Abstract
BACKGROUND: Hemoglobin and hematocrit are the 2 most common biomarkers used to identify anemia in clinical settings, but their results do not always agree. OBJECTIVES: To examine agreement between hemoglobin and hematocrit in identifying anemia among children aged 1 to <5 y and pregnant persons. METHODS: Pregnant persons and children aged 1 to <5 y with hemoglobin and hematocrit results from the same whole blood sample in National Health and Nutrition Examination Survey (1999-2020) were included. We used the Centers for Disease Control and Prevention anemia cutoff values for children, pregnancy status, trimester, and smoking adjustments. We examined concordance of anemia, sensitivity, and specificity among those with anemia based on ≥1 test overall and by race/ethnicity, sex, and income level. Cohen's kappa was used to measure concordance. RESULTS: Analytic samples included 7052 children and 1437 pregnant persons, of whom 1119 had trimester data. Among children, anemia prevalence was 3.7% [95% confidence interval (CI): 3.1, 4.3] based on hemoglobin and 5.5% (95% CI: 4.7, 6.3) based on hematocrit. Among pregnant persons, anemia prevalence was 7.7% (95% CI: 5.9, 9.5) based on hemoglobin and 12.4% (95% CI: 10.1, 14.6) based on hematocrit. Kappa scores overall and by sociodemographic characteristics ranged from 0.64 to 0.75 (moderate concordance) among children and from 0.53 to 0.78 (weak to moderate concordance) among pregnant persons. Among those with anemia on ≥1 test, 53.5% of children and 61.5% of pregnant persons had anemia based on both tests. CONCLUSIONS: We found substantial discordance between the 2 biomarkers; ∼50% of children and 40% of pregnant women were identified by only 1 of the 2 biomarkers. Because hemoglobin and hematocrit may be used interchangeably in the clinical setting, individuals with anemia may be missed, not receive treatment, and therefore be at higher risk of adverse pregnancy, birth, and developmental outcomes.