1272-P: Population Modeling of Clinical and Economic Impact of the Removal of Sulfonylurea Therapy

Introduction: Advances in decision support have allowed for simulated medical decision making on a population. These models allow for simulation of policy implementation in disease management. Avoidance of high insulin therapy is frequently mentioned as a preferred strategy for the management of type 2 diabetes. Methods: Using the GlucosePATH decision support software, a sample of 368 patients drawn from the software’s database was clinically and economically optimized for scenarios matching the full therapeutic options, a scenario excluding sulfonylurea and glinides (NoSU), and a scenario excluding high dose insulin, sulfonylurea and glinides (LowInsulin). Results: Baseline patient data showed an A1c of 8.7% with mean monthly cost of $28 on standard commercial insurance, $84 on basic commercial insurance, and $354 on Medicare. Optimized with full options resulted in an estimated A1c 7.0% on standard, 7.2% on basic, and 7.5% on Medicare. Monthly cost was on $73 on standard, $168 on basic, and $371 on Medicare. Both the NoSU and LowInsulin scenarios did not show significant difference for basic or standard commercial insurance in terms of cost and A1c. In Medicare, A1c was unchanged for the NoSU model, but increased to 7.8% (+0.3%, p=0.01) in the LowInsulin model. Cost increased to $412 in NoSU model(+$41, p=0.004) and $445 in LowInsulin model (+$64, p=0.015). In the basic commercial and Medicare plan, TZD and insulin use increased. Use of incretin and SGLT2 agents did not demonstrate significant change. Discussion: Traditional commercial models allow for near ideal implementation of modern diabetes strategies in terms of patient cost, challenges develop when coinsurance or shared-risk models are implemented. It will be important to be able to measure benefit in longitudinal economic terms in addition to laboratory markers.