OBJECTIVES : Effectiveness of antidiabetic drugs is related to adherence to the drug. Due to the design of clinical trials, adherence tends to be better compared to real life. In health economic models mostly, the efficacy of clinical trials is used. In the new version (9.5) of the IQVIA-Core-Diabetes-Model (CDM) an adherence module (AM) is included. The current study aims to explain how this functionality operates and how its use influences health economic outcomes.

METHODS : The AM requires three inputs per treatment 1) adherence rate, 2) HbA1c change per 10% change in adherence rate, 3) BMI change in Kg/m2 per 10% change in adherence rate. Adherence rate directly impacts treatment costs. All inputs influence the first-year treatment effect of each treatment independent of the treatment line. Baseline characteristics and treatment effects from the observational EDGE-study comparing metformin+vildagliptin (M+V) with metformin+sulphonylurea (M+S) were employed. Two adherence rates (100% and 80%) were tested. The impact of adherence per 10% change on HbA1c (-0.08%-points) and BMI (DPP4= -0.0576 Kg/m^2; SU: 0.0072 Kg/m^2) was taken from literature. The economic analysis employed UK 2018 costs and applied a 3.5% annual discount rate was on costs and outcomes.

RESULTS : Diminishing adherence from 100% to 80% hardly changed life expectancy. Quality adjusted life years reduced from 6.93 to 6.91 in M+V and increased from 6.87 to 6.88 in M+S (due to more BMI decrease with lower adherence). Linked to the adherence rate, treatment costs diminish accordingly, however complication costs increase. Impact on ICER was minimal (slight increase).

CONCLUSIONS : Lower adherence did not impact the ICER significantly, because of opposite effects on BMI, the same adherence rates in M+V and M+S arms, and patients switch to next line therapy quickly. Applying different adherence rates and other BMI effects, would result in considerable impact on final outcomes.