ABSTRACT
BACKGROUND: In countries with a high prevalence of undernutrition, timely, accurate screening at the community level is essential to identify children with wasting. The World Health Organization recommends using either weight-for-height z-scores (WHZ) and mid-upper arm circumference (MUAC) or both measures and signs of edema to be used to identify children with severe acute malnutrition for treatment. We compared WHZ and MUAC cutoffs to identify wasting among children aged 6-59 months in Nepal, using WHZ as the reference standard. METHODS: We used cross-sectional anthropometric data for 3,169 children aged 6-59 months from a 2017 cross-sectional dataset, representative of 42 of Nepal's 77 districts. We used descriptive statistics, receiver operating characteristic (ROC) curves, and kappa statistics to compare the use of MUAC and WHZ to identify wasting. The Youden index was calculated to determine the optimum MUAC cutoffs. RESULTS: The prevalence of wasting was 3.1% and 10.5% using MUAC and WHZ, respectively. We found 13.6% sensitivity for severe acute malnutrition (SAM) (MUAC <115 mm) and 21.0% sensitivity for moderate acute malnutrition (MAM) (MUAC ≥115 to <125 mm), with specificity of 99.7% and 91.2%, respectively. The sensitivity of MUAC for children aged 6-23 months was higher than for children aged 24-59 months. The total area of the ROC curve was 0.53 for the MUAC cutoff for SAM and 0.56 for MAM. The optimum MUAC cutoffs for SAM and MAM were 125 mm and 132 mm, respectively. CONCLUSIONS: Although MUAC can be used as a rapid screening tool to detect wasting in children aged 6-59 months, using the recommended MUAC cutoffs captures only a small proportion of the total number of wasted children. The poor sensitivity and specificity of MUAC compared to WHZ suggests a need to refine admission and discharge criteria for acute malnutrition management programs to ensure that wasting among infants and children in Nepal is consistently and accurately diagnosed and treated.
