Using 3 years of data, CDC places annual flu deaths at 5,000 to 27,000
Using a new method to estimate influenza-related hospitalizations and deaths, the Centers for Disease Control and Prevention (CDC) calculates that somewhere between 5,000 and 27,000 Americans died of flu-related causes during each of the three flu seasons from 2010-11 through 2012-13.
Writing yesterday in PLoS One, the authors noted that flu-related hospitalizations and deaths are undercounted because some people who contract flu don't seek medical care or get tested. During the 2009 H1N1 flu pandemic, they said, the CDC developed a multiplier method to provide more timely estimates of flu hospitalizations and deaths than were possible with older methods.
The authors said they adapted this method for use during non-pandemic seasons, when flu awareness may be lower. Five sites in the CDC's Influenza Hospitalization Surveillance Network (FluSurv-NET) collected data on the frequency and sensitivity of flu testing over two seasons to estimate under-detection of cases.
Using population-based rates of flu-related hospitalizations and intensive care unit (ICU) admissions for 2010-13 from FluSurv-NET, the authors extrapolated them to the US population and corrected them for under-detection. Flu deaths were calculated using a ratio of deaths to hospitalizations.
The team estimated that annual hospitalizations for the three seasons ranged from 114,192 to 624,435, ICU admissions ranged from 18,491 to 95,390, and deaths ranged from 4,915 to 27,174. The burden fell mainly on elderly people, who accounted for 54% to 70% of hospital cases and 71% to 85% of deaths.
"After accounting for the under-detection of influenza, our estimates represented 2.0–5.6 times the level of influenza morbidity as reported by influenza hospitalization surveillance," depending on the age-group, with under-detection greatest in the elderly, the report says.
The authors say their estimates are similar to previous CDC estimates of 86,494 to 544,909 hospitalizations per year from 1979 to 2001 and 3,349 to 48,614 deaths per year from 1976 to 2007, which were based on models of excess flu-related morbidity and mortality.
Mar 4 PLoS One report
Aug 26, 2010, CIDRAP News story on earlier CDC estimates
Guillain-Barre syndrome linked to influenza illness
Being hospitalized for influenza and pneumonia (P&I) was associated with hospitalization for the neurologic disorder Guillain-Barre syndrome (GBS), but influenza vaccine was not, according to an ecological study that used nationally representative data and was reported yesterday in Vaccine.
CDC researchers analyzed monthly hospitalization data for GBS and P&I from 2000 through 2009, as well as data on seasonal flu vaccination coverage from the 2004-05 through 2008-09 seasons.
The authors determined GBS seasonality using Poisson regression and GBS and P&I temporal clusters using scan statistics. They examined the association between P&I and GBS hospitalizations in the same and the following month using negative binomial regression.
The investigators found that vaccine coverage increased from 19.7% to 35.5% in the study period, but GBS did not follow a similar pattern. They also noted a seasonal GBS pattern, with winter months having higher rates.
They found a significant correlation between monthly P&I and GBS hospitalizations that held true for the current month but not the following month. Monthly vaccine coverage, however, was not associated with GBS hospitalization in adjusted models.
Mar 4 Vaccine abstract