The epidemiology of pneumonia in a birth cohort of children living on the Thai-Myanmar border

Background: There are an estimated 150 million episodes of childhood pneumonia per year resulting in up to 20 million hospital admissions and 1.9 million deaths, making pneumonia the commonest cause of death in children under-five. In order to tackle this huge global health problem robust regionally specific pneumonia epidemiology data must be known. Methods: We did a longitudinal birth cohort study based in Maela camp for displaced persons on the Thai-Myanmar border. 999 mothers were enrolled during pregnancy and their infants followed from birth until 2 years of age. As well as routine monthly visits the mothers were asked to bring their child to the clinic during an illness. Pneumonia was diagnosed using the WHO criteria. All children with pneumonia had a complete blood count, C-reactive protein (CRP), nasopharyngeal aspirate (NPA) and chest x-ray (CXR) performed. Chest x-rays were interpreted using the WHO standardised method. PCR was performed to detect respiratory syncytial virus (RSV), adenovirus, Influenza A, influenza B and human metapneumovirus. Results: There were 1,085 clinical episodes of pneumonia over 1,495 child-years of observation, giving an incidence rate of 0.73 episodes/child-year. The incidence of CXR primary endpoint pneumonia (PEP) was 0.22 episodes/child year and was more commonly found in children >1year (p<0.001). PEP was also associated with a neutrophil count >7.5x109/L and a CRP ≥40mg/L (OR 2.41; p<0.001). A virus was detected in 61.3% of pneumonia episodes, most commonly RSV. Adjusting for age, RSV was associated with more severe clinical disease (OR 2.18; p=0.007) and the presence of other infiltrate on CXR (p=0.001). Using a multivariate model, having a mother <18 years (p=0.001) and a small volume of living space per person (p=0.03) were found to be risk factors for pneumonia. Distance to next house (p=0.04) was a risk factor for multiple episodes of pneumonia and distance from stove to the infants bed (p=0.03) was a risk factor for PEP. Conclusion: The study demonstrated a high incidence of pneumonia in this population with a particularly large burden of viral-associated infections. Important risk factors for developing pneumonia were both maternal and infant age and also living conditions.