Influenza and Hospitalizations in Children

Publisher Website
Google Scholar
Abstract
The quantitation of mortality and morbidity caused by influenzavirus infections has traditionally relied on estimates of the number of infections that occur in excess of the number expected for the season. The sharp seasonality of the annual influenza epidemics allows such estimates. Two reports in the January 27 issue of the Journal, one by Neuzil et al.1 and one by Izurieta et al.,2 assessed the serious morbidity that might be attributed to influenzavirus infections in children. These estimates of morbidity are substantial but conservative. Both groups of investigators limited their estimates because of concern about overlap of respiratory syncytial virus infections. In their Tennessee study, Neuzil et al.1 used rates for the “peri-influenza season” as the main base-line values. The rate for the peri-influenza season was subtracted from the rate for the epidemic period of influenza. This method relies on a false assumption that all of the respiratory pathogens involved in the winter respiratory-disease season are layered proportionally, with influenzavirus added on top.3 The activity of both respiratory syncytial virus and parainfluenza virus peaks in the peri-influenza season. Therefore, subtracting the rate for the peri-influenza season leads to a large underestimate of the rate of influenza-attributable disease. Furthermore, this study combined data for the period from 1973 to 1984, a period when influenza epidemics were less intense and less frequent.4 Limiting the analysis to the epidemics that occurred from 1984–1985 to 1992–1993 would provide information that is more representative of the current situation.