摘要：Most estimates of US deaths associated with influenza circulation have been similar despite the use of different approaches. However, a recently published estimate suggested that previous estimates substantially overestimated deaths associated with influenza, and concluded that substantial numbers of deaths during a future pandemic could be prevented because of improvements in medical care. We reviewed the data sources and methods used to estimate influenza-associated deaths. We suggest that discrepancies between the recent estimate and previous estimates of the number of influenza-associated deaths are attributable primarily to the use of different outcomes and methods. We also believe that secondary bacterial infections will likely result in substantial morbidity and mortality during a future influenza pandemic, despite medical progress. ANNUAL OR SEASONAL INFLU enza epidemics are caused by influenza viruses that incorporate slight antigenic differences from recently circulating viruses (i.e., antigenic drift). 1 By contrast, a pandemic is caused by influenza A viruses that contain a new hemagglutinin for which there is little or no preexisting human immunity (i.e., an antigenic shift). 1 Among the four pandemics in the past 120 years, the 1918 pandemic caused the greatest morbidity and mortality. 2 Data from previous influenza pandemics and epidemics provide useful information for preparing for future influenza pandemics. For example, understanding age-specific mortality rates from past pandemics has aided the US government in planning for the distribution of vaccines and antivirals during future pandemics. 3 – 5 Such data, when used to inform policy decisions, are most useful when evaluated carefully in light of their strengths and limitations. The Centers for Disease Control and Prevention (CDC) has estimated deaths associated with influenza epidemics for more than four decades. Estimates from a recent study that used data from the 1990–1991 through 1998–1999 respiratory illness seasons have been cited widely. 6 That study used deaths coded for respiratory and circulatory causes and influenza virus surveillance data to estimate that approximately 36 000 influenza-associated deaths occurred annually in the United States during this period. Approximately 90% of these deaths occurred among people aged 65 years or older. These estimates are consistent with estimates made by several other groups, 7 , 8 but are higher than estimates reported by Doshi and others. 9 , 10
