Health Institute, Helsinki, Finland; 3Novartis Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: To estimate the cost-effectiveness of the insulinotropic agent nateglinide vs. metformin using an epidemiologic risk model that quantifies the relationship between glucose spikes as measured by 2 hour postprandial blood glucose (2h-BG) and risks for all cause mortality, acute myocardial infarction (AMI), and stroke in diabetic patients. METHODS: We used data from the DECODE study database (N=22,474, with up 25 years of follow-up) to estimate parametric failure time models predicting the risk for death, AMI, and stroke for 2h-BG, and other CVD risk factors. The risk equations were used to develop a decision model that projected risks, costs, and years of life for up to 40 years for men and women with and without an intervention specifically to control 2h-BG (results for men with 2h-BG >11 mmol/L reported below). Costs included the intervention and the costs of CVD events. All costs are expressed in Swiss francs (CHF), and were discounted at 3%. Clinical efficacy was taken from a randomised clinical trial of nateglinide versus metformin. RESULTS: When results were projected for 15, 25, and 40 years, incremental costs were 9,137, 10,047, 10,133 CHF, respectively (1 CHF ~ $0.60). Discounted years of life saved for these same intervals were 0.15, 0.24, and 0.26. The ratios of cost per year of life saved were 59,600, 41,500, 38,400 CHF. CONCLUSIONS: Initial modeling suggests that therapy with nateglinide among individuals with elevated levels of 2h-BG reduces the risk for death and CVD events and has acceptable cost-effectiveness ratios compared to metformin.