available data without making a large number of assumptions around the data.