An economic evaluation of erythropoiesis-stimulating agents in CKD.

BACKGROUND The objective was to determine the cost-effectiveness of treating anemic patients with chronic kidney disease (CKD) with erythropoiesis-stimulating agents (ESAs) to a low (9-10.9 g/dL), intermediate (11-12 g/dL), or high (> 12 g/dL) hemoglobin level target compared with a strategy of managing anemia without ESAs. STUDY DESIGN Cost-utility analysis. SETTING & PARTICIPANTS Publicly funded health care system. Anemic patients with CKD, overall and stratified into dialysis-/non-dialysis-dependent subgroups. MODEL, PERSPECTIVE, & TIMEFRAME Decision analysis, health care payer, patient's lifetime. MAIN OUTCOME Cost per quality-adjusted life-year (QALY) gained. RESULTS For dialysis patients, compared with anemia management without ESAs, using ESAs to target a low hemoglobin level is associated with a cost per QALY of $96,270. Given a lack of direct trials comparing low and intermediate targets, significant uncertainty exists between these strategies. Treatment to a high hemoglobin target was always associated with worse clinical outcomes and higher costs compared with a low hemoglobin target. Results were similar in non-dialysis-dependent patients with CKD, with a cost per QALY for a low target compared with no ESA of $147,980. LIMITATIONS Given limitations in the available randomized controlled trials, we were able to model only 4 treatment strategies, balancing the need to consider relevant targets with the requirement for accurate estimates of clinical effect. We assumed that the efficacy of the different strategies would continue over a patient's lifetime. CONCLUSIONS Using ESAs to target a hemoglobin level > 12 g/dL is associated with worse clinical outcomes and significant additional cost compared with using ESAs to target lower hemoglobin levels (9-12 g/dL). Given a lack of studies comparing low (9-10.9 g/dL) and intermediate (11-12 g/dL) hemoglobin targets for clinical outcomes, including quality of life, the most cost-effective hemoglobin level target within the range of 9-12 g/dL is uncertain, although aiming for higher targets within this range will lead to higher costs.