Estimates of Deaths Associated with Seasonal Influenza --- United States, 1976--2007

Influenza infections are associated with thousands of deaths every year in the United States, with the majority of deaths from seasonal influenza occurring among adults aged ≥65 years (1--4). For several decades, CDC has made annual estimates of influenza-associated deaths, which have been used in influenza research and to develop influenza control and prevention policy. To update previously published estimates of the numbers and rates of influenza-associated deaths during 1976--2003 by adding four influenza seasons through 2006--07, CDC used statistical models with data from death certificate reports. National mortality data for two categories of underlying cause of death codes, pneumonia and influenza causes and respiratory and circulatory causes, were used in regression models to estimate lower and upper bounds for the number of influenza-associated deaths. Estimates by seasonal influenza virus type and subtype were examined to determine any association between virus type and subtype and the number of deaths in a season. This report summarizes the results of these analyses, which found that, during 1976--2007, estimates of annual influenza-associated deaths from respiratory and circulatory causes (including pneumonia and influenza causes) ranged from 3,349 in 1986--87 to 48,614 in 2003--04. The annual rate of influenza-associated death in the United States overall during this period ranged from 1.4 to 16.7 deaths per 100,000 persons. The findings also indicated the wide variation in the estimated number of deaths from season to season was closely related to the particular influenza virus types and subtypes in circulation.

The current study extends estimates of influenza-associated deaths from two previous CDC studies (2,3) by adding data from four more influenza seasons for a total of 31 influenza seasons (1976--2007). Estimates are provided for three age groups (<19 years, 19--64 years, and ≥65 years) and for two categories of underlying cause of death codes: 1) pneumonia and influenza causes and 2) respiratory and circulatory causes. Deaths from pneumonia and influenza causes are highly correlated with the circulation of influenza (1) and can be considered a lower bound for deaths associated with influenza (2,4). However, a diagnosis of influenza virus infection often is not confirmed with sensitive and specific laboratory diagnostics, particularly among older persons, and even when identified is rarely recorded on death certificates (5). Many deaths associated with influenza infections occur from secondary infections such as bacterial pneumonia or complications of chronic conditions such as congestive heart failure and chronic obstructive pulmonary disease (6). Therefore, estimates using underlying respiratory and circulatory mortality data (which include pneumonia and influenza causes) can provide an upper bound for influenza-associated deaths (2,7).

Using methods published previously (2,3), CDC estimated the numbers and rates of influenza-associated deaths by virus type and subtype by using Poisson regression models that incorporated weekly national respiratory viral surveillance data. Weekly influenza test results by virus type and subtype were provided by approximately 80 World Health Organization (WHO) and 70 National Respiratory and Enteric Virus Surveillance System (NREVSS) collaborating laboratories in the United States (8). Prominent influenza type and subtype were defined as at least 20% of all isolates that were tested in that season. Mortality data were obtained from the National Vital Statistics System and reflect the underlying cause of death recorded on death certificates (9). Deaths were categorized using the International Classification of Diseases eighth revision (ICD-8), ninth revision (ICD-9), or 10th revision (ICD-10), as appropriate. Weekly estimates of the U.S. population by age group were used as part of the model to correspond to the weekly viral surveillance estimates. All data for deaths with underlying pneumonia and influenza causes and respiratory and circulatory causes were actual counts based on the death certificate ICD codes. To estimate the proportion of deaths that were influenza associated, the average annual number of deaths estimated by the model was divided by the average annual counts of death with underlying pneumonia and influenza causes and respiratory and circulatory causes.

For deaths with underlying pneumonia and influenza causes during 1976--2007 in the United States, the models estimated an annual overall average of 6,309 (range: 961 in 1986--87 to 14,715 in 2003--04) influenza-associated deaths (Table 1). For these underlying causes, the average annual rate of influenza-associated death was 2.4 deaths per 100,000 (range: 0.4--5.1).

Among persons aged <19 years, an estimated annual average of 97 (range: 41 in 1981--82 to 234 in 1977--78) influenza-associated deaths with underlying pneumonia and influenza causes occurred (Table 1). The average annual rate of influenza-associated deaths for this age group was 0.1 deaths per 100,000 persons (range: 0.1--0.3). Among adults aged 19--64 years, an estimated annual average of 666 (range: 173 in 1981--82 to 1,459 in 2004--05) influenza-associated deaths with underlying pneumonia and influenza causes occurred. The average annual rate of influenza-associated deaths for this age group was 0.4 deaths per 100,000 persons (range: 0.1--0.8). Among adults aged ≥65 years, an estimated annual average of 5,546 (range: 673 in 1978--79 to 13,245 in 2003--04) influenza-associated deaths with underlying pneumonia and influenza causes occurred. The average annual rate of influenza-associated deaths for this age group was 17.0 deaths per 100,000 (range: 2.4--36.7). Deaths among persons aged ≥65 years accounted for 87.9% of the overall estimated average annual influenza-associated deaths with underlying pneumonia and influenza causes.

