LAUSR

Abstract Objective: To estimate the avoided costs associated with reductions in end-stage kidney disease (ESKD), certain CV events (non-fatal myocardial infarction [MI], non-fatal stroke, hospitalization for heart failure [HHF]), and renal and CV death for patients treated with canagliflozin versus placebo, based on CREDENCE trial results. Methods: Renal (including ESKD) and CV events averted, based on the differences in adjusted rates of events between the canagliflozin and placebo arms in CREDENCE, were projected to the proportion of the members of a managed care organization (MCO) fitting the inclusion criteria in CREDENCE (i.e. diabetic nephropathy, at least 30 years old). The number of events averted for the population was multiplied by the unit-cost of the event, extracted from a targeted literature review, to obtain costs avoided per member per year (PMPY). One-way sensitivity analysis provided a range for the cost avoided PMPY, based on variations in the events averted, unit cost and size of the projected population. Results: Costs avoided PMPY were $2.92 for ESKD with a range of $1.28–$4.20. Costs avoided PMPY were $0.54 (-$0.28–$1.16) for non-fatal MI, $0.30 (-$0.22–$0.65) for non-fatal stroke, $1.56 ($0.80–$2.11) for HHF, $0.06 ($0.05–$0.07) for renal death, and $0.51 ($0.00–$0.91) for CV death. For non-fatal MI and non-fatal stroke, the lower bound of the range is interpreted as an incremental cost. Conclusions: Positive costs avoided for each of the outcomes considered were predicted in the main analysis, with ESKD as the outcome predicted to have the greatest costs avoided at $2.92 PMPY.