Meeting Brochure and registration form      SMDM Homepage

Tuesday, October 23, 2007 - 9:45 AM
E-6

EQUITY WEIGHTING IN COST-EFFECTIVENESS ANALYSES

Ahmed M. Bayoumi, MD, MSc and Jeffrey Hoch, PhD. St. Michael's Hospital, University of Toronto, Toronto, ON, Canada

Purpose: Although equity is an important consideration in resource allocation decisions, methods to incorporate equity into economic analyses have not been fully developed. Because equity-weighted health effects conflate survival, preferences, and values, we were interested in exploring equity-weighting of the societal willingness-to-pay (WTP) threshold using mathematics and a worked example. Methods: We postulated a equity weighting function for health outcomes, f, such that Q represents the unweighted measure of incremental health effect and f(Q) represents the equity-weighted health effect. We similarly defined g, λ, and g(λ) as the equity weighting function for the willingness-to-pay threshold, the unweighted, and equity-weighted WTP threshold, respectively. We examined the relationship between f(Q) and g(λ) using the incremental net benefit approach. To determine equity weights, we used results from a conjoint analysis study which measured societal preferences for resource allocation decisions and defined a plausible reference case, for whom no equity weighting would be desired (λ=g(λ)). We next examined the allocation decisions for three interventions: sildenafil for erectile dysfunction, bevacizumab for metastatic colon cancer, and enfuvirtide for advanced HIV with published incremental cost-effectiveness ratios of $11 200, $195 000, and $69 500 / quality adjusted life year (QALY), respectively. We examined what the allocation decision would be at different WTP thresholds with and without equity weighting. Results: We demonstrated algebraically that the ratio of equity-weighted to unweighted health-effects is equivalent to the ratio of equity-weighted to unweighted WTP thresholds. The equity weighting factors for sildenafil, bevacizumab, and enfuvirtide were 0.11, 2.6, and 1.2. At a WTP threshold for the reference case of $50,000/QALY, the decisions for sildenafil changed from No to Yes, for bevacizumab stayed at No, and for enfuvirtide changed from No to Yes. At a reference case WTP threshold of $75,000/QALY, the decision for sildenafil was still No, for bevacizumab was Yes, and for enfuvirtide was Yes. Conclusions: Equity weighting the WTP threshold is feasible and potentially offers decision makers more transparent and defensible processes, particularly where equity concerns conflict with efficiency. The scaling factor for λ is identical to the QALY weighting factor, independent of the functional form of the equity weighting function. The main limitations to applying this approach are measurement issues and the need for consensus regarding a reference case.