#2652 Hemoglobin excursion outside of the target range and their associations with all-cause mortality in hemodialysis patients

Abstract

Abstract Background and Aims Based on earlier analyses (Mermelstein et al., ASN Renal Week 2023), where we found an improved survival with hemoglobin levels above the currently targeted range of 10 to 11 g/dL quantified as an area under the excursion curve, we hypothesized an independent association between all-cause mortality hazard ratio and a) the number of excursions, and b) the number of days outside of the target range. Method We explored this relationship in a retrospective cohort study of incident hemodialysis patients receiving erythropoietin stimulating agent (ESA) commencing treatment Fresenius Kidney Care clinics in the US. We included patients who were given a long-acting erythropoietin stimulating agent within 90 days of dialysis initiation, considered the first six months as baseline and the following 18 months as the follow-up period. Only patients with at least 5 hemoglobin values over the first 75 days were included. We calculated the ‘area under the curve’ (AUC) above and below the target hemoglobin range of 10-11 g/dL for each patient during the baseline period, as well as days below and above range, and number of excursions below and above range. Cox proportional hazard models additionally adjusted for age, race, sex, diabetes, serum albumin, and phosphate were developed, and the hazard ratio estimates depicted in a forest plot. Results We studied 62,181 patients (57% male, 39% diabetic, mean baseline hgb 10.42 ± 0.72 g/dL, baseline AUC above target 44.36 ± 49.44 g/dL * days and 78.06 ± 86.26 g/dL * days below target). Forest plot of the HR estimates showed an increased risk of death with AUC below the lower limit of the target range (i.e. 10 g/dL), which also held true both for number of excursions and days exposed to lower hgb levels). Conversely, AUC, conversions and days spent at hgb levels higher than the upper limit of the target range (i.e. 11 g/dL), associated with a significant survival advantage. Conclusion In an effort to validate the robustness of our previously shown association between higher hgb levels and survival advantage, and to corroborate the accuracy of the utilized AUC quantification method, we separated both components of the AUC, namely into time and magnitude and quantified the independent associations with the risk of death. This allowed to discern between the associations of both to the risk of death. Consistent with the earlier results, avoiding levels below 10 g/dL and aiming for levels above the upper limit of 11 g/dL, seems to be beneficial for the anemia management of incident hemodialysis patients treated with a long-acting ESA.