For deaths with underlying respiratory and circulatory causes (including pneumonia and influenza causes) during 1976--2007, the models estimated an annual U.S. average overall of 23,607 (range: 3,349 in 1986--87 to 48,614 in 2003--04) influenza-associated deaths (Table 2). For these underlying causes, the average annual rate of influenza-associated death was 9.0 deaths per 100,000 (range: 1.4--16.7).

Among persons aged <19 years, an estimated annual average of 124 (range: 57 in 1981--82 to 197 in 1977--78) influenza-associated deaths with underlying respiratory and circulatory causes occurred (Table 2). The average annual rate of influenza-associated deaths for this age group was 0.2 deaths per 100,000 persons (range: 0.1--0.3). Among adults aged 19--64 years, an estimated annual average of 2,385 (range: 504 in 1981--82 to 4,752 in 2003--04) influenza-associated deaths with underlying respiratory and circulatory causes occurred. The average annual rate of influenza-associated deaths for this age group was 1.5 deaths per 100,000 persons (range: 0.4--3.1). Among adults aged ≥65 years, an estimated annual average of 21,098 (range: 2,344 in 1986--87 to 43,727 in 2003--04) influenza-associated deaths with underlying respiratory and circulatory causes occurred. The average annual rate of influenza-associated deaths for this age group was 66.1 deaths per 100,000 (range: 8.0--121.1). Deaths among persons aged ≥65 years accounted for 89.4% of the overall estimated average annual influenza-associated deaths with underlying respiratory and circulatory causes.

For both causes, the average mortality rates for the 22 seasons during which influenza A(H3N2) was a prominent strain were 2.7 times higher than for the nine seasons that it was not. The average annual number of influenza-associated deaths during influenza A(H3N2) prominent seasons was 7,722 for pneumonia and influenza causes and 28,909 for respiratory and circulatory causes, compared with 2,856 deaths for pneumonia and influenza causes and 10,648 deaths for respiratory and circulatory causes in seasons in which it was not.

The distribution of mortality across age groups was similar for the two groups of coded deaths. For pneumonia and influenza causes, the proportions of average deaths overall were 1.5%, 10.6%, and 87.9% for persons aged <19 years, 19--64 years, and ≥65 years, respectively. For respiratory and circulatory causes, the proportions were 0.5%, 10.1%, and 89.4%.

Based on an average annual count of 74,363 for all pneumonia and influenza deaths, and an average annual estimate of 6,309 deaths associated with influenza in this category, 8.5% of all pneumonia and influenza deaths were influenza associated. Based on an annual average count of 1,132,319 for all respiratory and circulatory deaths and an average annual estimate of 23,607 deaths associated with influenza in this category, 2.1% of all respiratory and circulatory deaths were influenza associated.

Reported by

MG Thompson, PhD, DK Shay, MD, H Zhou, MSc, MPH, CB Bridges, MD, PY Cheng, PhD, E Burns, MA, JS Bresee, MD, NJ Cox, PhD, Influenza Div, National Center for Immunization and Respiratory Diseases, CDC.

Editorial Note

This report updates estimates of the number of influenza-associated deaths from the 1976--77 through 2006--07 influenza seasons and demonstrates the substantial variability in mortality estimates by year, influenza virus type/subtype, and age group. The estimated rates of influenza-associated hospitalizations and deaths vary substantially from one influenza season to the next, depending, in part, on the characteristics of the circulating influenza virus strains (10). Because of this variability, a single estimate cannot be used to summarize influenza-associated deaths. This report provides estimates for two categories of underlying cause of death codes, pneumonia and influenza causes and respiratory and circulatory causes; if only one category is used to summarize the mortality effects of influenza, the respiratory and circulatory data likely provide the most accurate estimates. During the past three decades, the estimated number of annual influenza-associated deaths from respiratory and circulatory causes ranged from a low of 3,349 to a high of 48,614 deaths.

A previous study (2) presented an average annual estimate of 25,420 influenza-associated respiratory and circulatory deaths during a 23-season period; this study estimated an average of 23,607 annual influenza-associated deaths using the same model but over a 31-year period. The findings in this report are similar to those of previous CDC studies (2,3) and other cross-decade studies that used similar models (4,7).

