A little more than two pounds of anthrax spores spilled into the air over a city the size of New York could be expected to kill more than 120,000 people unless state and federal officials respond much more aggressively than they currently plan to, according to the first comprehensive computer model of such a terrorist act.
The new analysis concludes that health officials should rethink the current plan to distribute antibiotics only after an attack. Tens of thousands of lives could be saved, the study predicts, if people could stockpile the drugs in their homes.
The report also is critical of the rush to deploy "biosensors" to detect an attack. Sensors are pricey and in most settings would save few lives compared with improvements in drug distribution for those exposed, the analysis found.
The model, crafted by a trio of university researchers, uses medical details from the U.S. anthrax attacks in 2001, data from the former Soviet Union's anthrax program and the latest in "queue theory," which allows researchers to compare outcomes when products (such as pills) are distributed in various ways to people waiting in line.
Speed is crucial following an anthrax attack because, unlike smallpox, the disease progresses within just a few days to a stage that is nearly 100 percent fatal even with intensive care.
Although people with early symptoms can be saved, the hospital beds, ventilators and pulmonary specialists needed to save them are in relatively short supply. So, unlike the case in 2001, when every patient received intensive care, many of these patients would die. By contrast, everyone who takes appropriate antibiotics before symptoms arise can expect to survive.
"The person in charge, whether it's the president or whoever, needs to push the button really quickly," said Lawrence Wein, the Stanford Business School professor who led the research, published in yesterday's online edition of the Proceedings of the National Academy of Sciences. "We can't sit around a day or two waiting for confirmatory reports that this is indeed anthrax. Every day that it takes you to get the antibiotics started, you lose another 10,000 people."
The Centers for Disease Control and Prevention has said it can fly anthrax medicines (doxycycline or ciprofloxacin, brand name Cipro) to any U.S. city within seven hours. But getting it from the local airport to people's mouths can take days.
To see how distribution plans might affect the number of casualties, Wein and colleagues Edward Kaplan of Yale University and David Craft of the Massachusetts Institute of Technology created a computer program accounting for dozens of factors, from wind speed to the amount of air taken into a person's lungs per minute.
The model presumes a release of 2.2 pounds of anthrax bacteria -- a million billion spores -- 325 feet above a city of 10.8 million, with an additional 700,000 people in suburban areas. It predicts 1.5 million infections in an area 120 miles long and 11 miles wide.
After 48 hours -- the time to become aware of the attack, mobilize the CDC's pharmaceutical stockpile, set up drug distribution centers and get the first pills into people's mouths -- 12,700 people would already be sick, with 17 at the incurable stage.
Under current plans to serve 9,000 people at each neighborhood center, it would take four days to get the pills to everyone and 123,700 would die.
Lines would be shorter under a policy that offered pills only to those with symptoms, but that would be a huge mistake, the model shows, resulting in a total of 600,000 deaths. Under a compromise policy in which only those exhibiting symptoms are treated until 7 percent of the population has symptoms, after which everyone gets treated, the final death toll would be about 250,000.
If people had the antibiotics at home and no one had to wait on line, the death count would be about 60,000, the model predicts.
That number remains high in part because history shows that many people won't take the pills for the full 60 days needed to ensure effectiveness. And if compliance drops even 10 percent, the model shows, the death toll would increase 50 percent.
A policy allowing stockpiling has "equity problems," Wein conceded, because the wealthy can better afford to do so. But the poor would be helped too, he said, because the lines would be that much shorter.
But Richard Besser, who oversees anthrax preparedness for the CDC, is not sold. Many people might take the pills unnecessarily, he said, either because of a false alarm or to treat an infection for which less powerful drugs should be used. That could exacerbate the already problematic increase in drug-resistant bacteria.
"We're not averse to revisiting assumptions," Besser said, "but it's very concerning when large amounts of broad-spectrum antibiotics are in people's medicine cabinets."
Thomas Inglesby, deputy director of Johns Hopkins University's Center for Civilian Biodefense Studies, echoed that concern and noted that the study scenario, though not entirely unlikely, is extreme. "Anthrax attacks can happen in lots of ways," he said, "and most of them would be far far smaller than this."
He said the study is a good reminder that the nation still has no program in place by which it can rate community and state preparedness for an anthrax attack.
Glenn F. Webb, a Vanderbilt University medical modeler, countered that the analysis is conservative in that it assumes everything goes smoothly and that panic and disorder do not slow the process.
Wein recommends setting up teams of trained health professionals who could fly into affected areas to boost drug distribution rates. With one provider for every 700 people, he said, the death toll could drop to 1,000.