Validating the diagnostic performance of MUAC in screening moderate acute malnutrition and developing an optimal cut-off for under five children of different regions in Ethiopia

Abstract

BACKGROUND: Valid and reliable anthropometric indicator is useful for early detection and treatment for under nutrition. Although, mid upper arm circumference (MUAC) is used for screening of children with moderate acute malnutrition in Ethiopia, its performance for the different ethnic groups has not been evaluated. OBJECTIVE: To determine the diagnostic performance of MUAC for determination of moderate wasting among children of different ethnic background and develop optimal cut-off. METHODS: A community based cross-sectional study was conducted among under five children of the three regions namely: Somalia, Amhara and Gambella Regions. The diagnostic performance of MUAC was validated using weight for height Z-score< -2 as a gold standard binary classifier. Test variable is mid upper arm circumference (MUAC< 12.5cm) and weight for height Z-Score (WHZ) is standard variable. ROC analysis performed based on the assumptions of MUAC value lower the cut-off point indicates the undernutrition. Area under the curve and validity measures (sensitivity and specificity) was generated as parameter estimated. The results were presented using tables and ROC curves. RESULTS: Except in the Gambella region, there was fair agreement between MUC<12.5cm and Weight for Height Z score<-2 in diagnosing wasting in Somali (Sensitivity = 29.3%, Kappa = 0.325, P<0.001) and in Gambella regions (Sensitivity = 16.7%, Kappa = 0.19, P<0.001). In Amhara region there was fair agreement between the two measures in diagnosing moderate acute malnutrition (MAM) (Sensitivity = 16%, Kappa = 0.216). For the Overall sample, the sensitivity of MUAC<12.5cm was 20.6% (Kappa = 0.245, P<0.001. Based on ROC analysis, the optimal cutoff value of MUAC for diagnosing moderate acute malnutrition for the two regions namely for Gambella and Amhara was 13.85cm with sensitivity of 0.99 and 1.00, respectively. However, for Somali Region the optimal cut was 13.75cm (Sensitivity = 0.98 cm and specificity = 0.71). CONCLUSION: Findings revealed that the inter reliability of measurement for MUAC< 12.5cm and WHZ<-2 for diagnosing MAM was low among different ethnic groups with the cut-off varying in each region. The existing cutoff point is less sensitive for diagnosis of MAM. As Ethiopia is home of diverse ethnic groups with different body frame and environmental conditions, the new cut off points developed for each region recommended to be used for screening moderate acute malnutrition to prevent relapse of MAM and reduce chronic malnutrition.