Towards identifying malnutrition among infants under 6 months: a mixed-methods study of South-Sudanese refugees in Ethiopia

Abstract

AbstractObjectives:To determine (i) whether distinct groups of infants under 6 months old (U6M) were identifiable as malnourished based on anthropometric measures and if so to determine the probability of admittance to GOAL Ethiopia’s Management of At Risk Mothers and Infants (MAMI) programme based on group membership; (ii) whether there were discrepancies in admission using recognised anthropometric criteria, compared with group membership and (iii) the barriers and potential solutions to identifying malnutrition within U6M.Design:Mixed-methods approaches were used, whereby data collected by GOAL Ethiopia underwent: factor mixture modelling, χ2 analysis and logistic regression analysis. Qualitative analysis was performed through coding of key informant interviews.Setting:Data were collected in two refugee camps in Ethiopia. Key informant interviews were conducted remotely with international MAMI programmers and nutrition experts.Participants:Participants were 3444 South-Sudanese U6M and eleven key informants experienced in MAMI programming.Results:Well-nourished and malnourished groups were identified, with notable discrepancies between group membership and MAMI programme admittance. Despite weight for age z-scores (WAZ) emerging as the most discriminant measure to identify malnutrition, admittance was most strongly associated with mid-upper arm circumference (MUAC). Misconceptions surrounding malnutrition, a dearth of evidence and issues with the current identification protocol emerged as barriers to identifying malnutrition among U6M.Conclusions:Our model suggests that WAZ is the most discriminating anthropometric measure for malnutrition in this population. However, the challenges of using WAZ should be weighed up against the more scalable, but potentially overly sensitive and less accurate use of MUAC among U6M.