ASSESSING THE THREAT OF BIOTERRORISM
Raymond A. Zilinskas
Monterey Institute of International Studies
October 20, 1999
I thank the Subcommittee on National Security, Veterans Affairs, and International Relations for having been given the opportunity to testify on the subject of meeting the threat of terrorism armed with biological weapons. The specter of terrorists enhancing their already formidable arsenals by acquiring these weapons of mass destruction is truly a horrendous one, one that we all must do our best to prevent. So, how may prevention best be accomplished? A good start is to try to understand the threat of bioterrorism, use that understanding to craft appropriate barriers, and, should barriers be breached, craft effective means for alleviating ill effects resulting from an attack and the apprehending of the perpetrators. I hope that my testimony will further our collective ability to understand the technical aspects of bioterrorism and biocriminality.
To accomplish this objective, my presentation has four parts. The first provides a brief background of myself and my work in arms control. Second, findings from an ongoing project to assess applications of advanced biotechnologies for terrorist and criminal purposes are presented. Third, I comment on the recently issued report Combating Terrorism: Need for Comprehensive Threat and Risk Assessments of Chemical and Biological Attacks by the United States General Accounting Office (GAO). Fourth, thoughts on what the major biological threats to our society are and what might be done to meet them are presented.
After having graduated from California State University at Northridge with a BA in Biology (1962), and from University of Stockholm with a Filosofie Kandidat in Organic Chemistry (1963), I worked as a clinical microbiologist for 16 years before commencing graduate studies at the University of Southern California. Shortly after having earned a Ph.D. in 1981, I went to work for the Office of Technology Assessment (1981 - 1982). My subsequent jobs have been at the United Nations Industrial Development Organization (1982 - 1986) and the Center for Public Issues in Biotechnology, University of Maryland Biotechnology Institute (1987-1998). Since September 1998, I have worked as a Senior Scientist in Residence for the Center for Nonproliferation Research at the Monterey Institute of International Studies (MIIS). I also am an Adjunct Associate Professor at the School of Hygiene and Public Health, Johns Hopkins University where I teach courses on emerging issues in international health.
In October 1993, I was named William Foster Fellow at the Bureau of Intelligence, Verification and Information Support, U.S. Arms Control and Disarmament Agency (ACDA). In April 1994, ACDA seconded me to the United Nations Special Commission (UNSCOM) (see below). After the Foster fellowship ended on December 31, 1994, I returned to the Center for Public Issues in Biotechnology.
I began to think about biotechnology's possible negative effects in 1980, when performing research to complete my doctoral dissertation. At that time I was the supervisor of the clinical microbiology section of an acute care hospital's laboratory. Most of my responsibilities related to recovering and identifying pathogens, performing antibiotic testing on bacteria, and so on. I also served on the hospital's infection control committee, attempting to prevent and eliminate nosocomial infections. My work gave me opportunity to reflect on what I call the "pestilence triangle," which encompasses the complex interrelationships and balances between host, parasite (or pathogen) and the environment. Much of the activities in the medical sciences are focussed on trying to affect the components of this triangle for the benefit of the host. Thus, from the aspect of the host, some research seeks to devise methods whereby microbial invasions can be prevented or the host's ability to resist infection may be augmented. As to the parasite, the prevention of infectious diseases is sought through vaccine and anti-toxin R&D, while therapeutics, such as antibiotics, germicides, and other agents, weaken or kill the pathogen. Perhaps most important, the characteristics of the "natural" environmental that support the growth of pathogens past a certain critical threshold, or favor the dissemination or transmission of pathogens, may be affected by public health measures, such as proper waste disposal, water treatment, and air filtration. I came to realize that biological weapons, either when being developed in the laboratory or when actually used in the field, have the potential to upset the intricate balances that exist within the pestilence triangle by either altering the environment or upsetting parasite-host interactions. In either case, the host will suffer damage or death.
The objective of my dissertation was to analyze policy issues generated by recombinant DNA research, including the applicability of genetic engineering techniques to the weaponization of pathogens and toxins. Eventually, the part addressing BW was to take up fully one-third of the dissertation (Zilinskas, 1981). My thoughts on this subject were previously been developed in an article published in 1978, which included a discussion of terrorism and biological weapons (Zilinskas, 1978). I analyzed the Sverdlovsk anthrax epidemic in an article published in 1983 (Zilinskas, 1983). In 1986, I edited a book that contained an article in which the possible applications of biotechnology for BW were analyzed (Zilinskas, 1986a; Zilinskas and Zimmerman, 1986). I examined the difficulties pertaining to verifying the 1972 Biological and Toxin Weapons Convention (BWC) in a book chapter (Zilinskas, 1986b) published by the Stockholm International Peace Research Institute (SIPRI) (Geissler, 1986). In 1990, I wrote an article that investigated whether terrorists were likely to acquire biological weapons and concluded that they would do so within a fairly short time (Zilinskas, 1990a). That same year, I discussed biological weapons and the Third World and concluded that their allure to leaders of these nations was high (Zilinskas, 1990b). Also in 1990, I co-authored a United Nations report on the applications of biotechnology for arms control (Geissler and Zilinskas, 1990). In 1991, I co-authored a book chapter with a renowned Swedish scientist that argued for the activation of Article X of the BWC and the involvement of bioscientists in arms control activities (Zilinskas and Hedén, 1991); this chapter is found in a book published by SIPRI (Lundin, 1991). Also in 1991, I organized a conference called "The Microbiologist and Biological Defense Research: Ethics, Politics and International Security," which was held at the University of Maryland; its proceedings were published by the New York Academy of Sciences (Zilinskas, 1992a). My contribution to this book focussed on the need to establish an early warning system for suspicious and unusual disease outbreaks that might have been deliberately induced (Zilinskas, 1992b). In 1993, I was part of an interdisciplinary study group that considered the threat of biological events and analyzed whether local and state emergency personnel were trained and otherwise prepared to meat this threat (Bradford et al., 1993, 1994). Our conclusion was that they were not.
As noted above, I was a Foster Fellow at ACDA during late 1993 and all of 1994. However, the agency seconded me to UNSCOM, where I worked during April November 1994. Here I was responsible for setting up a database containing data about key dual-use biological equipment in Iraq and developing the first draft of the protocol that was to guide UNSCOM's on-going monitoring and verification program in the biological field. I was also UNSCOMs representative on two biological weapons-related inspections in Iraq (June and October 1994) encompassing 61 biological research and production facilities. Drawing on my experience at UNSCOM, I have written four articles or chapters that bear on international arms control (Zilinskas, 1995ab, 1996a; 1997).
After returning to the Center for Public Issues in Biotechnology in January 1995, I wrote an article that discussed whether the attacks carried out by the Aum Shinrikyo was a paradigm for future terrorist operations (Zilinskas, 1996b). In addition to fulfilling academic responsibilities, I worked for ACDA on an ad hoc basis as a long-term consultant. At ACDA I have performed a costing analysis of the future BWC compliance regime (Zilinskas, 1995c), analyzed the utility of lists for the protocol now under development to strengthen the BWC (Zilinskas, 1995d), and assessed problems related to the conductance of challenge inspections (Zilinskas, 1995e). Most recently, ACDA asked me to investigated Cuban allegations of the U.S. having waged BW against its human, animal, and plant populations; the results of that study were published in September 1999 (Zilinskas, 1999a).
At MIIS my research is mostly focused on effective biological arms control, the proliferation potential of the former Soviet Unions BW program, and meeting the threat of bioterrorism and biocriminality. One of the first tasks I undertook was to revisit the problems pertaining to verifying the BWC and analyze how the protocol now under development might make this task more effective (Zilinskas, 1998a). In early October 1999, the book Biological Warfare: Modern Offense and Defense, edited by myself, was published (Zilinskas, 1999b); it includes a chapter that I co-authored on the ethics of BW-related research and the role of bioscientists in preventing illicit research and development (R&D); i.e., R&D for offensive military, terrorist, and criminal purposes (Colwell and Zilinskas, 1999).
Beginning in February 1999, I have been working with faculty at the Center for Counterproliferation Research at the National Defense University (NDU) on a project that aims to assess the applicability of modern biotechnology techniques for terrorists and criminals (this project is discussed below). We expect that its findings will be published in two reports. The first will address terrorism or criminality directed against humans; this will be published in early 2000. The second report, to be issued in the middle of next year, will analyze the possibilities of BW against, animals, plants, and materials.
3. Advanced Biotechnologies and Their Possible Applications for Terrorism and Criminality
The NDU/MIIS terrorism project aims to assess the impact of recent and anticipated advances in biotechnology on R&D undertaken for the purpose of perfecting biological weapons. In general, it is exceedingly difficult to forecast developments that can be expected in the future from rapidly evolving and growing scientific fields. Because of its proven value for identifying areas of consensus or disagreement on issues presented to the involved experts, the focus group approach was selected for forecasting (Morgan, 1993). Accordingly, the principal investigators established a focus group constituted by 16 natural and social scientists who possess a wide variety of expertises. The scientific disciplines represented in the focus group include aerobiology, biotechnology, bioprocessing, human medicine, meteorology, microbiology, molecular biology, phytology, toxincology, veterinary medicine, and virology. Specialists in criminality, industrial practices, and terrorism also participated.
The focus group was asked to analyze a number of newly developed and emerging biotechnology techniques in terms of their utility in R&D that aims to produce microorganisms of enhanced military or terrorist utility. However, lessons from the history of the pre-1969 U.S. BW program and the Soviet BW program (this program supposedly was terminated in 1992, but some analysts believe it continues in some form in present-day Russia) indicate that research, development, and production to weaponize pathogens and toxins is only a small part of the total process of acquiring biological weapons. The most important parts of the acquisition process are developing "formulations," merging formulations and munitions to produce an efficient weapon system, and designing and producing a mechanism for dispersing pathogens or toxins over a target. This being so, the NDU/MIIS focus group was also asked to examine whether advanced biotechnologies may be used to enhance the ability of agents to withstand stress brought about by storage, dispersal, and physical and chemical environmental factors. This, in turn, necessitated giving thought to how agents may be used and for what purposes. The focus group was asked to limit its consideration to the next five years; i.e., up to and including 2005. The focus group was not to consider "classical" microbiology except to provide background or for the sake of comparison.
Much of what follows in Section 2 is derived from my analysis of initial focus group deliberations. However, as the NDU/MIIS project is still underway, its findings are yet to be finalized. This being the case, the descriptions, opinions, thinking, and conclusions set forth below are mine and do not necessarily represent those of either the NDU or any of the members of the focus group.
Biological weapons can be designed and used to injure and kill not only humans, but also animals and plants. Some security analysts believe that the greatest biological threat facing the U.S. is, in fact, the possible use of biological weapons by terrorists or criminals to wage economic warfare by destroying animal and/or plant populations important in agriculture (Rogers et al., 1999). However, the expertise of the present NDU/MIIS focus group is such that it is best capable of analyzing scientific/technical advances pertinent to BW as waged against humans, so the possible uses of biological weapons by criminals and terrorists against animals, plants, and materials will be the subject of a subsequent NDU/MIIS study.
My analysis of focus group proceedings, as well as of information derived from other sources, leads me to draw certain conclusions about the types of biological attacks terrorists can mount, the most likely scenarios for biological attacks in the immediate future, less likely but still possible scenarios for biological attacks in the immediate future, the possible applications of advanced biotechnologies for terrorist or criminal purposes in the next five years, and the possibilities that advanced biotechnologies offer to terrorists and criminals in the more distant future.
A. Types of Biological Attacks by Terrorists and Criminals
In general, terrorists or criminals can carry out three types of biological attacks. First, the pathogen or toxin may be injected. This method is best used when the terrorist or criminal wishes to assassinate an individual. Since individual assassinations are not likely to stress our emergency response and health delivery systems, they are not considered further in this paper. Second, a quantity of pathogens or toxins may be used to contaminate or poison foods, beverages, or fomites (such as food supplements and medicines taken by mouth). If done skillfully, this method could cause hundreds of casualties. Third, pathogens or toxins may be suspended in a wet or dry formulation (see below) and dispersed over a target area as aerosolized particles. This type of attack could produce thousands of casualties, if three conditions were met: (1) the formulation was well designed for aerosol dispersal; (2) the aerosol particles produced by the dispersal mechanism were of optimal size and could withstand environmental stresses; and (3) meteorological conditions were just right for blanketing the target area with aerosol particles.
B. Scenario of Likely Biological Attacks
It is highly probable that biological attacks by terrorists or criminals utilizing foodborne and waterborne pathogens or toxic chemicals will occur in the next five years. Much like what has taken place in the past, these attacks are likely to cause casualties ranging in number from a few to hundreds. Examples of past attacks include the contamination of 10 salad bars by members of the Rajneeshee cult in Oregon in 1984, which caused 751 casualties (Török et al., 1997), and intentional food contamination in Texas in 1996, which harmed approximately 15 persons (Kolavic et al., 1997). Events such as these likely will take place with increasing frequency in the years ahead for two main reasons; unprotected, unmonitored salad bars and other food displays have become ubiquitous throughout the U.S. and the number of persons with at least a modicum of training in microbiology is ever increasing (although the population constituted by microbiologists probably is no more or less dishonest or unethical than other populations of professionals, a small proportion of it should be assumed to be willing to lend or sell its skills for terrorist or criminal purposes [see below]). There is nothing original about making this near certain prediction; it is done here mainly for the purpose of developing recommendations stated in the Conclusion.
C. Scenario of Low Probability Biological Attack
The probability that terrorists or criminals will carry out airborne attacks with pathogens in the next five years is low. The reasons are that it is technically difficult to formulate pathogens and toxins for airborne dispersal, to operate dispersal mechanisms successfully, and to ensure proper meteorological conditions for effective aerosol dispersal. For these reasons, this type of attack is too difficult for most terrorist and criminal groups to attempt. The example of the Japanese sect Aum Shinrikyo is illuminative in regard to both pathogens and toxic chemicals (Kaplan and Marshall, 1996; Tucker, 1996). In the biological field, despite having evil intent, a membership that included highly trained bioscientists and chemists, ample funding, and ample time to carry out appropriate R&D, the sect failed utterly to produce effective biological and toxin weapons. It appears that there are two explanations for this failure. First, the sect used an avirulent strain of Bacillus anthracis (the causative agent of the disease anthrax) in their weapons and, second, they used a formulation of pathogens and substrate that clogged up the nozzles of their sprayers.
The problem of formulation, especially formulations for airborne attacks, is a difficult one to overcome. Briefly, after they have been produced, pathogens and toxins must be suspended in formulations in preparation for storage or attack. Possibly the major remaining secret of both the pre-1969 U.S. and pre-1992 Soviet BW programs pertains to the formulation of BW agents. After much empirical experimentation, both programs were able to develop methodologies for suspending or dissolving optimal quantities of weaponized pathogens and toxins in special solutions containing preservatives, adjuvants, and anti-static chemicals. The final emulsion or mixture is what is commonly called formulation. A specific formulation is required for every weaponized pathogen and toxin. Without properly constituted formulation, pathogens or toxins in storage or being transported are likely to loose their virulence or toxicity after a relatively short time (days to weeks); during spraying, solutions containing pathogens or toxins might foul nozzles so that no aerosol is emitted; after being emitted through the spray nozzle, electrostatic attraction between particles made up of pathogens or toxins can cause them to clump (bacteria as colloidal particles have electric charges), after which the clumps will fall ineffectually to the ground; and/or environmental stresses, such as UV light and desiccation, will kill or inactivate the aerosolized pathogens or toxins.
D. Possible Applications of Advanced Biotechnologies for Terrorism and Criminality in the Next Five Years
After having examined the panoply of advanced biotechnologies, two of them appear to hold the most promise for applications by scientists and technicians intent on weaponizing pathogens and toxins -- DNA technologies and genetic and protein engineering.
There is a precedent. Scientists who worked for the Soviet Unions BW program are alleged to have used recombinant DNA to combine certain features of the smallpox and Ebola viruses (Alibek, 1999). The result might have been a new type of virus, one that combined the virulence of the Ebola virus with the hardiness and contagiousness via aerosol of the smallpox virus. Soviet scientists also worked at recovering the influenza virus from the corpses of persons who died of influenza in 1918 - 1919 and were buried in permafrost ground. If they were successful, they might have been able to insert this material in influenza viruses circulating in our time, to produce a new variant useful for BW. Site-directed mutagenesis may be employed in order to change the structure of proteins constituting a bacterium's cell wall so that the modified organism is more difficult to identify or will no longer be recognized by an immune system primed to defend against the parent organism.
Protein engineering might be used by a scientist to stabilize toxin molecules so they better resist the action of, for example, chlorine, do not dissociate if placed in water, or resist heat. Further, as many toxins consist of two subunits (one subunit that ferries the toxin molecule to the cell and/or anchors the molecule to the cell membrane and a second subunit that kills the host cell), the possibility exists that protein engineering could be applied to alter a toxins chemical structure for the purpose of increasing the efficiency of one or both subunits.
The foregoing are examples of what could be done, but I believe it is highly unlikely that any of them will be utilized in laboratories operated by, or working on the behest of, domestic terrorist groups or criminals to weaponize pathogens or operate production systems in the next five years. There are two reasons for this conclusion.
First, complex research undertaken to weaponize pathogens is risky because it is more likely to fail than achieve its objectives. The problem of pleomorphic effects is particularly daunting. Pleomorphic effects are manifested as undesirable characteristics that appear in a genetically engineered organism simultaneously to sought-after positive characteristics. Thus, even if a laboratory succeeded in genetically engineering a pathogen so it exhibited a new or enhanced characteristic desirable for weapons use, such as antibiotic resistance or added toxin production, the newly developed organism might simultaneously present a weakness to environmental stresses and/or decreased virulence. If so, a new cycle of research, development, and field-testing would have to be done to remove the pleomorphic effects while retaining the sought-after characteristics. If the researcher was unskilled and/or unlucky, he or she might have to undertake several subsequent research, development, and testing cycles before being able to field a strain of pathogen that had improved weapons capabilities over the parent strain. As a consequence of potential difficulties with pleomorphic effects, it is likely that in the next five years or more only well supported, long-term national BW programs would attempt genetic engineering projects for the purpose of weaponizing pathogens.
Second, sciences understanding of many natural phenomena, such as infectivity, pathogenesis, host-parasite relationships, and others, is rudimentary. Lack of fundamental information about these phenomena prevents the undertaking of much applied research to, for example, enhance the ability of organisms to infect target hosts, cause severe damage to host systems, and be more specific as to preferred hosts. Further, some important phenomena, such as virulence factors and the ability of a pathogen to penetrate the hosts skin or intestinal wall, are controlled by several or many genes; however, the present level of scientific capability allows bioscientists to transfer or modify only single genes. It therefore is impossible to modify phenomena controlled by multiple genes, thus severely circumscribing approaches to weaponizing pathogens.
E. Possible Applications of Advanced Biotechnologies for Terrorism and Criminality in the More Distant Future
As a corollary to the preceding finding, the likelihood of advanced biotechnologies being applied successfully for terrorist or criminal purposes will increase substantially during the time 2006 2010. The reason is that much basic research is currently being undertaken pertaining to mechanisms of pathogenesis, including pathogenicity islands; histocompatability complexes; retroviral control and ecology; and bacterial genomics. Information generated by this research will fill many of the gaps discussed in Section A, thereby providing markedly increased opportunities for research to weaponize pathogens. Further, the Human Genome Project will be completed in or shortly after 2003; when this occurs, the human genome constituted by 80,000 100,000 genes will have been mapped. As the human genome is being mapped, a new field called functional genomics is growing; functional genomics aims to clarify the functions of identified genes. It is probable that some of the findings generated by functional genomics could be applied for such purposes as disrupting or destroying physiological functions by designing pathogens, toxins, and naked DNA that will target genes controlling these functions.
It has been suggested that functional genomics some day will generate data that may be used to identify genetic markers peculiar to specific human populations. If this was done, pathogens and toxins might be developed that mainly affect persons of populations possessing specified genetic markers (Larson, 1970; Hammerschlag, 1974; Lancet editorial staff and International Advisory Board, 1996; British Medical Association, 1999). This type of weapons has been called "ethnic" weapons; i.e., weapons that preferentially harm or kill designated national or ethnic populations. While stories about ethnic weapons make for exciting reading, the research required to develop an ethnic weapon would be extremely difficult, have a high probability of failing, take a long time to carry out, and be expensive. Realistically, the probability of such research being undertaken at all is low; and even if it were to be done by, for example, a well-supported national program, it would probably take more than 25 years to realize findings meaningful for biological weapons development.
To conclude this section, there are six reasons why biological and toxin weapons are likely to become ever more attractive to criminals and terrorists as we move into the 21st century. First, as the biotechnology, pharmacology, environmental, and health delivery industries grow, the number of persons possessing expertise in microbiology and the biosciences will increase greatly. It is reasonable to expect that a small proportion of this population will be willing for reasons of greed, ideology, or fear to apply techniques in these disciplines for criminal or terrorist purposes. Second, information on how to produce and disseminate pathogens and toxins is readily available in open sources. Someone with a modicum of education and training in the microbiological and biotechnological sciences can easily access this information and probably would be able to adapt it for the purpose of weaponizing agents. Third, a tiny quantity of a pathogens or toxin delivered effectively can cause many persons to become ill and die. Fourth, tactical weapons utilizing pathogens or toxins can be designed so that they are easily hidden. Therefore, it would be unlikely that a terrorist or criminal transporting and using a biological weapon would be discovered by either police or nearby citizens. Fifth, the delivery and use of pathogens and toxins do not necessarily require sophisticated methods. In particular, it is not technically difficult to contaminate food or beverages, which could cause hundreds of casualties. Due to significant technical difficulties, it is unlikely that terrorists or criminals will be able to deliver pathogens by aerosol, so a biological attack utilizing the airborne method is unlikely to occur in the next five years. Sixth, there are no defensive technologies available that are, or could be, deployed at civilian facilities to detect and identify deliberately disseminated pathogens or toxins in real or near real time. The fact that a biological attack has occurred would therefore not become known until some time later, when the pathogens incubation period has passed and many individuals become ill nearly at once.
4. Comments on the Report Combating Terrorism: Need for Comprehensive Threat and Risk Assessment of Chemical and Biological Attacks
In view of limited time and space, I will concentrate on what I believe to be the reports main problem, namely, the performance a "sound" risk assessment. Specifically, although much is made of the need to perform a risk assessment of domestic terrorists deploying biological and chemical weapons, there is nothing said about how such a risk assessment should be performed.
To begin, the Report states:
At first glance, this reads fairly well. However, when examined, the statement raises several questions in the readers mind. The authors do not define key words, such as "valid" and "sound." The do not explain NIEs and how they are derived. Without this information, the statement becomes at best unclear and at worst meaningless.
More information on risk assessment appears to be provided in a paragraph that begins on the bottom of page 5. Unfortunately, its contents are hopelessly muddled, jumping from risk assessment to threat assessment, then to risk management, and back to risk assessment. Regard the following sentence: "A threat analysis the first step in determining risk identifies and evaluates each threat on the basis of various factors such as its capability and intent to attack an asset and the likelihood and the severity of the consequences of a successful attack." Can a "threat" have capability and intent? Can it attack? Do the authors equate "threat" with a terrorist group? Or a biological or chemical weapon? What in civilian-speak is an "asset"? Are populations, individuals, police, and/or cows assets?
A bit more information of risk assessment is provided on page 19. After assuring the reader for the third time that "Risk assessments are widely recognized as valid decision-making support tools to establish and prioritize program requirements" (there is that word "valid" again; would anyone use "invalid" tools for decision-making?), the authors suggest the use of a multidisciplinary team of experts to: "(1) generate valid attack scenarios; (2) assess and rank the risks (likelihood and severity of consequences) of attack scenarios; and (3) decide on actions and programs focused on reducing or otherwise dealing with risks as assessed." Continuing, the authors state "Risk assessment should include sound inputs and information, such as the best available intelligence and law enforcement information and analysis, including NIEs and Intelligence Community Assessments. Soundly established requirements could help ensure that specific programs and initiatives and related expenditures are justified and targeted, given the threat and risk of validated terrorist attack scenarios." (p. 20)
As can be realized, the authors of the GAO report argue for the performing of a "sound" risk assessment (versus an "unsound" risk assessment?) of possible future terrorist events that would generate more than 1,000 casualties using "valid" information from various trustworthy sources they name. However, nowhere in the report do they offer suggestions or advice on how such a risk assessment ought to be done. The authors write of interviewing what reads like a large number of experts in every sector of our society. It would seem that some of these experts could have provided concrete guidance on how to conduct the type of risk assessment recommended in the report, but that is not done.
Could it be that a "sound" risk assessment as recommended in the report would be extremely difficult or impossible to perform? I now take the opportunity to explore difficulties inherent to performing risk assessments.
During 1996 1998, I was involved in a multidisciplinary project that aimed to determine risks that would attend the introduction of genetically engineered macroorganisms (such as fish and shellfish) and microorganisms (mainly bacteria) into the open marine environment. The results of that project were published in September 1998 (Zilinskas and Balint, 1998). My main task was to consider genetically engineered marine microorganisms and the possible risks they would pose to human health and environment were they to be released into the oceans (Zilinskas, 1998b). As a result of this work, I learned something about performing scientific risk assessment.
In science, the formula for estimating risk is as follows:
RISK = HAZARD x EXPOSURE
(Risk is the magnitude and likelihood of adverse effect.
Hazard is the harm the agent will cause.
Exposure relates to what population will be exposed to the agent,
at what concentration, and for how long.)
Some risk assessments are relatively straightforward. For example, the estimating of risk associated with adding lead to gasoline would not be difficult for a trained, experienced risk assessor. Hazard can determined because the harm done by various concentrations of lead on humans is known and the amount of lead emitted in exhaust gases of automobiles can be measured. Similarly, exposure can be determined by counting the number of automobiles traversing a locale of interest during a measured time, the half-time of lead in the environment is known, historic meteorological data may be consulted to determine dispersal patterns of gases over the local of interest, and demographic data may be mined for information on the population of the local of interest.
Estimating possible risks attendant to the introduction of a genetically engineered microorganism into the open terrestrial environment was at one time very difficult to do. When the first such introduction was proposed, the U.S. Department of Agriculture (USDA) and the U.S. Environmental Protection Agency (EPA) had great difficulty developing a protocol for appropriate risk assessment that should be done before a decision could be made whether the proposed introduction would be forbidden or allowed to proceed. Eventually, the EPA was given the authority to regulate proposed introduction of genetically engineered microorganisms into the environment, while the USDA has authority over proposed introductions of genetically engineered plants. Both agencies developed risk assessment protocols for proposed introductions that have been followed by researchers and industrialists for over 10 years. As this is written, over 3,000 genetically introductions of genetically engineered plants and microorganisms have taken place in the U.S., with no apparent harm (hundreds of introductions have also taken place in many other countries of the world). Thus, it would appear as if the risk assessment protocols developed and used by the EPA and USDA has done what they were supposed to do, namely, they protect human health and the environment while allowing possibly risky but economically beneficial activities to proceed under specified conditions.
In 1990, the EPA developed "21 Points to Consider" (United States Environmental Protection Agency, 1990), which lays the basis for risk assessment. It is unnecessary here to provide a detailed listing of these points; it is sufficient to say that a developer of a genetically engineered microorganism must provide information that satisfy five criteria familiarity with the organism donating genetic material, familiarity with the organism receiving genetic material, familiarity with the environment of the site onto which the proposed introduction will take place, ability to contain the introduced organism to the designated site and, should containment fail, knowledge of the damage that the escaped organism would cause to human health and/or the environment. If information can be provided that satisfies these familiarity and containment criteria, the EPA is able to perform an adequate and appropriate risk assessment of the proposed action.
When I attempted to determine the risks that might attend or result from the introduction of genetically engineered marine microorganisms into the open marine environment, I found that EPAs 21 Points to Consider could not be satisfied. The main reasons had to do with lack of familiarity with the marine environment; the inability to contain microorganisms to the site of application because of currents, eddies, and other natural forces; and the lack of knowledge about possible damaging actions by escaped microorganisms. In other words, although the EPAs 21 Points to Consider appear to provide a satisfactory basis for risk assessment in the marine environment, the lack of fundamental scientific information about marine organisms and the marine environment precludes the performing of an adequate risk assessment of a proposed action involving an application of genetically engineered microorganisms in oceans.
The point of the foregoing is that it is not possible to perform a meaningful or adequate risk assessment in all cases because the information to do so is not available. If solid information is lacking, the tendency might be to substitute assumptions for information. The more assumptions that are made while performing a risk assessment, the less rigorous will be the analysis.
For illustrative purposes, it is worthwhile considering how one might go about trying to assess whether a domestic terrorist group is likely to use biological weapons capable of causing more than 1,000 casualties. The first question that would be answered is, which terrorist groups in the U.S. possess the necessary expertise to acquire and deploy biological weapons that depend on aerosol dispersion of pathogens for effect? I do not know how many terrorist groups exist in the U.S., but it may be a sizeable number. Perhaps the FBI and/or local police have information on each of these groups membership, including whether microbiologists, biotechnologists, or other technical people belong to them. However, since relevant expertise is the key ingredient to any endeavor in applied microbiology, be it peaceful or ill willed, without this basic information, one cannot perform a risk assessment of terrorist capabilities in applied microbiology.
To get around this stumbling block, let us assume that a certain proportion, say a conservative 0.1%, of any population of workers will consist of bad persons; i.e., individuals who would be willing to use their skills for nefarious purposes. The population of scientists and technicians trained in microbiology probably numbers approximately 100,000; based on the foregoing assumption about 100 of them would be willing to apply microbiology for terrorist or criminal purposes. We do not know how many of these individuals belong to terrorist or criminal groups and how many of them would prefer to be lone operators. Let us make another assumption; fifty belong to groups that might wish to mount biological attacks and 50 are potential microbiology equivalents to the Unabomber.
Assuming even distribution of ill-willed microbiologists, then 50 groups in the U.S. each has one microbiologist and thereby possesses the requisite capability to acquire biological weapons. This brings up the next difficult problem to the risk assessor; which of these groups have leaders that intend to deploy biological weapons? This problem may be illustrated by referring to the microbiology technician Larry Wayne Harris. During an interview conducted in September 1999 by a German reporter, Harris was asked whether he would use biological weapons. He replied "If God tells me to do it, I will." No risk assessor would be in a position to determine if and when God gives Harris, or others of his ilk, the requisite command.
The intentions of others have proven equally difficult to ascertain. For instance, no outsider, as far as is known, was able to divine the intent of the Aum Shinrikyo to use biological and chemical weapons before it actually deployed a chemical weapon in 1993 in Tokyo. No one outside Iraq knows why Saddam Hussein, who possessed sizeable BW and CW programs, decided not to arm saboteurs or terrorists with biological and chemical weapons before and during Desert Shield and Desert Storm. Might he decide to do so in the future? Might Fidel Castro attempt to strike back at the U.S. with biological weapons in retribution for the many biological attacks he perceives Cuba has experienced? A risk assessor cannot know the answer to questions such as these.
Whatever the weapons system, the question whether or not it will be used depends on the intent of the leader or leadership controlling it. No risk assessor will be in a position to know the intent of the leadership of microbiologically capable domestic terrorist groups. Therefore, more assumptions have to be made when assessing the risk these groups pose to our society. The easiest would be to say that if a group has the capability, it should be assumed that it will acquire biological weapons. It follows that if a group possesses biological weapons, it should be assumed that it will use them.
It can be seen from the discussion presented above that fundamental information about capabilities and intent of domestic terrorist groups is lacking and probably cannot be obtained. This being the case, the analyst would have to make a series of assumptions if he or she wished to perform a threat or risk assessment of any one group. The product of such a threat or risk assessment would, to my mind, be worthless. If our government cannot use results from risk assessment to guide decision-making, what might it do to meet the so far theoretical threat of bioterrorism?
The major biological threat facing U.S. society are infectious diseases of natural origin, in particular, emerging infectious diseases, reemerging infectious diseases, and transported infectious diseases (Lederberg et al., 1992). An example of the first was AIDS in the early 1980s and the Hantavirus outbreak in Four Corners in 1993. These types of diseases typically seem to appear out of nowhere and may cause tremendous damage and untold suffering among a susceptible population. Examples of the second type include the reemergence of cholera in South America after an absence in that continent since the early 1900s. There could be many reasons why diseases that have not been seen for a long time reemerge. In the case of cholera, a combination of factors probably was responsible, including an unusual El Nińo condition and a breakdown in sanitary systems (Colwell, 1996). An outbreak of Marburg hemorrhagic virus disease outbreak in Germany earlier this year and the just concluded outbreak of West Nile fever in the New York area are examples of the third type. In these cases, the causative infectious agents are transported from an area where they are endemic to a new site where they have never been detected previously. As with emerging infectious diseases, transported infectious diseases are likely to come into contact with a population that is immunologically naďve, and therefore eminently susceptible.
In comparison to the real and enormous threat of emerging, reemerging, and transported infectious diseases, the problem of deliberately caused disease is almost insignificant. From a public policy perspective, it would make sense to pay much more attention to the larger problem while not neglecting the smaller one. However, that is not the situation at present in the U.S.; the overwhelming attention of executive agencies, the legislative branch, and the concerned public is affixed on the theoretical problem of bioterrorism and not on natural infectious diseases. Fortunately, this is not necessarily a bad development.
When a disease outbreak is first detected, no one is in a position to know if it has a natural or laboratory etiology. Thus, the initial public health and medical response to a disease outbreak will be the same whatever its etiology. Public health practitioners get busy trying to determine the etiology of the disease of concern by applying classical and molecular epidemiological techniques, while health providers treat those who have taken ill and try to prevent secondary spread of disease.
Our societys response to a natural versus deliberately caused disease outbreak would differ only after there are clear signs that the disease of concern might be the result of a terrorist or criminal attack. An explosive outbreak of a disease striking hundreds or thousands of persons and whose etiological agent normally is spread by aerosol would be one such sign. Another, more subtle sign, would be a cluster of cases where the causative agent is a foodborne or waterborne pathogen and a specific food or beverage source seems to be its likely source. Once the suspicion arises among public health and/or health delivery personnel that the disease of concern was deliberately caused, then law enforcement officials will also become involved, gathering evidence that might lead to the arrest and prosecution of the perpetrators.
Since the initial response by the public health and health delivery systems would be the same to an outbreak of disease whatever its origin, it follows that if the ability of these systems to respond effectively and appropriately to any medical disaster were enhanced, our societys ability to cope with both natural disease outbreaks and the aftermath of terrorist attacks would increase. If the following three steps were taken, I believe a significant increase in the ability of public health and health delivery systems to respond to disease outbreaks would result.
First, there seems to be a general agreement among emergency responders that no municipality in the U.S. is prepared to deal with the aftermath of a massive outbreak of disease; i.e., one that generates thousands of casualties within a few days. I suggest that assessments be made by each major city and state to determine what it would take for the city or state in question to prepare for dealing with the aftermath of such a health disaster. The assessment might be conducted in phases. Thus the first phase, or the generic phase, might determine what is needed to process, treat, and house a thousand casualties whatever the cause of their illness. A second phase might do the same, but would consider 10,000 casualties. A third phase might consider the special conditions that would have to be taken into account if a contagious disease agent caused the 1,000 or 10,000 cases of illness. After that, the special situations brought about by individual diseases might be clarified.
Second, each municipality should perform a study that includes assessing its ability to respond to disease outbreaks of lesser magnitude than what is discussed in the preceding paragraph and clarify possibilities for receiving assistance should its response ability be surpassed. The purpose of each assessment would be to determine the maximum numbers of casualties that the municipality in question could handle without outside assistance, make known the assistance that it could count on to receive from the state should the maximum be exceeded and, should a massive disaster strike, set up procedures for requesting federal assistance, including necessary military forces. The federal government might consider providing funding and expert assistance to those municipalities who need them in order to conduct these studies.
Third, the reporting system in the U.S. for infectious diseases needs to be significantly improved. During the last three years, significant improvements have been made to this system as a result of the current bioterrorism threat. However, this is still not a high priority item; out of the approximately $ 10 billion being spent annually on countering the threat of terrorism, less than $ 200 million is earmarked for public health surveillance and reporting. The reporting system should be sufficiently build up so that it would generate information that could be quickly analyzed for indications of emerging, reemerging, transported, and deliberately caused infectious diseases. To do this, more intensive training on how to detect and report unusual disease outbreaks must be given to local public health personnel and emergency medical personnel. Further, a certain proportion of law enforcement people should be given training in public health to the extent that they could spot indications of deliberately caused diseases and know what evidence they need to collect to verify the cause of these diseased and to track down and apprehend perpetrators. The possibility of establishing a detection system for automatically and continually surveying the entire Internet for information indicative of suspicious and unusual disease outbreaks should be considered. The Canadian government has set up an Internet-based surveillance system to track influenza on a worldwide basis; this experience should prove useful when setting up the surveillance system proposed here.
To conclude, it is not likely that rigorous risk assessments can be done of threats posed by terrorists or criminals armed with biological and chemical weapons. This being the case, risk management of the terrorist threat has to be undertaken on an empirical basis. Some steps to this end are suggested in this paper, such as improving the ability of federal, state, and local police to analyze information indicative of illicit biological activity, improving the ability of public health and health delivery personnel to deal with the aftermath of disease outbreaks, and improving the national system for detecting, surveying, and monitoring disease outbreaks. What is not so obvious and what I try to make clear in this paper is that there is need for placing the threat of bioterrorism in perspective the greater biological threat facing the U.S. is not terrorists armed with biological weapons, it is, as it always has been, diseases of natural origin. If we can successfully meet and defeat the real threat of emerging, reemerging, and transported infectious diseases, then we have also gone a long way towards being able to handle whatever manifestation of bioterrorism that will occur.