Laboratory epidemiology of respiratory viruses in a large children's hospital: A STROBE-compliant article

Ye, Sheng; Wang, Tianlin

The study showed that among samples tested for respiratory virus, RSV had the highest positive rate (13.0%), followed by influenza B (9.9%), adenovirus (8.9%), and influenza A (8.0%). The positive rates for the 4 respiratory viruses changed in distinct seasonal patterns. Adenovirus was most likely to be detected in summer, followed by spring, winter and autumn. Influenza A was most likely to be detected in winter, followed by spring, autumn, and summer. Influenza B was most likely to be detected in spring, followed by autumn, summer, and winter. RSV was most likely to be detected in winter, followed by spring, autumn, and summer. The multivariable analysis confirmed these results. The seasonal pattern of these common ARIT viruses can help to develop annual vaccination policy for children. Vaccination can effec-tively reduce the incidence and severity of respiratory infection in children older than 6 months. [18] Taylor et al conducted an epidemiological study involving 17 centers in 8 countries, which showed very similar prevalence for influenza (15.8%), adenovirus (9.8%), and RSV (9.7%). While the prevalence of RSV in our study was higher than Taylor's study (13.0% vs 9.7%), the prevalence of influenza was lower than that study (8%-9% vs 15.8%). Consistent with Taylor's study, the rate of influenza virus increased with age, while the RSV declined with age. However, while our study showed increased positivity of adenovirus with age (OR: 1.02; 95% CI: 1.01-1.04; P = .001), Taylor's study showed declining prevalence of adenovirus with age. [7] The difference might be attributable to different methodological designs for the 2 studies. While Taylor's study was a population-based study that all eligible healthy children receiving H1N1 vaccine were enrolled, our study was based on a tertiary care hospital. Furthermore, Taylor's study analyzed their data in patient level, and each case was confirmed to have the diagnosis of virus infection. However, our study was based on swab or sputum samples, some positive samples may not represent a clinical infection and 1 patient can have several samples.