Analytical and Biological Variation in Measures of Anemia and Iron Status in Patients Treated With Maintenance Hemodialysis

Abstract

To make informed decisions in dosing erythropoiesis-stimulating agents and intravenous iron therapy, clinicians must determine whether differences between current and previous test results for anemia and iron status markers reflect expected variation, a significant change, or an actual trend. Prospective observational cohort study. 30 patients undergoing thrice-weekly in-center hemodialysis. Within-patient biological variations in hemoglobin (Hb) level, hematocrit (Hct), reticulocyte Hb content, transferrin saturation (TSAT), and ferritin level were determined over 12 consecutive treatment days. We separately measured same-sample analytical variation and within-patient biological variation (coefficient of variation), then calculated the number of sampling days needed to determine the true or homeostatic value for each analyte with 95% probability. We also evaluated whether results differed among the first, second, and third dialysis days of the week. Biological variation differed by analyte. Hb level (4.0%), Hct (4.0%), and reticulocyte Hb content (4.8%) showed much lower variation than TSAT (38.2%) or ferritin level (15.1%). Analytical variation ranged from 2.0%-6.9% for all analytes. We found that one sample day would be sufficient to establish the true mean Hb level or Hct within a level of closeness+/-20% and 95% probability. For the same levels of closeness and probability, one sample day would be needed for reticulocyte Hb content, 15 for TSAT, and 3 for ferritin level. No pairwise comparison for any of the 5 analytes yielded a significant difference between results obtained on the first, second, or third dialysis day of the week. These findings may not apply to other patient populations. Low biological variation renders Hb level, Hct, and reticulocyte Hb content, but not TSAT and ferritin level, suitable for trend analysis using results from 2 successive samples. TSAT and ferritin test results, unlike reticulocyte Hb content, have limited value in evaluating changes in iron status within individual hemodialysis patients.