BMJ  2006;332:177-178 (21 January), doi:10.1136/bmj.332.7534.177-a

Letter

US flu mortality estimates are based on solid science

Editor—Doshi implies that the use of statistical models to estimate flu related mortality is inappropriate.¹ Not so. Epidemiologists rely on statistical models because the International Classification of Diseases (ICD) code for influenza (ICD9 487) severely undercounts the true number of flu related deaths. Doshi also implies that the Centers for Disease Control and Prevention (CDC) deliberately exaggerates flu mortality for the benefit of the pharmaceutical industry, while other scientists stand by, meek and mute. The BMJ's decision to publish Doshi's commentary with no counterpoint from an expert in the field served only to sow unnecessary confusion.

We need statistical models partly because flu is rarely confirmed in the laboratory and partly because flu brings about many more deaths from secondary bacterial pneumonia or exacerbations of underlying chronic diseases than from primary flu pneumonia. The classic Serfling modelling approach measures the total seasonal number of flu related deaths as the excess mortality above an expected winter baseline of all winter deaths.^{2 3} Regression models guided by virus surveillance data have been developed more recently.^{4 5} These modelling approaches yield similar estimates when applied to the same mortality data for the same range of years (table). The deaths identified by the models only occur when flu causes epidemics² and are most numerous for flu seasons dominated by severe A(H3N2) viruses.³ The important point is that the deaths attributed to flu by these models are flu related in that they would not have occurred without flu.

View this table: [in this window] [in a new window]   Average seasonal estimates of flu related deaths for specific time periods, based on applying various models to all US deaths

 

Why, Doshi asks, are the newer CDC estimates much higher than the older estimates of seasonal flu related deaths? The disparity between older and newer estimates is simply a consequence of various authors studying different time periods over decades when the elderly population simultaneously grew and aged (table); specifically, the proportion of elderly people over 85 had quadrupled by the end of the 1990s, which led to increasing mortality estimates because the risk of dying from flu increases exponentially in the last decades of life.³

And why, Doshi asks, is the estimate of 34 000 deaths for the 1968 pandemic lower than the current CDC estimate for an average season? One cannot compare crude mortality estimates across decades because of the ageing of the population; what makes the mild 1968 pandemic stand out is the shift in mortality to younger people—nearly half the 1968 flu related deaths occurred in people under 65, compared to around 10% of deaths in recent seasons.

National Institutes of Health, Bethesda, MD 20892, USA LSimonsen{at}niaid.nih.gov

National Institutes of Health, Bethesda, MD 20892, USA

---

Competing interests: None declared.

References

1. Doshi P. Are US flu death figures more PR than science? BMJ 2005;331: 1412. (10 December.)[Free Full Text]
2. Simonsen L, Clarke MJ, Williamson GD, Stroup DF, Arden NH, Schonberger LB. The impact of influenza epidemics on mortality: introducing a severity index. Am J Public Health 1997;87: 1944-50.[Abstract/Free Full Text]
3. 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.[Abstract/Free Full Text]
4. Thompson WW, Shay DK, Weintraub E, Brammer L, Cox W, Anderson LJ, et al. Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA 2003;289: 179-86.[Abstract/Free Full Text]
5. Dushoff J, Plotkin JB, Viboud C, Earn DJD, Simonsen L. Mortality due to influenza in the United States—an annualized regression approach using multiple-cause mortality data. Am J Epidemiol 2006;163: 181-7.[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

Related Article

Are US flu death figures more PR than science?

Peter Doshi
BMJ 2005 331: 1412. [Extract] [Full Text] [PDF]

This article has been cited by other articles:

• Rothman, K. J, Evans, S. (2006). More on JAMA's policy on industry sponsored studies.. BMJ 332: 489-489 [Full text]  
• Senn, S. J (2006). JAMA's policy does not go far enough. BMJ 332: 305-305 [Full text]