Effectiveness of community case management of severe pneumonia with oral amoxicillin in children aged 2-59 months in Matiari district, rural Pakistan: A cluster-randomised controlled trial

Background: Pneumonia is a leading global cause of morbidity and mortality in children younger than 5 years. In Pakistan, the proportion of deaths due to pneumonia is higher in rural areas than it is in urban areas, with a substantial proportion of individuals dying at home because referral for care is problematic in such areas. We aimed to establish whether community case identification and management of severe pneumonia by oral antibiotics delivered through community health workers has the potential to reduce the number of infants dying at home. Methods: We did a cluster-randomised controlled trial in Matiari district of rural Sindh, Pakistan. Public-sector lady health workers (LHWs) undertook community case management of WHO-defined severe pneumonia. The children in intervention clusters with suspected pneumonia were screened by LHWs and those diagnosed with severe pneumonia were prescribed oral amoxicillin syrup (90 mg/kg per day in two doses) for 5 days at home. Children in control clusters were given one dose of oral co-trimoxazole and were referred to their nearest health facility for admission and intravenous antibiotics, as per government policy. In both groups, follow-up visits at home were done at days 2, 3, 6, and 14 by LHW. The primary outcome was treatment failure by day 6 after enrolment. We matched and randomly allocated 18 clusters (union councils, the smallest administrative unit of the district) to either intervention and control using a computer-generated randomisation scheme. Analyses were done per-protocol. This trial is registered with ClinicalTrials.gov, number NCT01192789. Findings: 2341 children in intervention clusters and 2069 children in control clusters participated in the study, enrolled between Feb 13, 2008, and March 15, 2010. We recorded 187 (8) treatment failures by day 6 in the intervention group and 273 (13) in the control group. After adjusting for clustering, the risk difference for treatment failure was -5·2 (95 CI -13·7 to 3·3). We recorded three deaths, two by day 6 and one between days 7 and 14. We recorded no serious adverse events. Interpretation: Public sector LHWs in Pakistan were able to satisfactorily diagnose and treat severe pneumonia at home in rural Pakistan. This strategy might effectively reach children with pneumonia in settings where referral is difficult, and it could be a key component of community detection and management strategies for childhood pneumonia. Funding: US Agency for International Development through grants to John Snow Incorporation and Boston University, USA.