# Comparing HemoCue® and Quantitative Buffy Coat® and Coulter Counter-measured haemoglobin concentrations in African children with acute uncomplicated malaria: a Bland–Altman analysis

## TL;DR

This study compares three methods for measuring hemoglobin in children with malaria and finds that the HemoCue® device has large variability compared to lab methods.

## Contribution

The study provides a Bland–Altman analysis comparing HemoCue® with Coulter Counter and QBC® in African children with malaria.

## Key findings

- HemoCue® measured lower hemoglobin concentrations compared to Coulter Counter and QBC®.
- Limits of agreement exceeded clinically acceptable thresholds, suggesting potential risks for transfusion decisions.
- Negative correlations were observed between hemoglobin differences and mean concentrations for both comparisons.

## Abstract

Anaemia is a deleterious consequence of malaria, and its accurate diagnosis is crucial for effective management. However, laboratory methods for measuring haemoglobin (Hb) concentration, like the Coulter Counter and the Quantitative Buffy Coat® (QBC®), are costly and not widely accessible in resource-limited settings. The point-of-care HemoCue® test is a cheaper alternative and suitable in rural areas. The study aimed to determine the level of agreement between Coulter Counter/QBC® vs. HemoCue®-measured Hb concentrations by Bland–Altman analysis.

As part of a randomized, placebo-controlled trial of single low-dose primaquine in Ugandan and Congolese children with acute uncomplicated Plasmodium falciparum malaria, Hb concentrations were measured on days 0, 3, 7, and 28 using Coulter Counter (Uganda, n = 1880 paired values), QBC® (DR Congo, n = 1984 paired values) and HemoCue® Hb-301™. The predefined clinically acceptable limits were set at ± 0.5 g/dL.

The Bland–Altman analysis showed that the HemoCue® minus Coulter Counter mean Hb difference was − 0.15 g/dL with lower and upper limits of agreement of − 3.68 g/dL and 3.39 g/dL, respectively. Corresponding HemoCue® minus QBC® values were − 0.23 g/dL, − 1.66 g/dL and 1.22 g/dL. Linear regression of Hb concentration differences vs. mean Hb concentrations showed negative correlations: r = − 0.43 and r = − 0.34 for HemoCue® vs. Coulter Counter and HemoCue® vs. QBC®, respectively.

Compared to Coulter and QBC®, mean HemoCue® measured Hb concentrations were lower and, compared to the Coulter or QBC® methods, had an overall tendency to measure lower Hb concentrations with increasing Hb concentrations. Upper and lower limits of agreement were wider than the predefined clinically acceptable limits of ± 0.5 g/dL. HemoCue® should be used with caution in settings where decisions about blood transfusions are made.

The online version contains supplementary material available at 10.1186/s12936-025-05318-5.