PDB29 - Cost Utility Of Diabetes Drugs Using Hba1c As A Direct Predictor For Quality Of Life, Diabetes Complications And Mortality

Objectives:
Cost utility analyses (CUA) of diabetes treatments have traditionally been performed using HbA1c as a surrogate endpoint for diabetes complications and mortality. This study introduces a novel approach to CUA modelling of diabetes whereby blood sugar control as measured by HbA1c is used to directly predict quality of life utilities, diabetes complications and mortality.

Methods:
A microsimulation model was constructed which followed people with newly diagnosed diabetes over a period of 10 years. HbA1c level determined when a person was assumed to undergo escalation in diabetic treatment (from monotherapy with non-insulin hypoglycaemics to dual therapy to triple therapy and insulin). Information on efficacy and toxicity of therapies were derived from clinical trial data. Health care utilisation and costs were sourced from Australian government websites. Risk equations using change in HbA1c as a predictor for complications, quality-adjusted life years (QALYs) and death were derived from published data from large Australian diabetes cohorts. Probabilistic sensitivity analyses were undertaken. Two classes of drugs were investigated as alternatives to sulphonylurea when given in combination with metformin: DPP-IV inhibitors (sitagliptin, vildagliptin, saxagliptin, linagliptin, alogliptin) and SGLT-2 inhibitors (canagliflozin, dapagliflozin).

Results:
In general, the results for the CUA were similar between the two drug classes compared to sulphonylureas, with ICERs ranging from AU$40K/QALY to AU$50K/QALY. The proportion of diabetes complications dropped by 3-4%, insulin treatment was delayed on an average of 2-3 years and a drop of 1-2% in mortality was observed.

Conclusions:
This model illustrates a new way of assessing the cost utility of diabetes medications. Furthermore, it shows that both DPP-IV and SGLT-2 inhibitors represent cost-effective alternatives to sulphonylurea in combination with metformin.