VALUE OF INFORMATION ANALYSIS ON ERYTHROPOIESIS-STIMULATING AGENTS IN ANEMIC HEART FAILURE - GAUGING THE VALUE OF FUTURE RESEARCH DIRECTIONS

Purpose: In a model-based cost-utility analysis investigating the effect of erythropoiesis-stimulating agents (ESAs) in the treatment of anemic heart failure (HF) patients, we found the incremental cost-utility ratio of ESAs compared to standard therapy without treatment of anemia to be $35,900/QALY (95% credibility interval: $14,000/QALY to infinity). The results were sensitive to ESA costs and efficacy. Because of the remaining uncertainty we sought to elucidate the potential value of additional research.

 Methods: We used a probabilistic 5-state microsimulation Markov model based on hemoglobin strata that compares standard therapy without ESA to standard therapy plus use of ESA. Input parameters were derived from HF registries and trials including STAMINA-HFP, SOLVD, EPHESUS, Medicare reimbursement rates, and the published literature. The expected value of perfect information (EVPI) regarding all parameters was computed using a 2^nd-order Monte Carlo simulation with 1,000 outside iterations that included 50,000 nested 1^st-order MC-simulations per cycle. Expected Values of Partially Perfect Information (EVPPIs) were calculated in three-dimensional analyses with 50 iterations for the distributions representing the partially perfect information and using 1,000 2^nd-order MC simulations with nested 50,000 1^st order simulations.

 Results: Considering a population of 2.5 million eligible patients, a 3% discount rate, a technology duration time of 10 years, and willingness to pay (WTP) values of $50,000/QALY, $100,000/QALY, and $200,000/QALY, the population EVPIs were $7.09 billion, $8.70 billion, and $13.72 billion, respectively. The EVPPI for the same population and WTP values are $8.85 billion, $10.72 billion, and $17.33 billion for the clinical parameters employed in modeling (1.) the natural disease progression of HF, and (2.) treatment efficacy of ESAs in anemic HF patients, while it was insignificant for both the cost and health-related quality of life parameters.

 Conclusions: Potential avenues for further increasing the certainty about cost effectiveness of ESA treatment in anemic HF patients are: (1.) conducting cohort studies to elucidate the underlying epidemiological association between hemoglobin level and morbidity and mortality, and (2.) performing randomized controlled trials to assess the efficacy of ESA treatment. However, future research to better assess the health-related quality of life and cost parameters would increase the certainty around this decision only marginally.