Letters
By Science News
Redesigning flu mortality
In “Designer flu” (SN: 6/2/12, p. 20), researcher Michael Osterholmis quoted as saying that even if the actual kill rate of H5N1 is 20 times lower than the current estimate of 59 percent, H5N1 would still have a mortality rate that “far exceeds” that of the 1918 flu. Wikipedia gives a 1918 flu infection rate estimate of 27 percent, with 3 percent of the world’s population dying. Using the 3 percent mortality rate, “20 times lower” would require an assumption that the H5N1 infection rate is 100 percent, so the phrase “far exceeds” would require an infection rate of over 100 percent [for H5N1 to kill a larger percent of the world population], obviously impossible. Not that any of these numbers aren’t terrible and scary, but shouldn’t we avoid exaggerating the risk?
Linda Riley, Shawnee, Kan.
Mortality rates can be confusing. Osterholm was quoting the World Health Organization estimate that 59 percent of those infected with H5N1 die, which is different from the percent of the world population that dies. What’s more, estimates of the number of people killed by the 1918 flu vary widely, from 20 million to 100 million. If the estimates at the low end of that range are correct, then about 1 percent of the world’s population perished in the pandemic. If the H5N1 rate were a twentieth of the 59 percent estimate, or 2.95 percent, then H5N1 infection rates would need to be only about 50 percent to outstrip the total killed by the 1918 virus. Michael Osterholm’s larger point is that even if H5N1 is a far less deadly virus than WHO numbers suggest it is, if it were to become a pandemic it could kill millions more people than the 1918 Spanish flu did. — Tina Hesman Saey
Quick-change eggs
Regarding Erin Wayman’s article “Egg wars” (SN: 6/2/12, p. 12), what is the explanation for egg color changing in two species of birds over a mere 40 years? My understanding is that change, or evolution, occurs as a result of random mutations in DNA over very long periods of time. The environment then determines the fitness of the mutation. So let’s say that you are getting random mutations frequently: What are the odds that the right mutation in egg color will happen within 40 years? Wouldn’t frequent mutations on this scale threaten the eggs with bad mutations?
Audrey Irvine, via e-mail