The Impact of In¯uenza Epidemics on Hospitalizations Lone Simonsen,a Keiji Fukuda,

The traditional method for assessing the severity of in¯uenza seasons is to estimate the associated increase (i.e., excess) in pneumonia and in¯uenza (P&I) mortality. In this study, excess P&I hospitalizations were estimated from National Hospital Discharge Survey Data from 26 in¯uenza seasons (1970±1995). The average seasonal rate of excess P&I hospitalization was 49 (range, 8±102) /100,000 persons, but average rates were twice as high during A(H3N2) in¯uenza seasons as during A(H1N1)/B seasons. Persons aged!65 years had 57 % of all in-¯uenza-related hospitalizations; however, the average seasonal risk for in¯uenza-related P&I hospitalizations was much higher in the elderly than in persons aged!65 years. The 26 pairs of excess P&I hospitalization and mortality rates were linearly correlated. During the A(H3N2) in¯uenza seasons after the 1968 pandemic, excess P&I hospitalizations declined among persons aged!65 years but not among the elderly. This suggests that in¯uenza-related hospitalizations will increase disproportionately among younger persons in future pandemics. In¯uenza epidemics occur seasonally in temperate climates and result in substantial morbidity and mortality. Since William Farr's [1] studies of in¯uenza-related mortality in Londo