OBJECTIVES

Cost-effectiveness is inferred when appropriate ICERs are below threshold WTP values. Economic evaluation literature clearly defines an appropriate ICER. One often sees otherwise apparently reputable research fall short of standards for calculating ICERs and thereby potentially misleading clinicians/policy makers regarding economic efficiency. This work overlays appropriate ICER calculation on a previous economic evaluation (Herman, 2005) of the DPP and indicates just such an example where active interventions’ (metformin- MET and lifestyle- LS) ICERs are inappropriately calculated versus the same placebo (PBO) comparator, misleadingly concluding as the ICERs were less than presumed WTP thresholds, that both active interventions were generally cost-effective. This research applies appropriate methods to calculate appropriate ICERs to form correct conclusions.

METHODS

We were able to calculate relevant cost and efficacy parameters for the three interventions from Herman for their overall analysis and for the 13 analytical versions from data reported. Interventions were ranked by efficacy to determine if any were dominated; ICERs were calculated between remaining alternatives and compared to thresholds.

RESULTS

In all 14 versions of the analysis MET is dominated/extended dominated, implying that its ICER vs. PBO is inappropriate to calculate. In all cases, LS is cost-effective compared to PBO. The overall sample ICER is $1114/QALY and the LS ICER never exceeded $12,000/QALY in any subgroup.

CONCLUSIONS

Based on these data, contrary to the Herman DPP analysis, MET is never cost-effective in diabetes prevention. It may be claimed that the MET/PBO ICER is appropriate to calculate in the absence of a LS intervention (some patients may reject it). However, the interventions were not studied in such patients (they would likely have refused randomization or withdrawn if allocated to the LS arm) and data from the study cannot pertain - while LS would clearly not be CE for such patients, MET may not be either.