November 8, 2002


New Bioterror Models Limit Threat From Smallpox Attack

When anthrax letters brought bioterror to America a year ago, we had just begun to learn the warning signs of lethal mail when experts raised a far more chilling specter: a smallpox attack.

Coincidentally, homeland-security experts had just run a smallpox simulation in June 2001, with results that seemed downright apocalyptic. Clouds of virus released at three shopping malls one December day immediately infected 1,000 people each, the simulation showed; by February some three million had been infected and one million were dead.

"Dark Winter" inspired hundreds of news stories and even docudramas with a single message: If bioterrorists loosed the smallpox virus (variola) through an airborne release or by dispatching "suicide patients" to Times Square, we'd be doomed. It also reportedly spurred Vice President Dick Cheney's push for universal vaccination in advance of a smallpox outbreak.

But as the White House weighs that and other options, some experts are trying to break through the hysteria with a starkly different message: Smallpox spreads slowly and is not very contagious. "It is imperative that we step back from the sensationalistic press and marketing hype emerging from the burgeoning biodefense industry and ask if such a vaccination scheme is a good idea," writes biodefense expert Peter Merkle in the journal Science.

About 10 U.S. teams are developing mathematical models of a smallpox attack, says biologist Ellis McKenzie of the National Institutes of Health's Fogarty International Center, who recently hosted a conclave of modelers. As they argue over the effect of closing highways or shutting schools, of limited and mass vaccination, what emerges is this: The most realistic models show the least dire consequences.

Consider the disputed "R(0)." Pronounced R-nought, it is the number of secondary cases per primary case in a susceptible population. Dark Winter used 10: Everyone infected in the terrorist attack spread smallpox to 10 others. But epidemiologist James Koopman of the University of Michigan, Ann Arbor, suspects R(0) today would be hardly greater than 1. When he helped eradicate smallpox in India 30 years ago, even people packed onto buses for long trips didn't catch smallpox from infected passengers. An R(0) just over 1 means an outbreak would spread so slowly as to be easily contained.

Even a higher R(0) might not spell doom. Scientists led by Martin Meltzer of the U.S. Centers for Disease Control and Prevention modeled an outbreak that started with 100 infected people, each of whom infected three others, who also infected three others, etc. "Quarantine alone could stop disease transmission," they found, even if authorities daily isolated only half of those with overt symptoms. Adding targeted vaccination helped even more. They calculate that 40 million doses of vaccine, not 280 million, would be enough to contain an outbreak.

Historically, smallpox spread most among people living together, and in hospitals that didn't quickly isolate cases and vaccinate staff. For everyone else, experts tell me you have to spend hours with, and probably within six feet of, an infected person. Sitting in adjacent cubicles might not do it. And people today don't live in eight-member households. In one model, based on travel and contact, some runs have no secondary cases: Carriers didn't meet anyone in a way that spreads disease.

"Modeling person-to-person interactions this way is much more realistic," says Dr. Koopman. It also suggests that smallpox would quickly peter out, contrary to models that assume everyone is in contact with everyone else and that the disease propagates indefinitely. In fact, an unpublished analysis says Dark Winter overestimated the toll by a factor of 100.

Other models using realistic parameters find ring vaccination (inoculating around an outbreak) as effective as mass vaccination. That holds even if officials don't recognize the outbreak until two dozen or so people are sick, say sources familiar with a soon-to-be-published model, as long as vaccination reached 80% of those exposed. "If we're prepared, we shouldn't have any trouble containing smallpox from individuals in the first wave," says Dr. Koopman.

Still, what's "realistic" today might seem naive tomorrow. This week, officials leaked a Central Intelligence Agency finding that Iraq, North Korea, France and Russia have covert stocks of variola. Some modelers believe it only prudent to lay out "the worst case," as Edward Kaplan of Yale University argues. Modeling an attack that initially infects 1,000 people, he finds that mass vaccination results in fewer deaths than isolating the sick and vaccinating their contacts. That partly reflects the difficulty of tracing contacts in a mobile, panicked populace.

If smallpox vaccine were risk-free, we would not be having this debate. But it kills two or three people per million, and injures more. "Smallpox scenarios are now being promulgated as 'ground truth' to lawmakers and the public and are being used to justify the potential vaccination of possibly hundreds of thousands of people," argued Dr. Merkle. If the best science is telling us that we can spare most people that risk and still contain a smallpox attack, we should listen.

You can e-mail me at [email protected]

URL for this article:,,SB1036710032278405788.djm,00.html

Hyperlinks in this Article:
(1) mailto:[email protected]

Updated November 8, 2002

Copyright 2002 Dow Jones & Company, Inc. All Rights Reserved

Printing, distribution, and use of this material is governed by your Subscription agreement and Copyright laws.

For information about subscribing go to

Go back to