Child malnutrition: eligibility for referral is not enough to demonstrate a benefit

A screening threshold, not an evaluation of treatment

Mid-upper arm circumference is used to screen for severe acute malnutrition in community settings. In this study, conducted in the Nouna district, a measurement below 11.5 cm triggered a recommendation for referral to a nutrition programme.

The researchers used a prospective regression discontinuity analysis: they compared outcomes among children just above and just below this threshold. The analysis was embedded in the cluster-randomised CHAT trial, but it was not a randomised comparison of nutritional treatment versus no treatment.

Results with no clearly detectable difference

Among 65,554 children screened during twice-yearly censuses, 226 had crossed the referral threshold at least once. The analyses near the threshold included 2,901 observations for mid-upper arm circumference and 2,628 for mortality.

At the next census, the estimated mean difference in mid-upper arm circumference at the threshold was 0.21 cm, with a 95% confidence interval ranging from −0.22 to 0.65 cm. This interval does not allow a clear distinction between a benefit and no effect.

For deaths occurring before the next census, the odds ratio was 0.95, with a confidence interval ranging from 0.03 to 31.37. This substantial uncertainty prevents any conclusion that mortality was reduced, but also prevents ruling out a large effect.

The key limitation: use of care was not measured

The study examines eligibility for referral, not receipt of treatment. Without measuring actual use of the programme, it cannot establish whether referred children received care. It therefore does not demonstrate that malnutrition treatments are ineffective.

Most sensitivity analyses produced similar results. However, the analysis using the widest fixed comparison window produced inconsistent and unstable estimates. The conclusions apply to children near the 11.5 cm threshold, not necessarily to those with more severe malnutrition.

What we can take away

The findings highlight the need to distinguish screening, referral and actual care. The authors suggest strengthening the link to care, but this study does not demonstrate which approach would improve access to treatment.