The Etiology of Childhood Pneumonia in Bangladesh: Findings From the Pneumonia Etiology Research for Child Health (PERCH) Study

Pneumonia remains the leading infectious cause of death among children <5 years, but its cause in most children is unknown. We estimated etiology for each child in 2 Bangladesh sites that represent rural and urban South Asian settings with moderate child mortality. METHODS: As part of the Pneumonia Etiology Research for Child Health study, we enrolled children 1–59 months of age with World Health Organization–defined severe and very severe pneumonia, plus age-frequency-matched controls, in Matlab and Dhaka, Bangladesh. We applied microbiologic methods to nasopharyngeal/oropharyngeal swabs, blood, induced sputum, gastric and lung aspirates. Etiology was estimated using Bayesian methods that integrated case and control data and accounted for imperfect sensitivity and specificity of the measurements. RESULTS: We enrolled 525 cases and 772 controls over 24 months. Of the cases, 9.1% had very severe pneumonia and 42.0% (N = 219) had infiltrates on chest radiograph. Three cases (1.5%) had positive blood cultures (2 Salmonella typhi, 1 Escherichia coli and Klebsiella pneumoniae). All 4 lung aspirates were negative. The etiology among chest radiograph–positive cases was predominantly viral [77.7%, 95% credible interval (CrI): 65.3–88.6], primarily respiratory syncytial virus (31.2%, 95% CrI: 24.7–39.3). Influenza virus had very low estimated etiology (0.6%, 95% CrI: 0.0–2.3). Mycobacterium tuberculosis (3.6%, 95% CrI: 0.5–11.0), Enterobacteriaceae (3.0%, 95% CrI: 0.5–10.0) and Streptococcus pneumoniae (1.8%, 95% CrI: 0.0–5.9) were the only nonviral pathogens in the top 10 etiologies. CONCLUSIONS: Childhood severe and very severe pneumonia in young children in Bangladesh is predominantly viral, notably respiratory syncytial virus.