Only limited data are available in France on the incidence and health expenditure of type 2 diabetes. The objective of this study, based on national health insurance administrative database, is to describe the expenditure reimbursed to patients newly treated for type 2 diabetes and the proportion of expenditure attributable to diabetes. The study is conducted over a 6-year period from 2008, the year of incidence of treated diabetes, to 2014. Type 2 diabetic patients aged 45 years and older are identified on the basis of their drug consumption. To estimate expenditure attributable to diabetes, a matched control group is selected among more than 13 million beneficiaries over 44 years old not taking antidiabetic treatment. The expenditure attributable to diabetes is estimated by two methods: simple comparison of reimbursed health expenditure between both groups, and a difference-in-differences method including control variables. The cohort of incident type 2 diabetic patients comprises 170,013 patients in 2008. Mean global reimbursed expenditure is €4700 per patient in 2008 and €5500 in 2015. Expenditure attributable to diabetes, estimated by direct comparison with controls, is €1500 in the first year. We, thus, observe a decrease in the following year due to decreased hospitalisations, and then expenditure increase by an average of 7% per year to reach €1900 in the eighth year after the initiation of treatment.

about the real-world treatment patterns and patient characteristics at time of insulin initiation, even if this has substantial implications for current and future HTA assessments as insulin therapy is typically the standard of care in later T2DMstages. We aimed to collect this data. METHODS: T2DM insulin-naïve patients (at least one inpatient/two outpatient diagnoses by diabetologists, no insulin in 12month pre-index period) were identified in a German claims dataset (01/01/2012-31/12/2016). We evaluated the prescribed antidiabetic drugs (ADs) before and after insulin initiation as well as the initiated insulin regimen. For a patient subgroup participating in a disease management program, clinical parameters at baseline could be investigated. RESULTS: 284,878 T2DM-patients were identified. Of these 27,340 (9.6%) initiated insulin treatment (mean age: 72.2; 51.4% female). Most common ADs in the 6 months before insulin initiation were metformin (54.0%), DPP-4i (37.6%) and sulfonylureas (29.5%). 23.2% did not receive any AD prescription previously. 10,953 patients (40.1%) started insulin without concomitant other ADs. Their most common initiated insulin regimens were fast-acting (27.7%), longacting (25.5%) and ICT (intensified therapy, 25.5%). Of 16,387 patients with additional ADs, 14.9% received only metformin (insulin regimens: 42.0% long-acting, 20.3% fast-acting, 18.1% ICT), and 23.4% received metformin plus at least one further AD (insulin regimens: 72.3% long-acting, 9.4% fast-acting, 7.7% ICT). Clinical parameters were available for 15,173 patients. Mean weight/BMI of patients was 85.8kg (SD:18.9)/30.6 (SD:6.1). This is 9.5 kg heavier than currently assumed in German AMNOG processes. Mean Hb1Ac-value at baseline was 8.4 (SD:1.8). CONCLUSIONS: Real-world patient characteristics, in particular patients' weight, deviate substantially from assumptions currently used in German HTA processes. Furthermore, we identified many patients receiving insulin treatments not in line with current treatment guidelines.

No abstract

A 3 4 7 -A 7 6 6 on the comparison between expenditures of the diabetic and non-diabetic populations. Then, we observe over 8 years the dynamics of the costs attributable to diabetes. Results: We identify 220,000 persons who were newly treated for type 2 diabetes in 2008 (45 years and older). This provides an incidence rate of 1,160 per 100,000 insured persons. Among people with newly treated diabetes, 52% are men. The mean age is 64 years (62 years for men and 65 years for women). Based on this cohort, we aim to assess the change over time of overall cost and costs attributable to diabetes due to the natural history of the disease. ConClusions: Describing the impact of the natural history of the disease on the cost of type 2 diabetes on such a large cohort of patients will give policy maker a better understanding of how diabetes, its complications and health care affect expenditures over time.