ABSTRACT
OBJECTIVES: Malnutrition in infants less than six months is increasingly recognized. However, the WHO criteria for identifying malnutrition have not been fully evaluated against the risk of in-patient mortality. The observational study was conducted to evaluate the predictability of in-patient mortality of different anthropometric criteria and combination of criteria in order to understand which diagnostic criteria or combination of criteria most accurately predict in-patient mortality. METHODS: Data from a cohort of infants aged one to six months, admitted to Kalawati Saran Children's Hospital, New Delhi between February and December 2018 was analyzed. The discriminatory ability of different anthropometric indexes [weight-for-age Z score (WAZ), weight-for-length Z score (WLZ) and mid-upper arm circumference (MUAC)] and their combinations to predict in-patient mortality was assessed using Receiver operating characteristic (ROC) curves. RESULTS: A total of 1813 infants aged one to six months were admitted during the 11 mo period, of which 107 (5.9%) died in the hospital. Of all admissions, 39.9%, 26% and 23.4% were severely underweight, severely wasted and severely stunted, respectively. WAZ < -3 was the most sensitive predictor of mortality [sensitivity: 74.8%; specificity: 62.3%; area under the curve (AUC): 0.69, 95% CI: 0.64-0.74]. CONCLUSIONS: WAZ < -3 was the most sensitive predictor out of all individual and combined parameters/indexes in identifying infants less than six months at high risk of mortality which suggests that, it should be used to identify at-risk infants between one to six months on admission to in-patient care. Children identified as falling into this category should be properly evaluated and treated during their in-patient stay.