Screening for childhood anaemia using copper sulphate densitometry

Abstract

The objective of this study was to evaluate copper sulphate densitometry as a screening method for anaemia in children. The accuracy of copper sulphate densitometry was also compared to clinical assessment for the presence of pallor and haemoglobin measurement with a BMS-haemoglobinometer. Different observers performed these three screening tests independently. For the purposes of this study, anaemia was defined as a laboratory haemoglobin (Hb) concentration below 10 g/dl. A cross-sectional screening study was undertaken, where the results of the different screening tests were compared to laboratory haemoglobin determination (gold standard). The study sample consisted of one hundred consecutive children, aged between 6 months and 6 years, whose parents had given informed written consent for participation. The study was conducted in the Paediatric Outpatient Department of Pretoria Academic Hospital (73 children) and a local creche (27 children). In this study sample, the prevalence of anaemia (Hb < 10 g/dl) was 17% (95% Confidence Interval (95%CI) 10.2; 25.8). Clinical assessment by students for the presence of pallor had a sensitivity of 41.2% (95%CI 19.4; 66.5), specificity of 81.9% (95%CI 71.6; 89.2), positive predictive value of 31.8% (95% CI14.7; 54.9) and negative predictive value of 87.2%(95%CI 77.2; 93.3). The likelihood ratio for detection of anaemia by clinical assessment was 2.3. Copper sulphate densitometry had a sensitivity of 88.2% (95%CI 62.3; 97.9), specificity of 89.2% (95%CI 79.9; 94.6), positive predictive value of 62.5% (95% CI 40.8; 80.5) and negative predictive value of 97.4% (95%CI 90.0; 99.5) to screen for anaemia. The Likelihood Ratio of a positive copper sulphate-screening test was 8.17. On average, haemoglobin concentration was underestimated by 0.29 g/dl with the BMS-haemoglobinometer, with the 95% limits of agreement ranging from underestimation by 1.3 g/dl to over-estimation by 1.9 g/dl. Logistic regression analysis revealed that both the copper sulphate test and measurements with the BMS-haemoglobinometer predicted anaemia accurately. The area under the Receiver Operating Characteristic (ROC) curve for the haemoglobinometer was 0.94 (95%CI 0.87; 1), while the area under the curve for copper sulphate densitometry was 0.89 (95% CI 0.73; 1). Used together, the area under the ROC curve was 0.95 (95% CI 0.89; 1). In resource-poor settings, copper sulphate densitometry could be an accurate, inexpensive and simple screening method for anaemia in children.