The effectiveness and cost-effectiveness of cash plus interventions to prevent acute malnutrition in Somalia: evidence from an adaptive cluster randomised control trial

Clinical Snapshot

Key Findings

• P Value: Not reported in abstract

• Effect Size: 5.9 percentage point reduction in arm 2 (cash plus SBCC) vs 2.0 percentage point reduction in arm 1 (cash alone)

• Primary Outcome: Child acute malnutrition prevalence at 3 months

• Nnt Or Sensitivity: Number needed to treat approximately 26 children to prevent one additional case of acute malnutrition (based on 3.9 percentage point difference between arms 1 and 2)

• Confidence Interval: Arm 2: 9.1% (95% CI: 6.8, 11.8); Arm 1: 13.0% (95% CI: 10.3, 16.1)

Clinical Application

Requires established cash transfer infrastructure and trained personnel for behaviour change communication Limited direct applicability to mainstream Australian healthcare, but relevant for Indigenous communities, refugee populations, and international aid programmes supported by Australian government Vulnerable communities at risk of malnutrition, particularly in humanitarian or resource-constrained settings

Abstract

BACKGROUND: Acute malnutrition affects millions of children aged <5 years, as well as pregnant and lactating women globally, especially in humanitarian settings. Although cash plus interventions - cash or food transfers combined with complementary components - are widely implemented, evidence remains limited on which combinations and durations are most effective at preventing acute malnutrition. METHODS: We conducted a three-arm cluster-randomised trial within the 'Save the Children's Cash Plus for Nutrition' programme in Somalia. Monthly support was provided as cash alone (arm 1), cash plus social and behaviour change communication (arm 2), or cash plus an additional cash top-up (arm 3). Further, we randomised 33 villages across arms, targeting approximately 1500 households. Primary outcomes included the prevalence and incidence of acute malnutrition in children aged 6-59 months and in mothers, assessed at baseline, midline (three months), and endline (six months). We also conducted market monitoring, qualitative data collection, and analysis. RESULTS: Child acute malnutrition prevalence was approximately 15.0% in each arm at baseline. After three months, prevalence declined by 2.0 percentage points in arm 1 to 13.0% (95% confidence interval (CI) = 10.3, 16.1), representing a relative reduction of 13.3%. In arm 2, prevalence declined by 5.9 percentage points to 9.1% (95% CI = 6.8, 11.8), a relative reduction of 39.3%. In contrast, prevalence in arm 3 remained essentially unchanged, increasing to 15.1% (95% CI = 12.1, 18.6). By endline, there was little change from midline in all arms. Maternal malnutrition improved most in arm 2, but differences were not statistically significant. All arms showed improvements in dietary diversity and food security, but only arm 2 achieved sustained nutrition gains. Household livelihood conditions appeared to improve overall, though monthly expenditures nearly doubled. Arm 2 was the most effective and cost-effective. CONCLUSIONS: Adding social and behaviour change communication to cash transfers significantly improved child nutrition compared to cash alone, highlighting that integrated approaches can enhance nutritional outcomes and at minimal additional cost in humanitarian settings. REGISTRATION: ClinicalTrials.gov: NCT06642012.

References

1. 1.Akseer, N., Alier, K. K., Grounds, S., Garretson, S., Mohamud, S., Khattak, Q., Tripaldi, M., Loddo, F., Mohamoud, S. A., Mahdi, A. Y., Nur, M. A., Abdiqadir, S. M., Mitchell, E., Kees, A., Mahat, M. B., Gure, M., Jibril, D. I., Gedi, D., P'Rajom, M. O., Omer, M., Farah, A., Osman, M. A., Adan, A. M., Mohamoud, F. M., Mohamoud, A. M., Mohamed, A. A., Abdi, A. A., Abdulkadir, A., & Walton, S. (2026). The effectiveness and cost-effectiveness of cash plus interventions to prevent acute malnutrition in Somalia: evidence from an adaptive cluster randomised control trial. Journal of Global Health, 16, 04111. https://doi.org/10.7189/jogh.16.04111