When describing the severity of seasonal influenza epidemics, examining seasons with the same circulating influenza virus type is useful. For example, during seasons with prominent circulation of influenza A(H3N2) viruses, 2.7 times more deaths occurred than during seasons when A(H3N2) viruses were not prominent. An annual estimate of 36,155 influenza-associated respiratory and circulatory deaths often is quoted from an earlier CDC study (2); however, that average was calculated for the period 1990--1999, when more severe influenza A(H3N2) viruses were prominent for eight of the nine seasons.

Variations in influenza-associated mortality by age group also should be noted. As reported in this and other studies (2,3), approximately 90% of influenza-associated deaths occur among adults aged ≥65 years. An estimated annual average of 124 persons aged <19 years and 2,385 aged 19--64 years die from influenza-associated respiratory or circulatory causes. Future research that considers years-of-life-lost is needed to better communicate the mortality burden of influenza in these younger populations. Future research also is needed to estimate and communicate the risk for influenza-associated mortality among different demographic and health risk groups.

The findings in this report are subject to at least four limitations. First, the models do not account for cocirculating pathogens such as respiratory syncytial virus (RSV). Future research should replicate and extend models that distinguish between deaths associated with influenza versus RSV (2). Second, estimates over time might not be comparable because the influenza virus surveillance data used to model mortality rely on national influenza testing practices, which have changed over the past decade (8). Future research should consider how trends in testing practices can be included in these models and cross-season estimates of influenza-associated mortality. Third, increases in the U.S. population aged ≥65 years during the study period could have contributed to a general increase in influenza-associated mortality. Age-adjusting future estimates or estimating deaths in smaller age categories among older adults could address this issue. Finally, because the estimates made in this report rely on national death certificate data and these data currently are available only through 2007, preliminary estimates of 2009 influenza A(H1N1)-associated deaths are not directly comparable with these results.*

Debate will continue regarding the most appropriate statistical models and cause of death categories to use in estimating the number of influenza-associated deaths (1,7). This study's provision of estimates for more narrow (pneumonia and influenza causes) and more broad (respiratory and circulatory causes) categories continues the strategy of comparing and contrasting results from different models as advocated by CDC (1--3) and others (7).

Influenza infections are associated with substantial medical costs, hospitalizations, lost productivity, and thousands of deaths every year in the United States. Annual influenza vaccination is the best way to reduce the risk for complications from influenza infections and is now recommended for all persons aged ≥6 months in the United States. Prompt treatment with influenza antiviral medications can reduce the risk for severe illness and death among persons at increased risk for influenza or who are hospitalized with suspected or confirmed influenza.

References

1. Thompson WW, Moore MR, Weintraub E, et al. Estimating influenza-associated deaths in the United States. Am J Pub Health 2009;99:S225--30.
2. Thompson WW, Shay DK, Weintraub E, et al. Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 2003;289:179--86.
3. Thompson WW, Weintraub E, Dhankhar P, et al. Estimates of US influenza-associated deaths made using four different methods. Influenza Other Respi Viruses 2009;3:37--49.
4. Simonsen L, Reichert TA, Viboud C, Blackwelder WC, Taylor RJ, Miller MA. Impact of influenza vaccination on seasonal mortality in the US elderly population. Arch Intern Med 2005;165:265--72.
5. Wiselka M. Influenza: diagnosis, management, and prophylaxis. BMJ 1994;308:1341--5.
6. Schanzer DL, Langley JM, Tam TW. Co-morbidities associated with influenza-attributed mortality, 1994--2000, Canada. Vaccine 2008;26:4697--703.
7. Newall AT, Viboud C, Wood JG. Influenza-attributable mortality in Australians aged more than 50 years: a comparison of different modelling approaches. Epidemiol Infect 2010;138:836--42.
8. Brammer L, Budd A, Cox N. Seasonal and pandemic influenza surveillance considerations for constructing multicomponent systems. Influenza Other Respi Viruses 2009:3,51--8.
9. CDC. Mortality data, multiple cause detail, 1972--2007. Public use data tapes contents and documentation package. Hyattsville, MD: US Department of Health and Human Services, CDC, National Center for Health Statistics; 2007.
10. Russell CA, Jones TC, Barr IG, et al. Influenza vaccine strain selection and recent studies on the global migration of seasonal influenza viruses. Vaccine 2008;26(Suppl 4):D31--4.

All MMWR HTML versions of articles are electronic conversions from typeset documents. This conversion might result in character translation or format errors in the HTML version. Users are referred to the electronic PDF version (http://www.cdc.gov/mmwr) and/or the original MMWR paper copy for printable versions of official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

**Questions or messages regarding errors in formatting should be addressed to mmwrq@cdc.gov.