Portland, Ore. The scenario makes public-health officials sweat: Avian flu breaks out here in Portland. By day 24, there are 104 dead and 6,414 infected. By day 42, the death toll has risen to 846 or more and the infection level to 33,246. Had the city followed the vaccinate-and-quarantine regimen prescribed by the Models of Infectious Disease Agent Study (Midas), however, just 362 would be dead and 2,564 infected, with 34,559 citizens under quarantine.
The quest to diagnose flu cases early, as the first step in containing an outbreak, has a new ally in an emerging breed of microarrays and labs-on-chip. In the face of dire warnings about bird flu from the World Health Organization, biochip makers are gearing up to provide custom microarrays to screen for avian flu strain H5N1 and its close cousins.
The version built at the University of Colorado, Boulder, under contract with the National Institutes of Health, has just come out of testing at the Centers for Disease Control and Prevention (CDC), where a prototype microarray was exposed to a sample from a patient infected with avian flu. Meanwhile, STMicroelectronics says that its recently launched lab-on-chip, called In-Check, can be used to identify viruses like the flu.
The nightmare scenario outlined for Portland is no fantasy. It comes straight from Midas' epidemic modeler, EpiSims, based on earlier simulations of 1.6 million Portland residents frequenting 180,000 locations where they might infect one another. EpiSims, designed to model smallpox outbreaks, has been retooled to model pandemic influenza of the bird-flu type.
"EpiSims is now set up as a very good tool for modeling the spread of H5N1 the avian or bird virus," said James P. Smith, EpiSims project leader at Los Alamos National Laboratory, which programmed the simulator. "It could easily model the quicker turnaround time of special H5N1 microarrays, or flu chips, [using] just some small additional code to account for how you get your test results back more quickly, so people know right away whether they have been infected."
Another modeling tool comes from IBM Corp. According to James Kaufman, manager of Healthcare Informatics at IBM's Almaden Research Center (San Jose, Calif.), just a few changes in the parameters of its downloadable code (www.alphaworks.ibm.com/tech/stem) could measure the use of flu chips in the fight against avian flu. The Java program, which IBM is offering free for noncommercial use, is called the Spatiotemporal Epidemiological Modeler (Stem). Users can tweak any number of parameters to try out their own ideas on containment.
"Stem now comes preconfigured for modeling the spread of avian flu across not only the U.S. where we have accuracy down to the county level as of the last census but the entire globe," said Kaufman. To model the flu chips' effect, he said, "you would just change the time evolution of a disease, given the response of people to becoming informed right away that they are infectious. The difficult question is predicting with confidence what that response would be. In other words, how many people will stay home if they know they have the flu but do not yet show symptoms?"
It's only a matter of time until we see a global outbreak of pandemic influenza, according to Klaus Stohr, the director of flu surveillance at the World Health Organization, who declared last November that the outlook was grim. "We know the recipe, and all the ingredients are there," he said.