A new simulation model to estimate the long-term health and cost outcomes of interventions for people with type 1 diabetes

Abstract

OBJECTIVE: To develop a U.S.-based microsimulation model for assessing the cost-effectiveness of interventions to manage type 1 diabetes.

RESEARCH DESIGN AND METHODS: We developed risk equations for 14 diabetes-related complications and mortality, 12 risk factor progression equations, and one equation for utilities associated with 14 complications using data from the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) studies and the Epidemiology of Diabetes Complications (EDC) study. We integrated all equations into a simulation model. We conducted internal and external validation and demonstrated the utility of the model using a real-world example. Main model-generated outcomes included cumulative incidence of diabetes-related complications, life years, quality-adjusted life years, medical costs, and incremental cost-effectiveness ratios.

RESULTS: The model generates long-term clinical and economic outcomes from changes in risk factors of type 1 diabetes complications. Internal validation comparing modeled outcomes to observed data used to develop the model yielded good prediction accuracy, with mean absolute percentage error across all complications of 9% and correlation of cumulative failure rates above 0.9. External validation results were mixed, with occurrence of slight under- or overprediction across complications and studies. We illustrated the model with a case study estimating the effects of expanding the use of an insulin pump with continuous glucose monitoring to all people with type 1 diabetes.

CONCLUSIONS: Our new comprehensive type 1 diabetes simulation model can generate valid and accurate results for assessing the long-term cost-effectiveness of interventions to manage type 1 diabetes in the U.S.