The Singapore Diabetes Prevention Program (DPP) Cost-Effectiveness Analysis (CEA), Re-Evaluated and Corrected: Metformin Is NOT Cost-Effective When Correct Methods Are Used
Author(s)
ABSTRACT WITHDRAWN
OBJECTIVES A 2014 Singapore-based diabetes prevention CEA of interventions of placebo (PBO), metformin (MET), each including basic lifestyle modification advice, and an intensive lifestyle alternative (LS). It concluded that LS and MET were both cost-effective, comparing each intervention to PBO and forming conclusions based on the relatives sizes of the ICERs. This is an inappropriate method. We applied proper methods of ICER calculation/interpretation to examine appropriate conclusions using the data provided in the study. METHODS Following the DPP authors, we analyzed both healthcare-system and societal perspectives. We applied proper ICER calculation methods to the DPP data (costs and QALYs): ranking interventions by efficacy (PBO, MET, then LS), eliminating dominated alternatives and suggesting which treatment was cost-effective based on the Singapore authors’ stipulated willingness-to-pay (WTP) of $53,000. Net Monetary Benefit (NMB) calculations were also performed. RESULTS Under the healthcare-system perspective MET was extendedly dominated. The LS ICER vs. PBO was $16,920. Under the societal perspective, ICERs were $8400 (MET/PBO) and $43,000 (LS/MET). NMB values at the Singapore author’s assumed WTP of $53,000 were $96,890, $97,139 and $98,694 for PBO, MET and LS for the healthcare-system perspective and $78,297, $78,743 and $79,143 for the societal. CONCLUSIONS The original paper’s authors erred in ICER calculation/interpretation. Correct ICERs and stipulated WTP value imply that LS is cost-effective from both perspectives. Proper ICER calculations/interpretations are apparently not straightforward to many analysts. NMB calculations are simpler than ICERs and had they been performed by the original authors, no error would have been made as the NMB decision criterion of maximizing NMB, implies a clear choice of LS. At a WTP of $53,000, LS is the only cost-effective intervention. Implementing policies that focus on MET inappropriately based on the original authors’ flawed conclusions would imply needless monetary and health costs compared to the LS alternative.
Conference/Value in Health Info
2021-05, ISPOR 2021, Montreal, Canada
Value in Health, Volume 24, Issue 5, S1 (May 2021)
Code
PDB4
Topic
Economic Evaluation
Topic Subcategory
Cost-comparison, Effectiveness, Utility, Benefit Analysis
Disease
Diabetes/Endocrine/Metabolic Disorders