Predictive Factors of Time-to-recovery in Children With Severe Acute Malnutrition Treated in Rural Area in Mali: an Application of Cox Regression

Abstract

Background: Malnutrition is a serious public health issue in Africa. West Africa has the highest prevalence of malnutrition. In Mali, access to care is a challenge in remote areas. Thus, the mortality rate of severe acute malnutrition (SAM) is high. The aim of this study was to access outcomes mainly time-to-recovery in children with uncomplicated SAM treated in traditional health facilities compared to community health workers (CHWs) management and identify its predictive factors. Methods: A multicentre, randomised intervention study which followed up prospectively two groups of children between February 2015 and February 2016 in Kita. The intervention group was exposed to treatment delivered in traditional health facilities or CHWs and the control group was exposed to the treatment delivered in traditional health facilities only. Recovery time was determined using Kaplan Meier curves and compared between strata (intervention group, age group and gender). Cox proportional-hazard analysis was used to identify independent predictive factors. Results: Median recovery time of the entire cohort was 42 days (95 % CI: 39–42). There was no significant difference in median time-to-recovery between control group and intervention group (p-value=0.124). After adjustment, independently significant predictive factors of recovery time were weight (AHR = 1.767, 95% CI: 1.47-2.12) and height (AHR = 0.917, 95% CI: 0.89-0.95). Conclusion: Recovery rate was higher than the minimal international standard. Predictive factors of time to recovery were weight and length/height and treatment group was not found significant factor. Thus, including management of uncomplicated SAM would be a good alternative to improve access to care and reduce mortality due to the disease. Key words: uncomplicated severe acute malnutrition, community health workers, time-to-recovery