Association between diagnostic criteria for severe acute malnutrition and hospital mortality in children aged 6-59 months in the eastern Democratic Republic of Congo: the Lwiro cohort study.

Background: Few studies have assessed the relationship between weight-for-height (WHZ) and mid-upper arm circumference (MUAC) with hospital mortality considering confounders. The particularity of MUAC for age (MUACZ) is less documented. Objective: This study aims to investigate this relationship in a region endemic for severe acute malnutrition (SAM). Methods: This is a retrospective cohort based on a database of children admitted from 1987 to 2008 in South Kivu, eastern DRC. Our outcome was hospital mortality. To estimate the strength of the association between mortality and nutritional indices, the relative risk (RR) with its 95% confidence interval (95% CI) was calculated. In addition to univariate analyses, we constructed multivariate models from binomial regression. Results: A total of 9,969 children aged 6 to 59 months were selected with a median age of 23 months. 40.9% had SAM (according to the criteria WHZ < -3 and/or MUAC<115 mm and/or the presence of nutritional edema) including 30.2% with nutritional edema and 35.2% had both SAM and chronic malnutrition. The overall hospital mortality was 8.0% and was higher at the beginning of data collection (17.9% in 1987). In univariate analyses, children with a WHZ < -3 had a risk almost 3 times higher of dying than children without SAM. WHZ was more associated with in-hospital mortality than MUAC or MUACZ. Multivariate models confirmed the univariate results. The risk of death was also increased by the presence of edema. Conclusion: In our study, WHZ was the indicator more associated with hospital death compared with MUAC or MUACZ. As such, we recommend that all criteria shall continue to be used for admission to therapeutic SAM programs. Efforts should be encouraged to find simple tools allowing the community to accurately measure WHZ and MUACZ.