The inequality-adjusted incremental cost-effectiveness ratio

OBJECTIVES: To adjust the incremental cost-effectiveness ratio (ICER) to capture trade-offs between maximisation of total health and impact on social inequality in health. METHODS: We used equity weights derived from a social welfare function to calculate an inequality-adjusted ICER, both indirectly and directly. We demonstrated equivalence of decision rules based on the inequality-adjusted ICER and equity-weighted net health benefit. We defined a health inequality modifier (HIM) to the cost-effectiveness threshold - the ratio of the ICER to the inequality-adjusted ICER - and illustrated its application to hypothetical treatments for 1,336 diseases in England. We used hospital admissions to estimate disease prevalence and health benefits by five social deprivation groups, assumed no social gradient in health opportunity costs and examined "high", "medium" and "low" health inequality aversion scenarios based on UK general public sample estimates. RESULTS: Assuming medium inequality aversion, the inter-percentile HIM range was 0.96-1.18, indicating that inequality adjustment would reduce the threshold by 4% at the bottom percentile and increase it by 18% at the top percentile of diseases. The threshold was reduced by at least 10% for only 2 (/1,336 = 0.15%) diseases and raised by at least 10% for 142 (/1,336 = 10.6%). High inequality aversion widened the inter-percentile HIM range to 0.93-1.31. CONCLUSION: Use of the inequality-adjusted ICER would effectively modify the cost-effectiveness threshold to consider health inequality impact. In England, the modification would rarely exceed 30%, even assuming high inequality aversion, but could exceed 10% for about one tenth of diseases assuming medium aversion.