Overview of the BioWatch Program
To begin the workshop, three speakers provided an overview of the BioWatch program and background on the nature of the bioterrorism threat. This chapter summarizes these presentations and the ensuing general discussion among the workshop participants. Several key points were highlighted in this session, including
• The biothreat has not diminished; it is evolving, becoming more “democratized.”
• BioWatch, as part of a comprehensive approach, provides an opportunity to mitigate the consequences of a bioattack.
• Coordination and cooperation between federal, state, and local stakeholders is a requirement and a benefit of a successful BioWatch program.
• The aim of an autonomous detection system is to be “better, faster, and cheaper”; reducing time to response could save lives.
BIOTHREAT: A HISTORICAL OVERVIEW
Robert Kadlec, former special assistant to the President for homeland security and senior director for biological defense policy on the White House Homeland Security Council and now a consultant with RPK Consulting, discussed the historical background about bioweapons concerns. At the end of World War II, George W. Merck, who headed the nation’s (now-defunct) offensive bioweapons program at the time, made a number of statements about the threat of bioweapons that still hold true today. Merck said then that it is possible to develop agents for biowarfare
without vast expenditures of money or the construction of huge facilities, that the development of bioweapons could proceed under the guise of legitimate medical research, that such weapons could overwhelm existing defenses, and that expenditures to counter the threat of biowarfare should be viewed in the light of the harm that might come to an unprepared America through a sneak attack. The program Merck headed was stymied, Kadlec explained, because the technology needed to ensure the safety of the researchers or to safely produce biowarfare agents on an industrial scale did not exist then.
Research continued, however, leading into the 1960s, the so-called golden era of U.S. bioweapons research; the program produced several agents and conducted both small- and large-scale environmental tests (e.g., validation tests for nonnuclear alternatives and atmospheric tests). These tests are germane to BioWatch, Kadlec said, because they demonstrated the feasibility of using environmental samplers to recover organisms released in large quantities into the environment. The United States terminated its bioweapons program in 1969 when President Nixon renounced the use of bioweapons and ordered the destruction of more than 71,000 dual-agent–filled munitions.
But although the United States ended its efforts, the Soviet Union continued its offensive biological warfare activities into the 1980s, Kadlec said. The Soviet program has been detailed in two books: The Dead Hand, for which author David Hoffman (2009) won the Pulitzer Prize in 2010, and, more recently, The Soviet Biological Weapons Program: A History, by Milton Leitenberg et al. (2012). During this period, the United States believed bioweapons would be used in low-intensity conflicts, such as the Vietnam War, with the intent of incapacitating people, not killing them. Experts also believed, Kadlec said, that biowarfare was a superpower-only capability.
In the 1990s, at the time of the first Gulf War, intelligence indicated that Saddam Hussein had a biological weapons program. Kadlec, who was a United Nations weapons inspector at the time, said that it was known that Iraq had started producing bioweapons in anticipation of its invasion of Kuwait in March 1990. “Iraq had probably 10,000 or more liters of liquid anthrax that they had distributed across the country in 1,100-liter spray tanks for aircraft,” he said. “How do we know that? We found them.” He showed a video clip from January 1990 of an Iraqi Mirage F1 jet spraying Bacillus globigii, an agent used to simulate biological warfare agents (Center for Research Information, 2004), in a final operational test to demonstrate the capability to release a large amount of
agent against a series of targets the Iraqi military had identified in Israel, Kuwait, and Saudi Arabia.
Based on computer simulations of the projected lay-down pattern, a single airplane fitted with one 1,100-liter tank filled with anthrax would have probably affected a significant number of American and coalition forces in Saudi Arabia. “Remember, there were 10 of these aircraft and the intent was to use all 10,” Kadlec said. “It would have been a very significant event.” The good news, he said, was that the U.S. Air Force claimed air superiority early in the conflict and prevented these planes from flying. The lesson to be learned from this experience was that Iraq was not a superpower, yet it had gained the capability to develop and use bioweapons.
