The Bug Net

ike Kuliasha knows a simple way to keep terrorists off airplanes: nets.

Kuliasha is the new director of homeland security at the Energy Department's Oak Ridge National Laboratory in Tennessee, where engineers have built a machine to detect explosives residue on airline boarding passes. It works like a grocery store scanner. But instead of being used by clerks to scan bar codes on cereal boxes, the new machine would be operated by airline gate agents to scan passengers' passes as they board planes. Within seconds, the scanner could determine whether there is any trace of explosives on the pass, a clear indication that its holder has handled explosive material.

"At that point, I'd drop a net down on them at the gate," Kuliasha says.

The scanner has its flaws. It will detect nitroglycerin, a main ingredient in dynamite that also happens to be used in heart medicines. Kuliasha admits that heart patients aren't the kind of people you want to surprise with a massive takedown by security guards. But no system is perfect, he says. Though Kuliasha has a playful streak, his vision of counterterrorism is part of a serious effort by all levels of government to build a nationwide defense against biological and chemical attacks. This vast bug net would include scanning and sensing technologies and would depend on human instinct as well.

The strands of the net are being woven now. Across the country, engineers, emergency responders and epidemiologists at federal and local agencies are installing sensing and monitoring devices in subways, airports and hospitals to alert government officials to attacks. Some devices sniff out lethal gases that might have been released in a subway station, others search for patterns in disease reports that could signal an attack is under way. In the same way that smoke detectors warn of fire and set in motion evacuation and containment procedures, chemical and biological attack warning devices and systems show authorities where an attack has occurred and what has been released, giving them a better chance to reduce casualties.

Federal agencies, including the Energy Department, the Centers for Disease Control and Prevention and the Defense Department, have been building and perfecting chemical and biological agent detection systems in conjunction with local agencies and manufacturers for several years. They've been successful in isolated areas. But funding for an aggressive, nationwide effort didn't exist, mainly because the threat of chemical and biological terrorism seemed remote.

A year ago this month, all that changed. After the shock of last October's anthrax mailings (which public health officials refer to as "10-4," the day the first anthrax infection was confirmed), a new federal security strategy began to take shape. The anthrax mailings killed five people, including two Postal Service employees. The Bush administration and Congress responded by injecting more than a billion dollars into chemical and biological weapons defense programs. Much of the spending has cascaded down to state and local governments and is fueling a new technology market.

Today, the weavers of the bug net are working overtime.

TIMING IS EVERYTHING

The way Sgt. Leslie Campbell sees it, the difference between success and failure in containing a chemical or biological attack on a subway station is information. Campbell is an officer on the police force that protects Metro, the subway system that serves metropolitan Washington. He heads the transit agency's weapons of mass destruction defense program. For the past four years, Metro has worked with the Energy Department, the National Institute of Justice and the Federal Transit Administration to deploy chemical sensors in some stations. The program is the first of its kind in the world, and it could be a model for national chemical weapons defense.

Knowing exactly where an attack has occurred the moment it happens, and knowing which chemical agent has been used is the best way to save lives, Campbell says. He believes this because he's seen what happens when that information is lacking. Metro officials have studied Tokyo's response to a 1995 attack on its subway by religious cult members. The terrorists released a lethal nerve gas called sarin in subway cars on three different rail lines during morning rush hour. One milligram of the deadly chemical paralyzes the diaphragm and suffocates a person in minutes.

In the view of Washington's Metro officials, the Tokyo government's response to the attack was a mess. Emergency responders and hazardous materials teams didn't enter the contaminated areas of the subway for 45 minutes. By that time, sarin fumes already had affected hundreds. Panicked victims stumbled to local hospitals, where clueless physicians had no idea chemicals had been released. Eleven people died in the attack, and more than 5,000 were injured. The terrorists, for reasons that still stump observers, used a very diluted form of sarin; that probably reduced the death toll. But considering the government's clumsy response, Tokyo was lucky.

Campbell isn't relying on luck. In 1998, Metro began its chemical sensor program, known as PROTECT (Program for Response Options and Technology Enhancements for Chemical/Biological Terrorism). Two factors made Metro an early mover. Washington is an enticing target because of its symbolic value. In addition, Metro trains carry a special cargo. During any given weekday rush hour, there are more federal workers aboard the system's trains than in any place other than a few large federal buildings or military bases. During peak morning and evening travel times, at least 100,000 workers are zooming through the subway tunnels, large numbers of them headed to or from the Pentagon or the dense concentration of federal buildings downtown and on Capitol Hill. Like the city itself, those people are symbolic targets. So how does Metro protect them? Campbell is hesitant to reveal too many details, but he is willing to play a videotape of a drill Metro ran in December.

