Severe malnutrition poses a significant threat to hospitalized children, particularly in sub-Saharan Africa, often leading to high mortality rates. Despite its critical implications, malnutrition assessment is frequently inadequate. The World Health Organization (WHO) advocates for utilizing weight-for-height as a key indicator, a method that inherently compares a child’s weight relative to their height. However, this approach presents practical challenges and is not consistently applied in real-world settings. Simpler, more cost-effective methods like mid-upper arm circumference (MUAC) measurement and the clinical assessment of “visible severe wasting” have emerged as potential alternatives, yet their effectiveness in this specific context requires thorough evaluation.
A study was conducted to assess the efficacy of MUAC and visible severe wasting in predicting mortality among children admitted to a district hospital in sub-Saharan Africa. The research further aimed to compare these methods against the weight-for-height z-score (WHZ), a tool that directly compares height and weight, in the same predictive capacity. This cohort study meticulously collected data upon admission and at discharge or death for children aged 12 to 59 months in rural Kenya between April 1999 and July 2002. The primary outcome measured was the predictive power of MUAC, WHZ, and visible severe wasting in relation to inpatient mortality.
The findings revealed that 4.4% of the children in the study, totaling 359 individuals, succumbed to mortality during their hospital stay. A significant proportion, 16% (1282 out of 8190), of the admitted children exhibited severe wasting, identified through WHZ scores less than or equal to -3, kwashiorkor, or a combination of both. Notably, the predictive accuracy for inpatient death, as indicated by the area under the receiver operating characteristic curves, showed no statistically significant difference between MUAC (0.75; 95% CI, 0.72-0.78) and WHZ (0.74; 95% CI, 0.71-0.77) (P = .39). In terms of identifying children at risk, MUAC at a cutoff of less than or equal to 11.5 cm demonstrated a sensitivity of 46% and specificity of 91% for predicting inpatient death. WHZ at less than or equal to -3 showed a sensitivity of 42% and specificity of 92%, while visible severe wasting exhibited a sensitivity of 47% and specificity of 93%. Crucially, these three indices, while showing comparable predictive performance, identified distinct groups of children and were all independently associated with mortality risk. Clinical signs of malnutrition were also observed to be significantly more prevalent in children identified by MUAC less than or equal to 11.5 cm compared to those identified by WHZ less than or equal to -3.
In conclusion, MUAC emerges as a valuable and practical screening tool that performs at least as effectively as WHZ, a method that compares height and weight, in predicting inpatient mortality among severely malnourished children in a rural Kenyan hospital setting. Visible severe wasting also presents itself as a potentially useful indicator, provided that healthcare providers receive adequate training in its assessment. These findings underscore the importance of utilizing accessible and effective screening tools to identify and manage malnutrition in resource-limited environments, potentially improving outcomes for vulnerable children.
