February 23rd, 2012
03:57 PM ET
The consensus among many scientists has been that the strain of bird flu currently circulating – H5N1 – is not only highly infectious, but potentially deadly. That is based on the nearly 600 cases confirmed by the World Health Organization, more than half of which have resulted in death.
But a new study analyzing WHO data suggests that H5N1 may not be as virulent as previously thought, and that mild infections could be slipping under the radar because of less-than-ideal detection methods.
Were those mild cases included, the death rate due to H5N1 could be dramatically lower, according to research appearing in the journal, Science.
According to Wang, strictly-defined parameters must be met to confirm an H5N1 case: a high fever, verifiable exposure to the H5N1 virus within seven days of diagnosis, and confirmation of infection by a WHO-approved laboratory.
"These criteria were designed to be extremely specific for identification of H5N1 disease, but they are not sensitive enough to identify less severe cases," said Wang.
H5N1 is lethal to certain species of birds, and in its present form, rarely infects humans. But WHO data suggest that when H5N1 does jump to humans, the potential for death is high. Of 586 people infected by H5N1 thus far, according to WHO data, 346 died.
Researchers at Mount Sinai School of Medicine analyzed 20 studies, which included 12,677 participants, but in their analysis eliminated the need for patients to show overt symptoms of H5N1 infection, as is usually prescribed by the WHO. Their analysis suggests that the rate of infection due to H5N1 could be between 1-2%.
What does that mean? According to the study, potentially, "millions of people who have been infected worldwide." If millions of people have been infected, that could dampen the impact of a mere 346 deaths.
One challenge is uniform and accurate testing, said experts. For one thing sensitive tests required to identify H5N1 are difficult to perform in rural areas where bird-to-human interaction is high (and where H5N1 is most likely to spread), so the question about the true number of cases of H5N1 is still just that - a question.
And there are issues with the tests themselves. According to experts, current tests for H5N1 reap a lot of false-positives.
Richard Webby, director of the World Health Organization Collaborating Center for Studies on the Ecology of Influenza in Animals and Birds, agrees that better testing is needed.
"There are almost certainly asymptomatic cases of H5N1 as there are with all influenzas," said Webby, associate member of the Infectious Diseases Department at St. Jude Children's Research Hospital, by email. "The issue is that the current [WHO] figures are based upon laboratory confirmed cases which biases this to people who are sick enough to seek medical attention."
Nathan Wolfe, author of The Viral Storm and director of the Global Viral Forecasting Initiative, agrees that the true prevalence of H5N1 is likely higher than the cases confirmed by WHO.
But questions linger about how much lower the mortality rate for H5N1 truly is, and what a lower rate would mean.
"While [study authors] suggest a 'lower' mortality rate, the question is how much lower?" said Wolfe. " The 58.6% figure [according to current WHO data] is massive so if it's really in the 20 or 30% range that's still huge."
Wolfe added that the 1918 flu pandemic was likely less than 2% mortality, so whether the mortality rate is 58.6% or it is closer to 20%, "...either way there is still plenty to fear from H5N1."
And further complicating this question are two unpublished studies, currently being held up by the journals Science and Nature, which may reveal more about H5N1's virulence. The data are being held because of concerns it could be used for terroristic purposes. A committee convened by WHO reported that both studies should be published in full, but did not specify when.
To get a clearer picture of how deadly the H5N1 virus truly is the next step is better testing, according to Wang.
"There is currently no organized way to detect mild or asymptomatic cases of H5N1 infection," said Wang. "The number of mild infections is important to know in order to calculate an accurate fatality rate associated with H5N1 infections."
About this blog