As a foreboding drumbeat rolls against the opening credits, Campbell sets the scene: During morning rush hour, "a terrorist with a political agenda" enters the Smithsonian Metro station carrying four plastic sports drink bottles filled with a toxic chemical agent. On the tape, a benign-looking actor plays the role of terrorist. He casually bends down next to a trash can, puts a bottle on the ground and kicks it over before moving further into the station to plant more bottles. The drill shows how PROTECT is supposed to work, but it also shows why things went so wrong in Tokyo.

Within minutes, gas from the spilled liquid triggers a sensor. An alarm sounds in Metro's central command center, and the sensor indicates what kind of chemical has been detected. Staffers immediately train surveillance cameras on the spot the sensors tell them was hit. The Tokyo subway didn't have chemical sensors, so the first signs of attack were passengers rubbing their eyes, coughing or collapsing as they stopped breathing Metro command center employees would look for those very same signs among their passengers. If they saw them, they would declare that an attack had occurred and call paramedics and hazardous materials teams.

In the drill, the trains were stopped immediately. When they are running, the trains push and pull immense air currents through subway tunnels, creating an invisible river that would carry a chemical cloud into neighboring stations. That's what happened in Tokyo. Whenever a train moved, it further dispersed the cloud of gas. In Washington, air vents leading to the streets also would be closed, Campbell says. That would prevent hot air rising from the station from wafting the cloud through the vents to the street outside, where the wind would fan it across the city.

With the trains stopped and the vents closed, the video shows hazardous materials teams and the fire department rushing to the scene, equipped with wireless laptop computers hooked into the video images from the surveillance cameras. A simulator on the computers also creates a model of the gas cloud, predicting its movement based on the current wind direction. All this comes less than five minutes into the drill. Paramedics already are treating the sick, and because the hazardous materials teams know precisely what chemical they're dealing with, they know how to decontaminate the area and the victims, and how to protect themselves.

By the time emergency teams responded in Tokyo, gas had been dispersing for an hour. Passengers were wandering out of the station or collapsing on the platform. Those who made it to hospitals contaminated emergency room staff with the sarin on their skin and clothes.

The PROTECT system is designed to stem that kind of mayhem, but it's far from perfect. There have been false alarms, Campbell says-only a few, but too many. Such errors could kill the sensor system. "It's like a car alarm," says Jim Zarzycki, the director of the Edgewood Chemical Biological Center, the Army's main research and development site for chemical and biological sensors at Aberdeen Proving Ground, Md. "How much attention do you pay when your neighbor's car alarm goes off? Pretty soon, the technology is useless."

But false alarms aren't the only worry. Chemical sensors were designed for the open air of the battlefield, not the dingy, dusty and dirt-filled air of a subway station. How can engineers ensure the sensors, which scan the air around them by sucking it up like a vacuum, don't become clogged and break down?

That's where Tony Policastro comes in. A mechanical engineer at the Energy Department's Argonne National Laboratory near Chicago, he leads the team of scientists modifying existing sensor technology to work in environments for which it never was designed. Policastro is even more reticent than Campbell. He won't reveal how many sensors Metro has installed. He won't say what stations they're in. He won't say what they test for. But he will say that when it comes to safeguarding against chemical attacks, "detectors alone can't solve the problem." Responders need surveillance cameras and laptop computers with video feeds and cloud trajectories. Sensors are just one piece of the system.

Sensor quality continues to improve. The first devices the Energy Department tested produced too many false alarms and didn't hold up under dirty subway conditions, Policastro says. But the manufacturer-and he won't say who it is-has remedied those problems. Now the detectors should last 15 years before they're replaced. Metro officials want to deploy PROTECT to a dozen more stations in the next year, he says. Energy's Sandia National Laboratory in New Mexico also is working on a chemical sensor deployment at San Francisco International Airport.

Policastro and Campbell have great faith in the accuracy of their equipment not just because it has been tested, but also because engineers have been developing and improving chemical sensors for decades.

Today, most sensors do their work by sucking a substance into a tube, where it bounces about, or placing it on a microchip or crystal, with which it reacts. The speed at which chemical particles move, the direction they go, and how much they weigh are unique to each substance. The sensors recognize those factors and use them to identify chemicals. Today's sensors can tag all the lethal agents, such as sarin, mustard gas, and VX, one of the most lethal chemicals ever created, within seconds. Chemical detection has evolved into an exacting and rapid science. But the same can't be said for detecting biological agents, which make much deadlier weapons.

IDENTITY CRISIS

Disease-causing bacteria and viruses are more frightening to Policastro and Campbell than chemicals for two reasons: Their effects are more devastating-biological agents are hundreds to thousands times more toxic than chemicals-and they can't be detected as quickly. Biological weapons cause diseases whose symptoms don't appear immediately after ingestion. People react to chemicals within minutes or seconds. But it took days for symptoms of anthrax to appear in last October's victims. Inhalation anthrax victims usually are beyond the help of medical treatment three days after exposure.

Engineers are probably five years away from developing rapid biological sensors like those now in use for chemicals, says Calvin Chue, a research scientist at the Johns Hopkins Center for Civilian Biodefense Strategies in Baltimore. Because living things don't always behave predictably and they don't all look alike, it's much harder to detect biological organisms. Some bacteria and viruses replicate at different speeds and patterns than others. Some excrete toxins, some don't. Some are harmless, while their relatives are deadly. Bacillus anthracis is the only human-killing member of a class of bacteria that includes hundreds of benign cousins, some found in common potting soil and pesticides. The innate variety of nature makes it hard to identify an organism without painstaking work.

Scientists generally identify a biological agent using three methods, Chue explains. First, they examine its unique genetic makeup, its DNA. But that doesn't tell them whether the substance is still alive and capable of reproducing. They also look for distinguishing proteins on the surface of an organism. The Argonne lab and some medical schools are building databases of those proteins to serve as guides. Ultimately, though, the most accurate way to identify a substance is to grow it in a lab. But that can take days. Lab technicians must physically inspect each culture for telltale signs of its identity. After incubating for 24 hours, for instance, an anthrax culture will form soft peaks like beaten egg whites when teased with a prod. Decontaminating the Brentwood Postal Service facility in Washington, through which anthrax tainted letters were routed, involved months of this kind of laborious sampling and testing.

There have been some significant advances in rapid, on-site testing for biological agents. In the early 1990s, the Navy built hand-held test kits, called assays, which can identify anthrax in a specific area within 15 minutes. Those kits were used to screen portions of the Senate Hart Office Building for anthrax last year. But the assays only detect clusters of 7,000 to 10,000 spores, the amount epidemiologists once believed produced enough bacteria to be lethal to humans. The October attacks showed far fewer spores could be lethal, undermining the reliability of the kits. The kits were only designed for initial testing of an area, not as the definitive means of detecting anthrax, Chue says.

A major breakthrough in detection came in late August, when scientists at Rockefeller University in New York, using funds provided by the Defense Advanced Research Projects Agency, developed an enzyme, which, when applied to suspected anthrax bacteria, yields an identification in 10 minutes or less at spore levels as low as 2,000. Dan Nelson, a member of the research team, says the new detection method moves scientists a big step forward in their efforts to create true biological sensors. Chue sees the most promise in a sensor that would contain a living cell, one that mimics the behavior of those found in human blood and nerves. If a biological agent landed on the cell and began attacking it, that reaction could trip an alarm. However, Chue estimates engineers are at least five years away from building such a device.

Since biological sensors can't provide total security, human beings must fill in. Epidemiologists and other physicians in the biodefense net must rely on instinct and a healthy dose of paranoia to keep a biological attack from turning into an outbreak.

THINK ZEBRAS

Michael Fraser has one less thing to worry about these days: smallpox. Fraser, an epidemiologist formerly with CDC's bioterrorism preparedness and response program, is one of the few people in this country to be vaccinated for the disease in the past 30 years, since routine vaccinations were stopped. He worked at the bioterror unit during the 2001 anthrax mailings, monitoring the effects of the bacteria and helping coordinate CDC's response. (The agency confirmed the first case of anthrax.) Fraser knows that in the event of a terrorist-caused smallpox outbreak, he could be sent into an infected population.

Today, Fraser focuses on the most important component of the government's response to bioterrorism: local public health agencies and doctors in the field. While the CDC provides national health policy leadership and maintains an elite force of epidemiologists to investigate outbreaks, local health departments and local doctors are the front lines of biodefense, says Fraser, now a senior adviser for the National Association of County and City Health Officials (NACCHO) in Washington.

Epidemiologists think differently than other doctors, Fraser explains. Clinicians diagnose illnesses by following the adage, "When you hear hoof beats approaching, think horses, not zebras." If a patient's symptoms look like flu, he probably has the flu, not anthrax. But when epidemiologists hear those same hoof beats, they look for zebras. And today, they're trying to convince doctors at the local level to start zebra-watching on a regular basis.

