Background: Severe forms of malnutrition have drastic effects on childhood morbidity and mortality in sub-Saharan countries, notably in Ethiopia. Although few studies have previously estimated treatment outcomes of severe acute malnutrition (SAM) in Ethiopia, their reports were widely varied and remained inconsistent. Thus, this study aimed to pool estimates of treatment outcomes and predictors of mortality among SAM children in Ethiopia.

Methods: Databases including PubMed, CINHAL, Web of Sciences; Cochrane, Psych INFO and Google Scholar were comprehensively reviewed to identify 21 studies (dating from 2000 to 2018) estimating treatment outcomes and predictors of mortality among SAM children, using a priori set criteria. PRISMA guideline was used to systematically review and meta-analyze eligible studies. Details of sample size, magnitude of effect sizes, including Hazard Ratio (HRs) and standard errors were extracted. Random-effects model was used to calculate pooled estimates in Stata/se version-14. Cochran's Q, I², and meta-bias statistics were assessed for heterogeneity and Egger’s test for publication bias.

Result: Twenty-one studies were included in the final analysis, which comprised 8,057 under-five children with SAM in Ethiopia. The pooled estimates of treatment outcomes, in terms of death, recovery, defaulter and transfer out were 10.3% (95%CI: 8.3, 12.3), 70.5% (95%CI: 65.7, 72.2), 13.8% (95%CI: 10.8, 16.9) and 5.1% (95%CI: 3.3, 6.9), respectively. Diarrhea (HR: 1.5, 95% CI: 1.1, 2.2), dehydration (HR: 3.1, 95%CI: 2.3, 4.2) and anemia (HR: 2.2, 95%CI: 1.5, 3.3) were statistically significant predictors of mortality among these children. No publication bias was detected.

Conclusion: Treatment outcomes in under-five children with SAM are lower than the national standard, where mortality is being predicted by comorbidities at admission. Children with SAM need to be adequately treated for diarrhea, dehydration and anemia at the primary point of care to reduce mortality.

Keywords: Severe acute malnutrition, treatment outcomes, meta-analysis, and Ethiopia