Common Source in H5N1 Tangerang Cluster Unlikely
Recombinomics Commentary 11:22
February 13, 2008
Lily Sulistyowati, the ministry's spokeswoman, said the teenager started to become sick a week after visiting her grandmother, who sells poultry.
"Some of the chickens and water fowl at her grandmother's house died suddenly, but her grandmother is fine," Sulistyowati told Reuters.
"We can say it is a cluster in a family but it is not a human-to-human case because they may have contracted the virus from the same source," said Lily Sulistyowati, the ministry's spokeswoman, adding that the mother's condition had since improved.
She said the teenager started to become sick a week after visiting her grandmother, who sells poultry. The girl's mother had also visited the grandmother.
"Some of the chickens and water fowl at her grandmother's house died suddenly, but her grandmother is fine," Sulistyowati told Reuters.
The above comments are from two reports on the familial cluster in Indonesia. The first report neglects to mention that the mother of the current case had been H5N1 confirmed and was hospitalized last month.
The more recent report acknowledges the status of the mother, but then tries to use a common source explanation for the two cases in the same family.
However, neither report gives the disease onset dates, which makes the common source explanation unlikely.
The Who update on the mother (index case) indicates she developed symptoms on January 23, before the daughter’s visit to her grandmother, which was a week before symptoms.
The daughter developed symptoms on February 2, 10 days after her mother developed symptoms. The 10 day spread in disease onset dates is typical for human to human transmission. A common source would produce symptoms at the same time (disease onset dates within 1-2 days of each other).
Five to ten day gaps in disease onset dates between relatives are common, signaling human to human transmission. In the past WHO explanations of clusters used extremely tortured logic to try to link clusters to a common source.
About two years ago, clusters became larger and more frequent in northern Vietnam. In addition, the cases tended to be milder and patients recovered. At that time it was fashionable to try to link clusters to duck blood pudding, even though epidemiological data excluded the dish. Only a subset fo those who ate the dish became infected, as did contacts who did not eat the dish. Moreover the disease onset dates made the dish an unlikely source of infection for any of the family members. Moreover, the 17 day gap between disease oinset dates was actually then incorported into WHO definitions of incubation times, which had no scientific basis.
More recently, the same attempt was made for a cluster in China, which was yet another example of human to human transmission based on disease onset dates.
Disease onset dates are the most reliable indicator, because the H5N1 in the family members are likely to be similar regardless of whether the infections were due to a common source of human to human transmission.
Thus, the current cluster in Indonesia is almost certainly human to human and the likelihood that the two family members were infected by independent infections ten days apart at the grandmother’s farm remains close to zero.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings