Validation of HemoCue method and development of a revised hemoglobin cut-off to detect anemia in children aged 6–24-months

Background Accurate hemoglobin (Hb) measurement is essential for determining anemia prevalence at the population level. The HemoCue method, commonly used in resource poor settings, has been shown to overestimate Hb levels compared to the direct cyanmethemoglobin (DCM) method, leading to an underestimation of anemia. We aimed to validate the HemoCue method against the DCM method and establish an optimal cut-off value for diagnosing anemia in children under two. Methods Data were collected from children under two years of age as part of the Etiology, Risk Factors, and Interactions of Enteric Infections and Malnutrition and the Consequences for Child Health and Development study in Bangladesh, conducted in Mirpur, Dhaka, from February 2010 to February 2017. Using HemoCue 201 instrument and DCM method, Hb levels were measured in 846 venous blood samples collected from 589 children. Quadratic regression was explored to derive a correction factor; however, substantial dispersion in HemoCue measurements prevented establishment of a reliable correction model. Youden’s Index was used to identify the optimal cut‑off. While it provided a strong threshold, we prioritized achieving at least 70% sensitivity and specificity to confirm balanced diagnostic performance. Results The median (IQR) Hb concentrations estimated by HemoCue and DCM were 11.6 (10.5, 12.5) g/dL and 10.8 (10.1, 11.5) g/dL, respectively. Using a cut-off of <10.5 g/dL for anemia, the prevalence was significantly higher with the DCM method (35.2% vs. 23.2%, p < 0.001). Considering DCM as the gold standard, the sensitivity and specificity of the HemoCue method were 93% and 53%, respectively. In the test/train analysis, the newly identified cut-off of 11.00 g/dL, indicated an anemia prevalence of 34.2% in the testing dataset, with a sensitivity of 84% and specificity of 69%. Conclusion This study highlights substantial variability between HemoCue and DCM hemoglobin measurements, which prevented development of a reliable correction factor. The proposed HemoCue cut-off of 11.0 g/dL may improve anemia detection in children under two years of age when compared with the currently recommended threshold.