Mystery Spike In US Pneumonia and Influenza Deaths
Recombinomics Commentary 23:19
January 31, 2010
Deaths from pneumonia and influenza across the country rose sharply in the weekly flu report released Friday by the Centers for Disease Control and Prevention, but a spokesman for the agency called it merely “a blip we’re checking out.”
Lyn Finelli, the disease centers’ chief flu epidemiologist, said most of the deaths were from pneumonia and were among the elderly. It was possible, she said, that some cities had delayed reporting deaths over the holidays, which would make later data look artificially high.
The above comments, from Donald McNeil’s NY Times piece really do not address the dramatic spike in P&I deaths reported in this week’s CDC flu report. The composite underlying data is available in MMWR weekly reportable diseases section, which indicates that cities with incomplete data are list as unavailable across the board (P&I deaths as well as all other deaths). Thus, missing data from a small number of cities would have a minimal impact on regional or national rates.
The data in the three most recent reports do show that the region (West North-Central) with the highest rate (12.2%) is markedly higher than each of the prior two weeks (8.3%). Other regions with double digit rates in week 3 are the Pacific, with steady increases from 9.1% to 10.6% to 11.0%, as well as New England, which was high for all three weeks (11.1%, 9.2%, 10,7%). South Atlantic, where reported flu levels are highest, had increases from 4.9% in week 1 to 7.0 and 6.8% in the two most recent weeks. Cities with high rates in week 3 include Bridgeport, CT (25.0%), Peoria, IL (21.1%), St Paul, MN (19.0%), and Glendale, CA (17.3%).
Thus, the high rates are widespread and higher than the peak levels in the fall, as well as peak levels for all season since 2005/2006, except 2007/2008, when the peak week broke 9%. However, even in that year, the week 3 rate was 7.1%. Thus, it is not clear why week 3 in 2010 would be more impacted by holiday anomalies than any other recent season, all of which had rates well below the 8.1% in 2010.
Moreover, the highest rates in the prior years were linked to peak levels of seasonal flu, and this season there is no seasonal flu. Over 99% of flu this season has been pandemic H1N1, and although reported levels have been declining, the P&I deaths are spiking to strikingly high levels.
The above CDC explanations do not address the size of the peak and the relationship to reported H1N1 levels, and raise concerns of serious undercounts,. Many hospitals continue to use the rapid tests, which in some instances only detect 10% of pandemic H1N1 infected individuals. The virus quickly moves to the lower respiratory tract in severe cases, and many media reports describe fatal cases that were first confirmed at autopsy.
Thus, the current testing methodologies may significantly under-represent pandemic H1N1 cases, and the true level may correspond to the current strikingly high spike in P&I deaths. The current MMWR tables break down total deaths by age, but the P&I deaths are listed as a composite number. An age breakdown of the P&I deaths in the 122 cities would be useful.
The above CDC explanations, as well as their “thought to” qualifier on the explanation in the week 3 report, raise serious questions that have not been credibly addressed.