Background: The causative role of respiratory viruses detected in upper airway secretions in childhood pneumonia needs further investigation.
Objective: To measure the association between infection with respiratory RNA viruses and pneumonia in children.
Methods: From March 2006 to July 2007, we conducted a case-control study of 680 pneumonia cases (WHO criteria) and 680 randomly selected, concurrently sampled age-matched controls among children aged 2-35 months in Bhaktapur, Nepal. A nasopharyngeal aspirate from each child was examined for 7 respiratory viruses using reverse transcription polymerase chain reaction. We calculated the matched odds ratios (MORs) for the detection of the individual viruses from a case compared with a control as measures of pathogenicity using conditional logistic regression.
Results: At least 1 virus was recovered in 248 (36.5%) cases and 48 (7.1%) controls. The MOR varied from 2.0 to 13.0; the highest associations were observed for parainfluenza virus type 3 (MOR 13.0; 95% confidence interval [CI] 6.0-28.0), respiratory syncytial virus (MOR 10.7; CI 4.6-24.6), and influenza A (MOR 6.3; CI 1.9-21.4). We observed that the association was lower for children age 2-5 months compared with older children for parainfluenza virus type 3 (P value for interaction 0.002).
Conclusions: All of the 7 respiratory viruses were associated with pneumonia, but their pathogenicity varied. Parainfluenza type 3, RSV, and influenza A were most strongly associated with pneumonia.