OBJECTIVES: The management of renal anemia remains challenging for nephrologists. The Anemia Control Model (ACM), a certified medical device based on Artificial Intelligence, assists nephrologists in optimizing dosage of erythropoiesis-stimulating agents (ESA) and iron for their patients. We evaluated the effectiveness of the ACM concerning hemoglobin (Hb) targets achievement and ESA inappropriate use.

METHODS: This historical cohort study analyzed electronic health records of adult patients receiving in-center hemodialysis in Fresenius Medical Care dialysis clinics (2015 and 2018). We compared Hb target achievement, inappropriate ESA use and severe anemia between ACM guided therapy and standard of care. Non-ACM patients were allocated in 3 reference groups depending on historical Hb target achievement rates of their clinic: A, international benchmark (<70% target achievement rate); B, above average (70-80%); C, excellent (>80%). We used propensity score matching to account for covariate imbalance across comparison groups.

RESULTS: After matching, each study group included 85512 patient-months with no covariate distribution imbalance. In the ACM arm, Hb target achievement rate was significantly greater than group A (Rate Ratio=1.18, 95%CI: 1.18–1.19), B (Rate Ratio=1.08, 95%CI: 1.08–1.09) and C (Rate Ratio=1.01, 95%CI: 1.01–1.02). Risk ratios for inappropriate ESA administration were 0.37 (0.35–0.38), 0.49 (0.47–0.51), 0.66 (0.63–0.69) for ACM versus group A, B and C, respectively. The risk ratios for severe anemia were 0.41 (95% CI: 0.39–0.44), 0.64 (95% CI: 0.61–0.68), 0.94 (95% CI: 0.88–1.00) for ACM versus group A, B and C, respectively. Median ESA consumption was significantly smaller in the ACM group.

CONCLUSIONS: ACM was associated with increased Hb target achievement rate, reduced inappropriate ESA usage, reduced incidence of severe anemia and reduced ESA consumption compared to standard of care.