Diarrhoea, lack of appetite predict mortality in children with severe acute malnutrition
A recent meta-analysis has identified six independent predictors of inpatient mortality in children with severe acute malnutrition (SAM), namely diarrhoea, shock, pneumonia, HIV infection, lack of appetite, and weight-for-height z score (WHZ).
“Early recognition of these prognostic factors in the community, with referral to inpatient facilities, and risk stratification at hospital admission may help to reduce inpatient mortality among children with SAM,” the researchers said.
Studies published between January 2000 and January 2020 were searched in Embase, Ovid Medline, the Cochrane Library, and clinicaltrials.gov using a prespecified protocol. Eligible studies enrolled children aged ≤59 months who were hospitalized with SAM and used multivariable analysis to examine baseline factors independently associated with inpatient mortality. The researchers used random-effects meta-analysis, stratified by the stated measure of effect, where >20 percent of studies included the same factor in analyses.
A total of 1,432 studies were identified, of which 28 met the eligibility criteria: 19 studies included all children with SAM and nine included specific subgroups of children with SAM. All 19 main studies were carried out in eight countries across Africa, with a median of 400 children per study. The mean inpatient mortality was 15.7 percent (95 percent confidence interval [CI], 10.4–21.0 percent). HIV prevalence had a range of 2.1–51 percent. [Am J Clin Nutr 2020;112:1069-1079]
The meta-analysis, stratified by hazard ratio (HR) and odds ratio (OR), revealed nine independent predictors of inpatient mortality: HIV infection (HR, 4.32, 95 percent CI, 2.31–8.08), WHZ (OR, 0.44, 95 percent CI, 0.24–0.80), diarrhoea (HR, 2.84, 95 percent CI, 1.40–5.75), pneumonia (HR, 2.16, 95 percent CI, 1.19–3.02), presence of shock (HR, 3.67, 95 percent CI, 2.24–6.03), and lack of appetite (HR, 2.16, 95 percent CI, 1.48–3.16). Child age and sex were not predictive of inpatient mortality.
The association between oedema and mortality was difficult to establish from the available studies.
“The independent factors we identified reflect the interaction of predisposing factors, comorbidities, and end-organ manifestations of SAM, which ultimately drive mortality in these critically ill children,” the researchers said. “We require a greater mechanistic understanding of the underlying causal pathways to ascertain whether early interventions targeting these factors may reduce mortality in this high-risk population.”
Moreover, the following factors also correlated with inpatient mortality in more than one study but were ineligible for meta-analysis: raised inflammatory markers, reduced consciousness, urban setting, absence of fever, presence of skin changes, and bacteraemia.
Although a widely measured variable, fever appeared to be a bad predictor since children with SAM do not usually show the clinical signs of infection, such as fever, partly due to impairment of the acute inflammatory response. [J Trop Med Hyg 1988;91:173-180]
“Malnutrition underlies 45 percent of under-5 deaths globally. SAM is the most serious form of undernutrition, characterized by wasting with or without oedema. Mortality remains high (10–40 percent) among children requiring hospitalization for complicated SAM,” the researchers said. [PLoS Med 2006;3:e500]