Interpretation of SMART survey results for Programming

[Anonymous]

Hi All,

In SMART assessments we find GAM prevalence by WHZ and MUAC. As an example we may find a GAM of 11.6% by WHZ and GAM of 6.5% by MUAC. In such a case where we have these 2 types of prevalence and we do realize that these measures are not necessarily capturing the same children in the population. what would be the recommended prevalence to be used for intervention where the protocol for admission is both for WHZ and MUAC. Is there a way to have a combined analysis of the prevalence for MUAC and WHZ to obtain a single prevalence for purposes of programming.

[SMARTKeymaster]

Hi Hassan,

MUAC is a good predictor of mortality among malnourished children, however, it shouldn’t be applied a solely criterion for acute malnutrition owing to its strong association with age, sex, stunting, and its low sensitivity to detect slim children as evidenced by previous studies irrespective of time and place. In your case WHZ would be a safer bet for programming. Recommendations are on course for prospective studies to come up with such combined prevalence for programming. A study by ACF on 16 cross-sectional surveys from South Sudan, Bangladesh, Chad and the Philippines found out that both indicators could only detect 28.5% of the malnourished children (https://nutritionj.biomedcentral.com/articles/10.1186/s12937-015-0074-4).

[Author]

Posts