Revised estimates of deaths associated with seasonal influenza in the US

ECDC comment

The current article describes the number and population rate of influenza related deaths over 31 seasons (1976 through 2007) in the United States (US).

Estimates of Deaths Associated with Seasonal Influenza — United States, 1976–2007CDC’s MMWR, August 27, 2010 / Vol. 59 / No. 33

The current article describes the number and population rate of influenza related deaths over 31 seasons (1976 through 2007) in the United States (US) [1].  Seasonal number of influenza deaths is correlated with the dominant influenza strain in order to identify a possible association. Deaths attributable to influenza are estimated using statistical models with data from death certificate reports [2,3, 4]. Two possible combinations of causes of deaths with different sensitivity and specificity are used to identify influenza related deaths: 1) pneumonia and influenza causes (P&I) and 2) respiratory and circulatory causes (R&C). The former is more specific but fails to capture deaths in individuals who die without being tested for influenza. The second is more sensitive as it captures influenza deaths from secondary infections or exacerbation of chronic conditions but may include deaths that are unrelated with influenza [5]. The resulting estimates are thought to represent the lower and upper bound of the true number of influenza deaths respectively.

The overall average number of death per season was 6,309 (range: 961 to 14,715) for P&I and 23,607 (range: 3,349 to 48,614) for R&C. The average annual rate of influenza-associated death was 2.4 deaths per 100,000 (range: 0.4-5.1) and 9.0 deaths per 100,000 (range: 1.4-16.7) respectively. Both the absolute number of deaths and mortality rates increased with age with a remarkable increase in individuals 65 years and older. In this age group an average annual number of 5,546 P&I and 21,098 R&C deaths were estimated corresponding to 17.0 and 66.1 deaths per 100,000 respectively. Deaths in elderly accounted for almost 90% of the overall estimated average annual influenza-associated deaths with both P&I and R&C underlying causes. Mortality rates were almost three times higher in seasons when influenza A(H3N2) was the predominant virus than in seasons when it was not. When compared to the actual average number of deaths coded as P&I or R&C [6] annually in the US, influenza contributed to around 8.5% and 2% of these respectively.

An accompanying editorial note explains the importance of estimating influenza mortality over a number of seasons. By doing so this study showed relevant variations in influenza mortality by year, influenza virus type/subtype and age group, suggesting that single estimates cannot be used to summarize influenza-associated deaths.

ECDC Comment (24th September 2010): At present an analysis such as this across the European Union cannot be done. Though the influenza and virological data are to be found though the European Influenza Surveillance Network (EISN) the death reporting systems are too diverse.  Progress is being made by the EUROMOMO project but there is a long way to go.[8]  There are a number of estimates based on EU national data, two examples of which are from Norway and Germany.[9,10] But they of course lack the power that comes with very large population.  An estimate of a plausible range for deaths attributable in Europe over the same period can be made for the thirty whole EU/EEA countries if a constant ratio of population is assumed between the USA and EU/EEA of around 506,276,000 to  310,347,000 (the estimated population figures for 2010)[6,7]. If applied to the US range of influenza attributable deaths that would mean increasing each by a ratio of 1.63. For the EU/EEA countries that would give a range from 5400 in the most benign year to 79,200 at the highest with an average of 38,500 influenza attributable deaths. This is a very crude estimate given the differences in age structure, other demographic differences and the underlying conditions between Europe and the USA but it is arguably the best that can be done for Europe at present.   What also needs to be born in mind is that all these estimates may not necessarily hold true with the ‘new’ seasonal influenza emerging in 2010 onwards.[11]                   

  1. CDC, Estimates of Deaths Associated with Seasonal Influenza United States, 1976 to 2007 MMWR Weekly August 27, 2010 / 59(33);1057-1062
  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. Thompson WW, Weintraub E, Dhankhar P, Cheng PY, Brammer L, Meltzer MI, et al. Estimates of US influenza-associated deaths made using four different methods. Influenza and other respiratory viruses. 2009 Jan;3(1):37-49.
  5. Warren-Gash C, Smeeth L, Hayward AC. Influenza as a trigger for acute myocardial infarction or death from cardiovascular disease: a systematic review. The Lancet infectious diseases. 2009 Oct;9(10):601-10.
  6. US Bureau of the census
  7. Eurostat Total Population Database estimates for 2010  
  8. Mazick A, Gergonne B, Wuillaume F, Danis K, Vantarakis A, Uphoff H et al Higher all-cause mortality in children during autumn 2009 compared with the three previous years: pooled results from eight European countries Eurosurveillance Feb 5th 2010
  9. Gran JM, Iversen B, Hungnes O, Aalen OO. Estimating influenza-related excess mortality and reproduction numbers for seasonal influenza in Norway, 1975-2004. Epidemiology and infection.  Mar 25:1-10.
  10. Zucs P, Buchholz U, Haas W, Uphoff H. Influenza associated excess mortality in Germany, 1985-2001. Emerging themes in epidemiology. 2005 Jun 21;2:6.
  11. Nicoll A, Sprenger M. The end of the pandemic – what will be the pattern of influenza in the 2010-11 European winter and beyond . Euro Surveill. 2010;15(32):pii=19637.