Bird flu tracked person to person in "Hutch" program

Seattle researchers have used a new statistical tool to confirm that avian influenza was spread person to person in the case of a family in Indonesia last year.

International investigators from the World Health Organization have already determined the deadly virus was spread from human to human in that particular family. But these findings, announced today by the Fred Hutchinson Cancer Research Center, represent the first scientific, statistical look at how the virus actually spread among the eight people, seven of whom died.

The senior author of the study, Ira Longini, a biostatistician at "The Hutch" and the University of Washington, said the statistical tool developed for this research, called TransStat, will be made available to public-health authorities for free. It will allow them to quickly discover whether human-to-human transmission of the virus is occurring during new avian-flu outbreaks, so they can move fast to contain its spread.

"This system is designed to do it in real time as the outbreak is expanding," Longini said. "We want to make the point that if it happened once, it can happen again. Let's be better prepared and use this tool next time."

The Indonesia outbreak, in May 2006, started with a 10-year-old boy in Northern Sumatra who likely caught the virus from an aunt who had been exposed to dead chickens and chicken feces. Researchers believe the boy then passed the virus to his father, and the virus was passed to other family members.

At the time, Longini and other researchers worried the virus would spread beyond the family. "It was extremely worrisome," Longini said. "We were alarmed."

Health authorities stepped in with antiviral medication and quarantined more than 50 relatives and close contacts.

But Longini said another such outbreak is inevitable. "It's a constant threat," he said. "How much more intense will it be the next time?"

The new statistical analysis of the Indonesian outbreak, Longini said, settles some uncertainties about what really happened in the family regarding transmission of the virus.

"Public-health people were saying, 'Yeah, it could have been person-to-person transmission,' but that was as far as it went," he said.

"It kind of needed to have the final nail in the coffin, statistically."

Researchers used the same statistical methods to look at another large avian-flu outbreak in Turkey in 2006. That also infected eight people, four of whom died.

Even though there may have been human-to-human transmission in Turkey, Longini said, the computer said it was too iffy to say for sure.

The TransStat software developed from the researchers' statistical models will be available soon online through MIDAS, the Models of Infectious Disease Agent Study, supported by the National Institute of General Medical Sciences.

The model uses data about certain specifics of an outbreak, including the duration of the outbreak, the virus' incubation period, and the length of the infectious period. It also looks at information about people in the community and whether they fell ill or not.

Then, the program analyzes the pattern of illness onset — who got sick and who didn't, where they lived, and how the cases were separated in time.

"It combines all that information to assess statistically whether there is person-to-person transmission," Longini said.

Dr. Jeff Duchin, the director of communicable-disease control for Public Health — Seattle & King County, said sporadic human-to-human transmission of avian flu has occurred in the past. But so far it has not shown signs of spreading easily that way.

"What you want to be able to pick up quickly is when the virus changes so it is able to move from person to person efficiently," he said.

Though Duchin said he has not taken a close look at Longini's model, it appears it could be helpful on the "front lines" of an avian outbreak when both humans and birds are involved. That could help public-health officials move quickly if the virus transforms into one that can easily move from human to human.

"I'd love to learn more about how we can use this tool," he said. "It sounds like it could have very useful potential."

Carol M. Ostrom: 206-464-2249 or costrom@seattletimes.com