Study population
A total of 240 patients received hemodialysis therapy between October 2017 and September 2019. Among them, 161 patients were eligible for the study. Twenty-nine patients were excluded due to bleeding events such as gastrointestinal hemorrhage (n = 7), obstetric hemorrhage (n = 4), and blood loss due to clotting of the dialysis filter (n = 8) during the study observation period. Additionally, patients undergoing intravenous iron treatment (n = 5) and those without data on %HRC, CHr, serum iron, serum ferritin, or TSAT (n = 5) were excluded. Thus, a total of 142 patients undergoing hemodialysis were included in the analysis (Fig. 1).
Baseline characteristics of patients
The demographic, clinical, and biochemical data of the patients are shown in Table 1. Among the 142 patients, 80 (56.3%) were men; the median age was 65 years, median dialysis duration before study enrollment was 51.8 months, and underlying cause of end-stage renal disease (ESRD) in 90 patients was diabetes (63.4%). The median %HRC was 3.1% (range, 0.5%–15.5%; Fig. 2). When patients were divided into two groups according to their median %HRC, the dialysis duration before study enrollment, iron concentration, and CHr were significantly higher in the patients with %HRC ≤3.1% than in those with %HRC >3.1%. However, there were no differences between the groups with respect to age, male sex, underlying ESRD cause, medication use, and values of ferritin, TIBC, PTH, calcium, phosphate, CRP, TSAT, and Kt/V.
Correlation between ΔHb and other variables
There was a significant negative correlation between the delta hemoglobin level (ΔHb1mo- baseline) and %HRC (r = -0.63, P < 0.001; Fig 3). Moreover, despite a significant positive correlation between ΔHb1mo- baseline and CHr (r = 0.21, P = 0.01), there were no correlations between ΔHb1mo- baseline and the concentrations of iron, ferritin, and TSAT.
Development of anemia according to %HRC
Among the 142 patients, 62 (43.7%) developed anemia at 1 month after study enrollment. The incidence of anemia was significantly higher in patients with %HRC >3.1% than in those with %HRC ≤3.1% (70.4% vs. 6.9%, P < 0.001). However, hemoglobin levels were significantly lower in patients with %HRC >3.1% than in those with %HRC ≤3.1% [9.8 (9.0 0–10.1) vs. 10.2 (10.0–10.6), P < 0.001] at 1 month after study enrollment.
Variables related to the development of anemia
The variables related to the development of anemia are shown in Table 2. In univariate analysis, %HRC and CHr were found to be significant predictors of the development of anemia. Moreover, age, male sex, use of non-calcium-based phosphate binders, and CRP levels were significantly related with the development of anemia. In contrast, the concentrations of ferritin, iron, and TSAT were not associated with the development of anemia. Specifically, %HRC remained to be a significant predictor of anemia, even after adjusting for potential confounding factors related to the development of anemia.
Comparison of variables to predict the development of anemia
To compare the ability of iron, ferritin, TSAT, %HRC, and CHr to predict the development of anemia, receiver operating characteristic (ROC) curve analysis with the area under the ROC curve (AUC) was performed. The AUC of %HRC was significantly larger than that of the other variables. ROC curve analysis with AUC was also used to obtain the optimal cutoff value of %HRC to predict the development of anemia. The best cutoff value for %HRC to predict the development of anemia was 4.3%, with a sensitivity and specificity of 67.74 (95% confidence interval [CI], 54.7–79.1) and 97.50 (95% CI, 91.3–99.7), respectively (Fig. 4).