Those watchers, whom doctors prefer to call "alert clinicians," were the key to fingering anthrax last October, says Dr. James Hughes, the director of the National Center for Infectious Diseases at CDC. The Florida doctor treating Bob Smith, the first of five people to die in the attack, found Smith's symptoms suspicious. A reasonable diagnosis could have been that Smith was suffering a respiratory ailment. But his doctor ordered a spinal tap to test for bacterial meningitis, Hughes says. Testers found bacillus bacteria, of which anthrax is one kind, and notified the Florida state health department. The health department consulted with the CDC, which diagnosed Smith with anthrax. Hughes says a 1993 outbreak of hanta virus in New Mexico, and the first signs of West Nile virus in New York in 1999, were spotted the same way, by alert clinicians looking beyond the obvious, ordering more tests and reporting their suspicions.

To help automate that reporting, CDC and local health departments use a technique called syndromic surveillance. Symptom reports from emergency rooms and doctors' offices, test results from laboratories, and pharmaceutical inventories showing which drugs are being prescribed most often are fed electronically into an online repository, giving public health officials a way to monitor geographic areas for suspicious illnesses. Doctors generally don't monitor patterns of illness beyond their practices or hospitals, let alone their cities or states. But with a bird's eye view of illnesses, epidemiologists can use syndromic surveillance to spot a zebra.

And they have. After the Sept. 11 attacks, CDC epidemiologists used their system to spot health problems among emergency workers at the World Trade Center site ranging from anxiety to dust inhalation. And during this summer's forest fires in Colorado, Denver health officials used a surveillance system to spot patterns of upper respiratory illness among firefighters.

Surveillance has more potential to improve public health generally than to counteract terrorism, NACCHO's Fraser says, so governments are willingly investing in the technology. Los Angeles County and New York, Salt Lake City and Boston have built their own systems. CDC used a program called LEADERS (Lightweight Epidemiological Advanced Detection and Emergency Response System) to monitor hospitals during the 2001 World Series in Arizona and after the Sept. 11 attacks in New York.

NO TIME TO LOSE

The nation's biological and chemical defenses won't be built by a single agency. They will be stitched together, piece by piece, by a vast array of hands. The private sector hopes to play a major role. Before last October, companies had little reason to spend money developing biological detection technologies that governments might never want to buy, says Randy Burkholder, an associate vice president with AdvaMed, a trade association of medical device manufacturers in Washington. But since the Sept. 11 attacks and the anthrax mailings, there has been a tremendous ramp-up in research and development by companies hoping to address federal, state and local government interest in biological counterterrorism. Right now, the private sector outspends government 10-to-1 in R&D for medical technology, Burkholder says. To take advantage of the work of sensor and other technology manufacturers, the federal government must make a serious funding commitment, he adds.

President Bush asked for a major spending increase for chemical and biological defense in his fiscal 2003 budget. Funding requested for the Defense Department's Biological Chemical Defense program totals $1.4 billion, a 60 percent increase over the previous year. More than 42 percent of the new funds are slated for R&D. As part of the program, in the fall and winter, the Pentagon will test open-air biological monitoring equipment in Albuquerque, N.M., and the Washington metropolitan area.

The urgency to construct defenses against chemical and biological weapons keeps mounting. New threats constantly emerge. Last summer, media reports revealed that Defense Secretary Donald Rumsfeld had warned Pentagon officials about the threat posed to the United States by the proliferation of cruise missiles that could be obtained by terrorists and equipped with chemical or biological warheads. The swirling debate about whether the United States should invade Iraq has prompted concern among Defense and public health officials about terrorist reprisal attacks with chemical and biological weapons on American soil or on U.S. troops and allies, especially in the Middle East and Central Asia.

Analyses by the National Research Council and the Energy Department agree that assaults using industrial chemicals such as cyanide, which can be more easily obtained than agents like sarin, pose the highest risk to the U.S. population. Sarin and its cousins, such as mustard gas and VX, are next on the list. Biological pathogens are the least likely to be used, but would be most devastating. In the overall scheme of terrorist threats, the risk of chemical and biological attack is thought to be relatively low. For instance, the National Research Council found in a study this year that a terrorist is more likely to attack a nuclear power plant by land or air than to launch an assault with chemical or biological weapons.

Those odds don't calm the weavers of the bug net. The risk may be low, but a biological or chemical attack could cause casualties that would eclipse those of Sept. 11.