Antidiabetic Drugs in Patients with Chronic Kidney Disease: Trends, Adherence and Economic Outcomes

Therapeutic management of Type 2 Diabetes (T2D) patients with comorbid Chronic Kidney Disease (CKD) is challenging given prescribing restrictions on certain antidiabetic medications and limited evidence regarding their comparative effectiveness. This dissertation aimed to evaluate selection of antidiabetic medication, adherence to these medications and impact of adherence on healthcare resource use in this population. The first study assessed antidiabetic medication use over the years and across the stages of CKD using National Health and Nutrition Examination Survey (NHANES) participants from 2003-2014. Increase in metformin, Dipeptidyl Peptidase-4 (DPP-4) inhibitors and insulin and decrease in sulfonylurea and thiazolidinediones was observed. For several of these medications, changes in their use over time and across stages of CKD generally agreed with safety warnings and changing guidelines. However, guideline discordant glyburide and metformin use observed in this study warrants interventions to mitigate such use in future. The second study compared medication adherence and persistence between two medication classes: DPP-4 inhibitors and pioglitazone. In this administrative claims database study, patient initiating DPP-4 inhibitors had better adherence (Proportion of Days Covered (PDC) greater than or equal to 0.80) as compared to those initiating pioglitazone. However, out-of-pocket cost (higher with branded DPP-4 inhibitors versus lower with generic pioglitazone) was identified as an important factor driving the difference in adherence between these classes. Persistence with pioglitazone changed substantially with calendar years in line with safety warnings from U.S. Food and Drug Administration (FDA) and approval of generic products. The third study evaluated the association between antidiabetic medication adherence and healthcare resource use. Adherent patients had lower odds of diabetes-related and all-cause hospitalization in the follow-up year. While diabetes-related costs were higher, the total healthcare costs were lower for those who adhered to their medication. In conclusion, although majority of changes in prescribing practice was in accordance with the guidelines, gaps observed with respect to inappropriate diabetes medication use demands efforts towards strengthening health policies. Further, selecting one diabetes medication class over the other may affect treatment adherence and persistence which in turn is related to subsequent healthcare resources use and cost.