Scandalous Monitoring of H5N1 Avian Influenza (Bird Flu)
Recombinomics Commentary
February 18, 2005
>> Dick Thompson, of the World Health Organization, told the BBC the findings were significant.
He said: "It means the range of illnesses we have been looking for when considering a diagnosis of avian flu will now be expanded.
"We will have to change the way we conduct our investigations, the management of hospital patients and even the way we deal with their bodily secretions." <<
Changing the way investigations are conducted is long overdue. The monitoring of bird flu is well into the scandalous area. There are glaring deficiencies in deciding who to test, the testing itself, and the reporting of the results. All of these deficiencies have lead to a gross underestimate of the incidence of avian influenza in southeast Asia, and little information on the spread of the virus throughout the world.
The latest report in the New England Journal of Medicine demonstrates that H5N1 can present clinically in ways that are not being currently monitored. Without this monitoring, the spread of H5N1 is largely unknown. However, even some of the most obvious cases are not tested. Even when they are tested, results are initially reported as inconclusive or negative.
Patients that die at home, arrive at hospitals with unusual symptoms or at hospitals that simply don't consider testing for avian influenza, are difficult to monitor. Most of the data on lack of testing or false negatives comes from the familial clusters involving 2-4 patients. Since the patients all have similar symptoms that are bimodal but still relatively close in time, there is a high level of suspicion. The symptoms are classic for H5N1, one or more in the cluster dies, and eventually one or more tests positive for H5N1.
Examples include the most high profile clusters that attracted media attention. Prior to this week, the most well known cluster was recently described in the New England Journal of Medicine involving the cluster in Thailand. Although the paper concludes that all 3 members in the cluster had H5N1, and the daughter transmitted the virus to her mother and aunt, none of the three tested positive initially. The index case was diagnosed as dengue fever and was never tested for H5N1. The mother was not tested while alive. After she died sequences were extracted from fixed tissue, and H5N1 was confirmed. The aunt initially tested as negative, but was subsequently positive. Since the mother was in Bangkok at the time of her daughter's infection, and didn't develop symptoms until after her daughter's death, the cluster was considered to be an example of human-to-human transmission. However, the case was a disaster from a monitoring point of view. Even though the daughter and aunt were from an area that had bird flu outbreaks, and the mother developed symptoms after visiting her daughter in the hospital before she died, neither were tested for bird flu before they died. Moreover, since the index case was not tested, she is not included in the tally of H5N1 cases.
The same scenario was seen in clusters in Vietnam and Cambodia. In an earlier cluster in Vietnam, the index case was never tested, and both of his sisters were initially inconclusive. In a cluster in Vietnam over the summer, both initial cases (cousins) were not tested. In the more recent cluster in Hanoi, the index case tested negative twice - he was only positive at autopsy. His brother initially tested negative. In the cluster from Cambodia, the index case was not tested.
In each of the above examples most of the cases were either not tested or initially tested negative. None of the cases that were not tested are in the official tallies. However, even in cases that were tested and found to be positive for H5N1, cases are not included because they were not confirmed by a second source. Thus, the current tally for Vietnam, Thailand, and Cambodia is significantly underestimated, even though many of the missing were from H5N1 positive clusters, and the cluster-linked patients all died with avian influenza symptoms.
The poor monitoring and reporting of the most obvious cases raises serious questions on the ability to monitor H5N1 infections with any reasonable degree of accuracy or confidence. Newer cases in Vietnam and Cambodia have been described as more complex. Earlier cases with flu-like symptoms coming to Vietnam for treatment were not being tested. The similar situation existed for patients crossing from Cambodia and Laos into Thailand.
In the 1918 flu pandemic, patients were misdiagnosed as having cholera, dengue fever, and typhoid: three diseases that are common in Southeast Asia. The latest finding of a patient with neurological and/or gastrointestinal involvement significantly raises the number that should be tested. Testing for H5N1 should be done for meningococcemia / meningitis cases in the Philippines and mainland China as well as for the large number of deaths of children in India.
Testing of cases presenting with a broad range of symptoms will be a challenge, but increased screening should help define how far and wide the H5N1 has spread in human and animal populations throughout the world.
This long overdue expanded screening is welcome.
Media link