A better understanding of the economic burden of diabetes constitutes a major public health challenge in order to design new ways to curb diabetes health care expenditure. The aim of this study was to develop a new cost-of-illness method in order to assess the specific and nonspecific costs of diabetes from a public payer perspective. Using medical and administrative data from the major French national health insurance system covering about 59 million individuals in 2012, we identified people with diabetes and then estimated the economic burden of diabetes. Various methods were used: (a) global cost of patients with diabetes, (b) cost of treatment directly related to diabetes (i.e., 'medicalized approach'), (c) incremental regression-based approach, (d) incremental matched-control approach, and (e) a novel combination of the 'medicalized approach' and the 'incremental matched-control' approach. We identified 3 million individuals with diabetes (5% of the population). The total expenditure of this population amounted to €19 billion, representing 15% of total expenditure reimbursed to the entire population. Of the total expenditure, €10 billion (52%) was considered to be attributable to diabetes care: €2.3 billion (23% of €10 billion) was directly attributable, and €7.7 billion was attributable to additional reimbursed expenditure indirectly related to diabetes (77%). Inpatient care represented the major part of the expenditure attributable to diabetes care (22%) together with drugs (20%) and medical auxiliaries (15%). Antidiabetic drugs represented an expenditure of about €1.1 billion, accounting for 49% of all diabetes-specific expenditure. This study shows the economic impact of the assumption concerning definition of costs on evaluation of the economic burden of diabetes. The proposed new cost-of-illness method provides specific insight for policy-makers to enhance diabetes management and assess the opportunity costs of diabetes complications' management programs.
Keywords: Cost of illness; Diabetes; Econometrics; Health administrative databases.