Abstract BACKGROUND AND AIMS Renal anemia management remains challenging for the nephrologist. Chronic inflammation, EPO resistance, uremia, malnutrition, iron deficiency, dialysis modalities, bleeding, infections, hydration status all interfere with erythropoiesis and erythrocytes half-life. Since 2013, >100 clinics in the FMC Nephrocare network have been using the Anemia Control Model (ACM), an artificial-intelligence based, certified medical device, decision support tool, to assist nephrologists in optimizing the monthly ESA and iron dose based on the specific characteristics of the patient and the amount of ESA dose already taken in the last 90 days. Based on this widespread experience, we sought to evaluate the real-world effectiveness of the use of ACM to facilitate hemoglobin target achievement among dialysis patients. METHOD In this historical cohort study, we analyzed electronic health records (EHR) of adult patients receiving in-center hemodialysis therapy in Fresenius Medical Care European dialysis clinics between 1 January 2015 and 31 December 2018 registered in the European Clinical Database (EuCliD®). We compared the rate of target achievement between patients treated with the ESA dose, suggested by ACM, against three control groups of patients treated in clinics without ACM and stratified by different target achievement rates for hemoglobin (Hb in the range of 10–12 g/dL or > 12 g/dL without ESA administration) (Fig. 1). To decrease indication bias, we matched patients belonging to the four comparison groups by a propensity-score assessing the likelihood of receiving the same dose suggested by ACM based on patients’ characteristics (i.e. demographic, comorbidities, anemia-related biomarkers and dialysis vintage). Centers where Hb target achievement < 70% were classified as Tier 1 centers; centers where Hb target achievement was between 70% and 80% were classified as Tier 2 centers, and centers where Hb target achievement was >80% were classified as Tier 3 centers. Performance was evaluated in the month before the index date for all the patients in any given clinic. For each matched group, the index date corresponds to the date of dose suggestion (ACM group) or to the last Hb assessment date (for all matched patients belonging to other three control groups). The study endpoints were Hb target achievement, inappropriate ESA administration (ESA administration despite Hb > 12 g/dL) and severe anemia (Hb < 9 g/dL). All endpoints have been evaluated in the month following the index date. RESULTS After matching, we obtained four groups accounting for 8 5512 patient-months each. There were negligible differences in patients’ characteristics across the four matched groups (Table 1). In the ACM group, the Hb target achievement rate was 87.52/100 patient-months 95% confidence interval (95% CI): 86.89–88.15], inappropriate ESA use rate was 4.13/100 patient-months (95% CI: 3.99–4.27) and the incidence of severe anemia was 2.14/100 patient-months (95% CI: 2.04–2.24); the Hb target achievement rate ratio (95% CI) was 1.18 (1.17–1.2), 1.08 (1.07–1.09), 1.01 (1–1.02) for ACM versus Tier 1 centers, ACM versus Tier 2 centers and ACM versus Tier 3 centers, respectively. Rate ratio for inappropriate ESA administration were 0.37 (0.36–0.39), 0.49 (0.47–0.51), 0.66 (0.64–0.69) for ACM versus Tier 1 centers, ACM versus Tier 2 centers and ACM versus Tier 3 centers, respectively. The risk ratios for severe anemia were 0.43 (0.41–0.46), 0.65 (0.62–0.7), 0.95 (0.89–1.02) for ACM versus Tier 1 centers, ACM versus Tier 2 centers and ACM versus Tier 3 centers, respectively. CONCLUSION In this matched historical cohort study, we observed an increased hemoglobin target achievement rate, decreased inappropriate ESA usage and a decreased incidence of severe anemia among patients treated according to ACM suggestions. Target achievement rate among patients treated according to ACM suggestions was 87%. The magnitude of the effect in favor of ACM was larger in centers with a lower target achievement rate.
Read full abstract