Abstract P062: Impact of Second-line Antidiabetic Treatments on the Risk of Microvascular Complications: Analysis of Real-world Data for 94,685 Patients Type 2 Diabetes

Abstract

Objective: To compare the risk of microvascular complications associated with second-line anti-diabetic medications by analyzing the national health insurance claim database. Methods: The Korean National Health Insurance (NHI) Service is the single payer which covers all Korean citizens and residents. We identified all patients who used sulphonylurea (SU), dipeptidyl peptidase-4 inhibitor (DPP4I), or thiazolidinedione (TZD) as a second-line oral anti-diabetic medication added to metformin (MET) therapy between January 2011 and June 2015 in the NHI database. Cox’s proportional hazard regression model was used to estimate hazard ratio and its 95% confidence interval for developing microvascular complications according to the types of second-line medications. Age, gender, calendar index year, duration of metformin treatments, comorbidities of hypertension and dyslipidemia, and Charlson Comorbidity Index were adjusted as potential confounders. Results: A total of 94685 initiators of a second-line add-on to MET of either a SU (n = 28887), DPP4I (n = 60780) or TZD (n = 5018) were identified. Diagnoses of diabetic retinopathy (n=7139), diabetic nephropathy (n=5290), and diabetic neuropathy (n=4265) were identified from the NHI claim database over a mean of 1.3 years of follow-up. Compared to the SU+MET, adjusted hazard ratio (95% confidence interval) for diabetic retinopathy was 1.03 (0.96-1.10) for the DPP4I+MET and 1.26 (1.10-1.44) for the TZD+MET. Adjusted hazard ratio (95% confidence interval) for diabetic nephropathy was 1.30 (1.19-1.42) for the DPP4I+MET and 1.63 (95% CI: 1.40-1.91) for the TZD+MET, compared with the SU+MET. Adjusted hazard ratio (95% confidence interval) for diabetic neuropathy was 0.78 (95% CI: 0.71-0.85) for the DPP-4i+MET and 0.81 (95% CI: 0.66-0.99) for the TZD+MET, compared with the SU+MET. Conclusion: In this analysis of nationwide real-world data, DPP4I+MET and TZD+MET therapies were associated with higher risk of diabetic retinopathy and nephropathy, but lower risk of diabetic neuropathy compared to SU+MET therapy.