Also during the 1990s, the Japanese religious cult and terrorist organization Aum Shinrikyo was producing biological agents and attempted, unsuccessfully, to conduct attacks with these weapons. What is significant about the near-miss, Kadlec said, is that the intelligence community and Japanese police authorities were totally unaware of this activity, just as the United States had been unaware of the size and significance of the Soviet Union’s program.
Biowarfare took a new direction in 1998 when an al Qaeda’s fatwa made specific mention of the use of strategic biological weapons. The organization created a laboratory dedicated to the development of bioweapons in Kandahar, Afghanistan, and it was in the process of attempting to make weaponized anthrax just prior to the attacks of September 11, 2001. This effort was overseen by Ayman al-Zawahiri, and there is no indication that al Qaeda has changed its strategic intent. Kadlec added that al Qaeda is known to have partnered with the Malaysian terrorist organization Jemaah Islamiyah, which has recruited a biotechnologist named Yazid Sufaat, who was trained at the University of Sacramento in the 1990s. The reality of the situation, Kadlec said, is that al Qaeda is more likely to procure a biological weapon than a nuclear weapon, and he added that other al Qaeda–affiliated groups have stated publicly that they want to use biological weapons.
As a final example to illustrate the threat of bioweapons, Kadlec discussed the anthrax attacks that occurred in 2001, which were not the work of al Qaeda but of one person, “a deranged scientist.” In a sense, he said, this incident represents the final step in a trend stretching from the 1960s, when bioweapons were the province of superpowers, to today, when they have been “democratized.” He noted that the terror organizations continue to promote the use of bioweapons and that the equipment
to produce pounds of anthrax can be bought over the counter for a few thousand dollars (Graham et al., 2008).
Returning to the 2001 letter-borne anthrax attack, Kadlec said that there were several cases of anthrax reported in the days before the first fatality, and he hypothesized that if today’s environmental detection system deployed by the U.S. Postal Service had been in place then, it is likely that anthrax would have been detected before any of those cases had developed. “This doesn’t say that environmental detection is the sine qua non silver bullet,” Kadlec said in concluding his presentation, “but used in conjunction with better point-of-care diagnostics that can enable clinicians to make that diagnosis, I think we have a more comprehensive chance of dealing with and mitigating these events.”
In response to a question about the effective human dose of anthrax, Kadlec said that one study suggested that 8,000 to 10,000 anthrax spores are needed to produce the disease in a human but that there is likely great variability in this number, depending on an individual’s health and physiology (Cieslak and Eitzen, 1999). For other organisms, such as Francisella tularensis, which produces tularemia, an incapacitating dose is approximately 5 to 10 organisms. Intake of perhaps 10 times that many could result in death. Jeffery Runge then commented that it is important not to become fixated on anthrax because it is not the ideal agent for a bioweapon. “The requirement for an automated detection system ought to be able to turn on a dime and be able to measure agents that are deemed to be the highest risk,” he said.
BIOWATCH PROGRAM HISTORY
Jeff Runge, former assistant secretary for health affairs and chief medical officer at the Department of Homeland Security (DHS) and now a principal partner at the Chertoff Group, said that his presentation on the history of BioWatch should serve as a context for considering how to develop the best possible system for protecting the nation, given the increased interest in biodefense compared with the situation when BioWatch was first developed. “History is, in fact, history,” he said. “It’s worthless unless it guides the future.” He also commended DHS for its willingness to take a fresh look at the available technologies and not simply cling to the legacies of the past. Runge said it was his hope that the workshop would help DHS develop a requirements-driven balance in the face of what he characterized as a politically charged environment.
Runge stated that in the run-up to Operation Iraqi Freedom, an intelligence assessment found that there was an increased risk that Iraq would export biological weapons and that, without environmental detection, the United States would not be able to effectively initiate post-exposure prophylaxis to the population. Approximately 30 days after the program was conceptualized, BioWatch deployed its first-generation sensors, which were based on a commercial off-the-shelf product used by the Environmental Protection Agency (EPA) that was modified to detect biological agents using polymerase chain reaction (PCR) technology. These sensors were deployed in the highest-risk urban areas and required the federal government to create a distributed network for installation, maintenance, filter collection, and sample analysis. “As I look back at this,” Runge said, “I think we can acknowledge that the requirement for speed afforded us less precision than optimal.” At the time, he explained, the scientific understanding of how bioweapons might disperse over time in specific locations was not well developed, and a formal biothreat risk assessment had not been performed.
