Paradigm Shift Intervention Monitoring
twitter
Live feed of underlying pandemic map data here
Commentary
Georgia H1N1 Surveillance Raises Pandemic Concerns
Recombinomics
Commentary 15:57
March 31, 2010
The above comments by Anne Schuchat from the CDC are from the press conference on the spike in hospitalizations in Georgia. The March 5 investigation confirmed that the hospitalizations were due to lab confirmed H1N1, and there was no seasonal flu found. The comments notes that many hospitalized patients with H1N1 are not tested, and as the CDC has warned previously, in some instances the rapid test, which is still in widely used worldwide, has a sensitivity as low as 10%. Thus, in those instance 90% of the H1N1 infected patients tested negative.
Although the study confirmed H1N1 as the cause for the spike in hospitalizations, the press conference did not address the increase in deaths described in the Georgia March 24 press release, which cited 72 deaths. Although this number is a gross under-estimate, it does not match the 60 deaths detailed in the table at the GA website and does not address the recent P&I data for Atlanta, which is dramatically higher than levels reported for the same time frame in 2009. In the past two weeks there were 28 deaths in Atlanta, compared to 6 reported in 2009.
H1N1 testing has been at historic lows in the US, even though the current pandemic was confirmed in the US almost exactly one year ago, supporting a new wave at this time. In the latest CDC weekly report (week 11), only 3050 samples were tested for the entire country. Although this number may be updated in this week’s report, the surveillance is significantly depressed and under-reporting cases and deaths.
As noted in the press conference, 90% of pH1N1 deaths are under 65, in contrast to seasonal flu, where 90% of deaths are over 65. The infection rate in the over 65 population is low, because of prior exposures to the 1918 virus, seasonal H1N1 circulating prior to 1950, or vaccination with the 1976 swine flu. Therefore, surveillance sites that submit samples from an elderly population have few pH1N1 samples, and testing would largely simply reduce the percentage positive. The GA data indicates the age group with the most positives is 25-49, and that age group does not typically seek medical treatment for influenza, so they may be significantly under-represented in the flu surveillance network.
Consequently, the surveillance system produces a low number of samples for testing, as well as a low percent positive, creating a perception of little influenza activity, at a time when the start of a third wave is likely.
Media Links
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings