Optimal screening of children with acute malnutrition requires a change in current WHO guidelines as MUAC and WHZ identify different patient groups
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Forlagets udgivne version, 868 KB, PDF-dokument
BACKGROUND: Timely treatment of acute malnutrition in children <5 years of age could prevent >500,000 deaths annually. Screening at community level is essential to identify children with malnutrition. Current WHO guidelines for community screening for malnutrition recommend a Mid Upper Arm Circumference (MUAC) of <115 mm to identify severe acute malnutrition (SAM). However, it is currently unclear how MUAC relates to the other indicator used to define acute malnutrition: weight-for-height Z-score (WHZ).
METHODS: Secondary data from >11,000 Cambodian children, obtained by different surveys between 2010 and 2012, was used to calculate sensitivity and ROC curves for MUAC and WHZ.
FINDINGS: The secondary analysis showed that using the current WHO cut-off of 115 mm for screening for severe acute malnutrition over 90% of children with a weight-for-height z-score (WHZ) <-3 would have been missed. Reversely, WHZ<-3 missed 80% of the children with a MUAC<115 mm.
CONCLUSIONS: The current WHO cut-off for screening for SAM should be changed upwards from the current 115 mm. In the Cambodian data-set, a cut-off of 133 mm would allow inclusion of >65% of children with a WHZ<-3. Importantly, MUAC and WHZ identified different sub-groups of children with acute malnutrition, therefore these 2 indicators should be regarded as independent from each other. We suggest a 2-step model with MUAC used a screening at community level, followed by MUAC and WHZ measured at a primary health care unit, with both indicators used independently to diagnose severe acute malnutrition. Current guidelines should be changed to reflect this, with treatment initiated when either MUAC <115 mm or WHZ<-3.
|Tidsskrift||P L o S One|
|Status||Udgivet - 2014|
CURIS 2014 NEXS 416
Antal downloads er baseret på statistik fra Google Scholar og www.ku.dk