Because of the need for a speedy deployment, the Centers for Disease Control and Prevention (CDC) took an active role in providing guidance to local public health laboratories, which received supplemental funding from BioWatch to perform sample analysis. This effort was led by the White House Office of Homeland Security—BioWatch moved to the Science and Technology Directorate within DHS when DHS was created—and involved extensive interagency cooperation among the CDC, EPA, the Federal Bureau of Investigation, and numerous state and local government agencies, including public health, emergency management, law enforcement, and local environmental protection. “As you can imagine, this was a monumental task to do in a hurry and is actually an interesting feat in history,” Runge said. At the same time, the Department of Defense (DoD), the U.S. Capitol Police, and the U.S. Secret Service established and maintained independent biosensor networks with little sharing of information about plans, systems, network design, technology, or results.
In April 2004 the White House issued Homeland Security Presidential Directive 10, Biodefense for the 21st Century, which was based on four pillars of activity: threat awareness, protection and prevention, surveillance and detection, and response and recovery (White House, 2004). Under this directive, DHS created the Office of the Chief Medical Officer and eventually moved the operations of BioWatch out of the Science and Technology Directorate and into the Office of Health Affairs
(OHA). The Science and Technology Directorate remains focused on research and development to support BioWatch autonomous detection.
Operational Challenges of BioWatch
From the program’s beginning, Runge said, there were operational issues with regard to the decision making that would result from a positive BioWatch signal. The first issue involved jurisdiction. Terrorism is a national issue with national jurisdiction, Runge explained, but local governments are at the “tip of the spear—it’s their skin in the game regardless of whose jurisdiction it is.” As a result, early conference calls often had more than 100 people (representing local, state, and federal stakeholders) on the line with no uniform, consistent operational construct. Nonetheless, the system appeared to be working, and it led to a number of interesting discoveries. For example, BioWatch sensors revealed that there is Brucella in the air around county fairs and that atmospheric levels of Francisella tularensis increase during times of drought. Runge noted that if funding had permitted, immunological studies could have been conducted in these populations to further understand the nature of such patterns. However, rather than investing time and money to make these types of discoveries, he said, “our [DHS’s] purpose was to protect the nation.”
Speaking of the top-down approach that was taken when BioWatch was first established, Runge likened it to setting up North American Aerospace Defense Command (NORAD) radars around the country and asking local police departments to monitor the radars and call a national clearing office if they detected incoming missiles. Though the local laboratories welcomed the additional funding and added laboratory resources that initially came with BioWatch, the system was deployed with little local input, and the technology was not designed with the requirements of local laboratories in mind. As a result, he said, resistance to the program developed among many jurisdictions, and the program suffered from a lack of a consistently coordinated operational construct. There was also some loss of faith and support among local jurisdictions, although BioWatch did compel interactions among agencies and individuals who, historically, had never spoken to each other.
All shortcomings aside, the basic requirement for biodetection—the ability to maximize survival in affected populations—has not changed, Runge said. What has changed, though, is that the level of understanding about biological weaponry has become more sophisticated, and, as a re-
sult, the technology and overall system need to change to reflect the advancement of knowledge. One of the problems that technology must address is that the timeline of an attack has changed (see Figure 2-1), given advances in the ability to shape the size and characteristics of particles so that they penetrate deeply into the lungs. This technology, which Runge said is readily accessible over the Internet, has made the bioweapon threat greater. “High doses and small particle size truncate the incubation period to the point where the geographical distribution of casualties may look more like an incendiary or nuclear detonation, with a cluster around which we won’t save anybody because of their dose,” he said.
In response to the current understanding of the characteristics of bioweapons, DHS has issued a set of requirements for BioWatch Generation 3 that includes autonomous detectors with technology that is at least as sensitive and specific as the current system using PCR. The detectors
FIGURE 2-1 Normalized timeline of an aerosolized anthrax attack.
NOTE: DHS = Department of Homeland Security; EPA = Environmental Protection Agency; IC = Incident Command; PEP = postexposure prophylaxis.
SOURCE: Runge presentation, June 25, 2013.
must operate in continuous mode to reduce response time and should be able to preserve the captured live agent to enable fingerprinting and forensics, Runge said. The cost targets at the time the requirements were issued were $50,000 per unit and $10,000 per unit in annual operational and maintenance costs. Runge said that it is important to note that the responders to the Generation 3 request for proposals were mainly smaller technology companies and defense and aerospace companies; the broader biomedical industry was not well represented. He also mentioned that a planning scenario for New York City put the cost of an aerosol anthrax attack at $1 trillion.
Runge briefly discussed the development of what was called Generation 2.5, a multiplexed assay that was designed to serve as an interim solution that could be deployed indoors on a limited basis. A few machines were purchased and deployed in New York City, again with limited coordination between federal and local agencies. An inconclusive result from an autonomous unit produced a crisis of confidence within the New York police department counterterrorism unit, and the system was turned off. At the same time, CDC declared at a national BioWatch meeting that it had no confidence in this multiplexed assay. As a result, no other machines were deployed.
In closing, Runge said that several important points can be drawn from the history of BioWatch. Above all, the timely detection of biological agents is crucial as the first step in enabling the timely deployment of medical countermeasures to the public, agent identification, and immediate notification and simultaneous situational awareness for federal, state, and local officials. Detection delays greater than 24 hours severely limit the usefulness of any biodetection system, as does a distributed network with scores of owners. Runge said that in his opinion, “We need central authority and central control, but with full integration and buy-in from local operational authorities.” From an operational perspective, all signals from any detector will require some adjudication by human experts, which suggests that sensors of various methodologies, costs, and discrimination may be useful in a layered warning system. Finally, he said, two important lessons from history are that careful joint operational planning provides ownership of the endeavor and that the hardware to accomplish the mission should be tailored to planning and training.
CURRENT BIOWATCH PROGRAM AND AUTONOMOUS DETECTION
Michael Walter, detection branch chief and BioWatch program manager in OHA at DHS, began his presentation by reminding the workshop audience that the BioWatch mission came out of President George W. Bush’s 2003 State of the Union address and was established by executive order. Its mission is to detect a biological attack, identify the agent used, and prepare a response in time to minimize the impact of the attack. Walter described it as “a very straightforward mission, but very complex to integrate and accomplish” and added that, over time, a number of other tasks have been tacked onto the program. Initially, BioWatch was deployed in 10 high-value cities, but program officials were soon asked if the system could be used for other high-value targets, such as events that might be under threat. They were also asked to accomplish the program’s mission better, faster, and more cheaply, which, Walter noted, was the impetus for the workshop. BioWatch is a state-of-the-art system, but it is labor-intensive and therefore expensive, so there are drivers to move from a manual to an autonomous system.
One area that has received a great deal of focus during the past few years has been the effort to provide guidance at the federal, state, and local levels about how to respond to BioWatch information. “We all know that there is no way to respond to a bioterrorism event on the fly,” Walter said. “This has to be planned out. Everybody has to have an understanding of what you are doing and what your actions are going to do to affect other agencies that are going to be responding.” Furthermore, BioWatch does not exist in a vacuum, so it must be interoperable with other systems that have been deployed by other agencies, Walter reiterated.
Currently, BioWatch is deployed in more than 30 jurisdictions nationwide, and it employs more than 150 people working in the field, more than 130 in laboratories, and more than 100 public health professionals who are integrated with the program, in addition to the dedicated program staff that organizes and coordinates the program. For the most part, BioWatch sensors are deployed outdoors with limited indoor operations. BioWatch sensors are deployed on a case-by-case basis at special events affecting national security (e.g., a presidential inauguration) and local special events such as parades, marathons, and even the Super Bowl at the request of local jurisdictions. Walter noted that although there has been controversy about what constitutes a false-positive, the program maintains that from a technical, analytical, and operational
standpoint, there has never been a false-positive. “We have been right 100 percent of the time,” said Walter. “What we have seen on our filters is there, and we have confirmed that with the CDC.”
Reiterating what Runge said earlier, Walter said that BioWatch depends on partnerships, and he added that he considers the state and local BioWatch Advisory Committees (BACs) the crowning glory of the program. These committees, he explained, force organizations that do not generally talk to each other, that often do not like each other, and that compete for funding with one another, to sit around the table and figure out how to deal with specific situations. These BACs have become the nexus for all hazard response committees and serve as a bridge not only between the federal government and local authorities but also between state and local agencies. He said that the BACs and BioWatch program officials have established a coherent, organized, and disciplined notification process that “lets us know what happened, where it happened, what’s being done about it, and how we are responding to it, plus what is needed from the federal government to start the response.” None of this structure will change when BioWatch transitions to an autonomous detection system, he said.
The key to the program, Walter said, is the planning and preparedness (see Figure 2-2) that will enable local jurisdictions to be the first to respond to a bioterrorism attack, with local authorities applying their expertise and knowledge of their own cities and counties. BioWatch assists in planning and preparedness by providing guidance documents that point out various considerations that the local authorities can use as they formulate their plans. Walter noted that the current structure for planning and preparedness activities was designed to address the chaotic operational structure that existed in the program’s early days.
BioWatch also provides training and test exercises so that local jurisdictions can see if their plans work. “If we do those two things correctly—if we plan and prepare and we train and exercise based on whatever technology is available—then deploying technology becomes more transparent and more effective for the state and local governments that have to use it,” Walter said. He added that public health agencies tend to be conservative about new technologies, so these activities are important for building trust and allowing public health officials to become comfortable with new technologies when they are deployed. Both planning and preparedness activities and training and exercise activities are coordinated by the public health and preparedness group in the BioWatch program office. Included in this office are jurisdictional coordinators who support
the BACs and help shoulder the organizing activities that state and local public health authorities would otherwise have to manage.
Through these activities, BioWatch program officials have developed a good idea of how individual jurisdictions are going to respond to an attack, but there is concern about how the system will operate in the face of multiple attacks or when an autonomous system that collects as many as 10 samples per day from every machine is deployed, Walter noted. “We need to get in front of that now, because we’re not going to catch up once we start a deployment,” he said.
FIGURE 2-2 BioWatch program scope.
NOTE: BAR = BioWatch Actionable Result; NRF = National Response Framework.
SOURCE: Walter presentation, June 25, 2013.
Operational Components of BioWatch
The daily operations of the program have three components: field operations, laboratory analysis, and public health and preparedness. At its root, Walter said, BioWatch is a tool of the public health domain. Field operations center on what is essentially a vacuum cleaner pulling air through a filter. This system is dependable, inexpensive, and easy to deploy, but it requires a person to go out every day to collect the filters and bring them to the laboratory according to strict standard operating procedures. Adherence to these standards is checked by independent audits by an outside evaluator.
The laboratory operations are the most sophisticated part of the BioWatch program. All analyses are performed using PCR that identifies target organisms through DNA signatures. The assays themselves are designed by DoD’s Critical Reagents Program and CDC’s Laboratory Response Network. Walter characterized these assays as being thoroughly validated, very specific, extremely sensitive, and highly dependable. Based on the results from those assays, the local laboratory director or designee can declare a BioWatch Actionable Result (BAR). Declaring a BAR is not a federal event, he noted, although CDC and the BioWatch program office can help with a technical discussion as to whether the assay results make sense. “But in the end, it is the state and local laboratory director that declares the BAR and initiates the response,” Walter said. The analytical strategy is straightforward, he added. A primary screen serves as a low-pass survey for organism identification. Positive results are followed by a verification panel that provides expanded analysis with highly discriminating DNA signatures.
Walter explained that a BAR does not necessarily mean that a terrorist attack has occurred or that a viable biological agent has been released (see Figure 2-3). It also does not mean that an agent is infectious or that there is a risk to the public’s health. What a BAR does mean is that a filter collected from a specific location contains genetic material from an organism that is tested by the BioWatch system. This result is actionable only after assessment by the state and local BioWatch advisory committee.
FIGURE 2-3 A BioWatch Actionable Result.
NOTE: BAR = BioWatch Actionable Result; PCR = polymerase chain reaction.
SOURCE: Walter presentation, June 25, 2013.
Future Autonomous Detection Systems
In the future, BioWatch plans to develop an autonomous detection system, which, Walter noted, is not necessarily the same as the Generation 3 program that has been put on hold pending an independent analysis of alternatives, which is currently ongoing. The primary reason to move to autonomous detection is time—the faster that an attack is detected, the more lives can be saved (see Figure 2-4). An additional reason is that autonomous detection would increase system capabilities and coverage without sharply increasing costs. Current technology, he said, can cut the casualty rate almost in half compared with a situation in which there is no detection system in place. Autonomous detection could reduce the casualty rate even further by cutting the time between when exposure occurs and when a BAR is declared. With the current system, the time between release and declaration of a BAR is somewhere between 12 and 36 hours. Newer technologies developed in the first phase of acquisition can produce results in 3 to 6 hours (see Figure 2-5). Walter noted that this reduction in the time it takes to get results “could dramatically decrease the time between potential exposure and confirmation of the result, giving more time back to state and local public health … for a more effective response.”
FIGURE 2-4 The primary benefit of early detection through autonomous detection lies in reducing the casualty rate and number of fatalities following a release.
SOURCE: Walter presentation, June 25, 2013.
FIGURE 2-5 Comparison of detection timelines: Event-to-detection and confirmation.
NOTE: BAR = BioWatch Actionable Result; PCR = polymerase chain reaction.
SOURCE: Walter presentation, June 25, 2013.
One effect of successfully developing an autonomous detection system will be that operational concepts will need to be redone in coordination with state and local public health authorities to account for the increased sampling and broader geographic coverage, noted Walter. “We need to make sure response plans are effective, based on that technology,” Walter said in his closing remarks. “We need to make sure we have exercises that link the technology and the response. The only way we will have a successful deployment of any technology is going to be if state and local public health have confidence in that system, which means we have to have a robust quality assurance program backing it up, and we need to be able to prove at the state and local level that these machines see what they are supposed to see and they don't see what they are not supposed to see.”
Commenting on Walter’s remarks, John Vitko, rector of St. Luke Church and former director of Biological and Chemical Countermeasures for the DHS Science and Technology Directorate, reiterated that the deployment of an autonomous system will not change the fact that local public health officials will still be responsible for declaring a BAR. He also stressed that CDC and DHS officials will still be available for consultations prior to declaring a BAR.
A workshop participant asked how BioWatch shares costs with local jurisdictions. Walter said that BioWatch has contractors who work for the program in local laboratories but the program does not cover laboratory costs at the state and local level. BioWatch does provide funding to collect filters and deliver them to the laboratories as well as to cover the costs of planning training exercises. Eric Eisenstadt, an independent technical consultant, asked about the uniformity of responses to a BAR across local jurisdictions. Walter responded that there are slight differences reflecting local situations but that the initial response across jurisdictions follows a template and occurs on the same timeline. Eisenstadt also questioned use of the term “BioWatch Actionable Result” as potentially confusing and wondered if there might be a need to change the language. Walter responded that all state and local authorities involved in the BioWatch program are familiar with this term and have a shared, clear understanding of what it does and does not mean. He noted, too,
that his office consulted with NORAD regarding the term “BAR” and was told that it was satisfactory.
Raymond Mariella, Jr., a senior scientist at Lawrence Livermore National Laboratory, asked how long it would take to develop a new assay for an autonomous system for a previously unidentified organism. Walter responded that it could take a year or longer because, in addition to the assay itself, there would need to be accompanying bioinformatics. Although reagents can be ordered quickly through CDC and DoD, the primary reagent suppliers for the BioWatch program, Walter said that validation is needed to ensure that the assay is working accurately, and this process could take a substantial amount of time.