Table 1.
Characteristics of patients at baseline.
Table 2.
Anemia related factors among ferritin categories.
Fig 1.
Relationship between Hb and ferritin levels among the patient groups, as stratified by serum ferritin levels.
A significant correlation was found between Hb and ferritin levels only in patients with ferritin levels <50 ng/mL (p<0.001, R = 0.89). No significant correlation was found between Hb and ferritin levels in patients with 50–100, 100–300, and ≥300 ng/mL ferritin.
Fig 2.
Factors associated with a high ERI (≥280).
In the LASSO analysis, high doses of intravenous iron, low serum calcium levels, female sex, low serum albumin levels, angiotensin-converting enzyme inhibitor (ACE-I)/angiotensin receptor blocker (ARB) use, and low serum pre-albumin levels were selected significant predictors of high ERI scores.
Fig 3.
ESA dose and ERI in patients with or without composite events.
(A) The ESA dose at the time of the composite event was significantly higher than that at the end of the observational period in patients without composite events. (B) The mean ESA dose (until the composite event) was significantly higher than the mean ESA dose during the observational period in patients without composite events. (C) The ERI at the time of the composite event was significantly higher than the ERI at the end of observational period in patients without composite events. (D) The mean ERI value (until the composite event) was significantly higher than the mean ERI value during the observational period in patients without composite events.
Fig 4.
The cut-off values for the ESA dose and ERI for composite events.
The cut-off values for the ESA dose (A) and ERI (C) at the time of the composite event were 5250 and 309.28 IU/week, respectively. The cut-off values for the mean ESA dose at the composite event (B) and mean ERI (D) until the composite event were 4,125 and 387.69 IU/week, respectively. According to the Hanley and McNeil method, the area under the curve for the ERI value at the time of the composite event was significantly higher than that of the ESA at the time of the composite event.
Fig 5.
Factors associated with composite events.
In the LASSO analysis, the ERI at the time of the composite events, low prealbumin, albumin and high density lipoprotein cholesterol (HDL-CHO) levels, a history of CVD, use of vitamin D at the time of treatment, and high hCRP levels were the predictors of composite events in maintenance hemodialysis patients.
Fig 6.
Correlation between mean or median levels of parameters and composite events.
In the LASSO analysis, the mean ERI during the study period, lower mean prealbumin levels (odds ratio, 0.9708), lower mean Hb levels, history of CVD, higher mean hCRP levels, lower mean albumin levels, and shorter median treatment time were predictors of composite events in maintenance hemodialysis.
Fig 7.
A: Relationship among the combined categories of ERI, ferritin level and composite events. The risk of composite events in patients with a high ERI (≥280) and a high ferritin (≥100 ng/mL) level was significantly greater (HR, 2.09; p = 0.033) than that of patients with a high ERI (≥280) and a low ferritin (<100 ng/mL) level. B: Relationship among the ERI value, intravenous iron dose and composite events. The risk of composite events in patients with a high ERI value (≥280) treated with a high dose of intravenous iron was significantly greater (HR: 1.76, p = 0.032) than that in patients with a high ERI value (≥280) treated with a low dose of intravenous iron. Frn: ferritin; low intravenous iron: treated with intravenous iron (<50 mg/week); high intravenous iron: treated with intravenous iron (≥50 mg/week).