The Scientist

The US Army has used virtual reality simulations for combat training since the late 1980s. Trainees are placed inside a module, three or four walls of which are filled with computer-generated screens of a battlefield terrain seen from different perspectives. Participants might be inside a replicated combat vehicle or walking a treadmill through a virtual cave. They wear electronic devices that monitor their movements as they encounter those designated as the enemy, who are operating other modules. Everyone is networked by simulated radio communications to battalion and brigade command posts. The military calls this "sweaty palms realism," because the soldiers react as though they're in a real battle, despite the knowledge that it's virtual. The numbers of such training stations initially were limited, but computer-based simulators now allow widespread access. Why couldn't this technology be adapted to drill health workers on the possibilities of a bioterrorist attack?

It can be, says David Siegrist, a research fellow at the Potomac Institute for Policy Studies, a nonprofit think tank in Arlington, Va., that specializes in national security issues. But such a system could do more than merely train people, he suggests. It might even answer whether America is prepared for bioterrorists. "I don't think we're ready, but I can't tell you exactly why in a rigorous way, whereas if you had a networked system of simulation, then pretty soon you could work it for end-to-end solutions," he says, fluidly employing the jargon of military analysis. "You can identify interdependencies and cascading effects, you can start to prioritize your spending and, at the end of the day, you can maybe come up with the answer to the questions, How much is enough? and How much of what?"

Siegrist, an outspoken advocate for adapting simulation to bioterrorism preparedness,¹ is not sanguine about the status quo. "I see a lot of agencies pursuing their own agendas, a lot of them very worthwhile, but the country needs an integrated system solution. We need to treat it as a problem with the system, put somebody in charge, and have [him/her] fix it. Right now, we have 40 different agencies working on it—each one, I would suggest, talking past one another."

Broad Initiative Mooted

Meanwhile, the pieces of a simulation package are scattered, awaiting assembly. For example, LANL has developed a virtual reality application that teaches emergency workers how to cope with victims of a bioterrorist attack. Through a headset, the trainee sees a scene that is also displayed on a screen for observers. It's a small airport, throughout which a pathogen has been released after a terrorist bombing. The first responder wears sensors that feed her activities into the simulation as she triages, diagnoses, and treats cyber-patients. Each victim has signs and symptoms ranging from visible wounds to inhalation exposure to psychological shock.

The Uniformed Services University of the Health Sciences in Bethesda, Md., is one of several institutions around the country where students work on mannequins that can imitate breathing and hearing, have reactive pupils, can move their arms, and will respond to anesthesia or other drugs. They can be intubated and ventilated; they have a pulse, mechanical lungs, heart sounds, and twitch response; and the unit's operator can talk as the patient, using a microphone and a speaker mounted on the mannequin. Most important, the dummies can die and die again without harm, until the trainee gets it right.

Such individualized instruction would be linked to increasingly broader levels of resolution in an antiterrorism simulation network, Siegrist suggests. For example, a plume model would determine the area in which people would be infected over time by a deliberately released pathogen, based on local weather conditions. Other existing modeling can discern the situation for critical infrastructures, such as when the closest hospital would know it was in trouble, how long it would take to transport and test the isolate before confirmation of a bioterrorism attack could be made, and what might happen in the meantime.

A big challenge is to predict what key people would do during a bioterrorism crisis. For instance, would hospital workers go home to their loved ones? If so, how many absentees, and in which jobs, would it take before the hospital reached system failure? Such information could be gathered by placing individual nurses, physicians, and others in a virtual emergency response module, then iterating their reactions throughout the hospital and onward to other hospitals in the city, taking into account the differing departmental strengths and weaknesses. Another advantage of this exercise is that it would help to clarify who's in charge at each phase of the incident. If the president, the FBI, the military, state and city governments, local police, hospitals, and the Red Cross were all involved, the exact chain of command might not otherwise be clear.

Improvements Needed

Until that futuristic day, science is left in the familiar position of relying on technology far beyond what it ever had, and far behind what might be. Nevertheless, Siegrist sounds the warning: "There are limitations to simulation, but there are no limitations to what you can do wrong when all you have is speculation."