Monday, 6 July 2009

Interesting flu case fatality rate paper

It may be flawed but it is accompanied by lots of facts and estimates.

The highest estimate is based on the assumption that swine flu is
equally virulent as common flu so data for common flu is used,
especially that collected by CDC.

The latest CDC data on common flu uses various models to estimate flu
fatality rates so for a population of 10 million in developed nations,
i.e. fully developed health care systems:

case fatality rates for common flu and therefore swine flu is
estimated to be 120-1800

Interesting estimate is that of New York. It predicted that swine flu
is much lower at 60 - 90.

The conclusion is simple. Either the CDC estimate for common flu is
just too high, or Swine flu deaths in New York is the real fatality
rate for common flu.

It would help if there were comparisons with normal flu deaths in
order to confirm the New York telephone survey data. Then we can
determine the range of errors in the telephone sample surveys and the
assumptions made.

One thing is certain though. The school closures and treatments with
respirators have helped reduce fatalities.

Let us hope that the swine flu figure is the correct one instead of
the CDC estimates for flu.

Malaysia, with a population of 30 millioin, will have fatalities
ranging from 180 to 270 with actions taken such as closing of schools,
using Tamilflu and respirators. That telephone survey data was also
made in early May, in Spring, which is only slightly worse than
Malaysia's hot summer. Fatalities in Malaysia will be even lower.

You can argue that it is much lower than deaths due to dengue, but the
question is what would happen if we don't close schools? Will we
approach the fatality of Mexico in the early days or Argentina

One thing is certain. Fatalities due to common flu is vastly
exaggerated by CDC estimates. Even when we closely monitor Swine flu
with almost exhaustive tests, we don't detect that much flu causing
deaths, over that of Swine flu.

No comments: