Abstract 13181: Horizon and Perspective: Critical Choices in Estimating Cost-Effectiveness of Cardiovascular Prevention

Abstract

Introduction: In order to capture the effects of preventing death, cost-effectiveness models must quantify extra years of life gained and healthcare costs incurred during those years. Different approaches to modeling these effects have been described, but their impact on estimates of cost-effectiveness of coronary heart disease (CHD) prevention strategies is unclear.

Hypothesis: We hypothesize that estimates of the cost-effectiveness of coronary heart disease prevention are sensitive to choice of time horizon and inclusion of healthcare costs unrelated to CHD.

Methods: We used an established CHD decision model to estimate the incremental cost-effectiveness ratio (ICER), in $/quality-adjusted life-years (QALYs) of a hypothetical cardiovascular prevention intervention that reduces CHD event rates by 30% and that costs $0.50/day. Focusing on two scenarios - 45 yo men and 65 yo women, both with a 10-year CHD risk of 5% - we varied time horizons and used different strategies for inclusion of healthcare costs unrelated to CHD.

Results: The cost-effectiveness of the intervention varied dramatically, from very cost-effective (<$10K/QALY) to not cost-effective ($100K/QALY), depending upon the strategy used (Table). Time horizon was an important driver, with shorter horizons not allowing as many life-years to accumulate after a prevented event. Using a lifetime horizon but applying the treatment effect for only the first 10 years gave estimates that were more stable across scenarios. Including direct medical costs unrelated to CHD increased the ICER, especially when age-dependent annual costs were included. An approach that stopped counting annual healthcare costs at age 70 provided a smaller ICER “penalty” for the final years of life, particularly for scenarios starting at older ages.

Conclusions: Without a standard approach to modeling time horizons and perspectives, direct comparisons of the ICER for different prevention strategies are less meaningful.