Recognizing Covert Exposure in a Population
Most of the previous chapter assumes a scenario much like the explosives-based incidents that have been typical of terrorism to datea sudden and highly localized event producing casualties almost immediately. This is a reasonable assumption for incidents involving most chemical agents, but it not likely to be accurate in the case of sulfur mustard, whose effects may not be obvious for several hours, and nearly all of the biological agents, whose effects are almost always delayed. This chapter will therefore focus on what we have previously termed scenario two, a covert attack with an agent producing signs and symptoms in those exposed only after an incubation period of several hours, days, or weeks, when the victims might be widely dispersed.
In these circumstances, effective medical response to many covert terrorist actions will be critically dependent not upon Hazmat teams, emergency medical technicians, and other ''first responders," but upon the ability of individual clinicians, who may be widely scattered around a large metropolitan area, to identify, accurately diagnose, and effectively treat an uncommon disease. Education about the threat posed by bioterrorism, and about the diagnosis and treatment of possible agents deserves high priority, but the agents most frequently considered threats are rarely seen in U.S. cities and are therefore likely to remain quite far down any differential diagnosis. The identification of a single outbreak or a series of unusual disease presentations or deaths by the local or state public health department may therefore be the first clue that a cluster of disease may be related to the intentional release of a biological agent,
unless the perpetrators reveal themselves. An analysis of the distribution and number of reported cases will provide important clues regarding the source of infection and can be used both to guide law enforcement and to help all physicians in the community make a rapid and accurate diagnosis of new cases and begin optimal treatment without delay. A very large and efficient attack may truncate the dispersion of cases in space and time (even in this case, victims will not be affected immediately, as they would be with chemical agents), making effective intervention difficult even as it makes it more obvious that the cause is a deliberate release. Rapid and accurate epidemiologic investigation will nevertheless be a key factor in minimizing suffering and loss of life in bioterrorist incidents. Surveillance systems for collecting reports of such cases and appropriately trained staff to monitor for disease outbreaks are the foundation of public health epidemiology. Yet over the last few decades there has been severe erosion in the capacity of public health departments to conduct disease surveillance and epidemiologic investigation.
Surveillance and Investigation of Biological Agents
A rapid evaluation by public health epidemiologists is absolutely critical. Delays in determining the scope and magnitude of the exposure may result in illness and deaths that might have been avoided if a rapid response, based on accurate and timely surveillance data, was made.
Surveillance systems can be passive or active. The large majority of surveillance systems in place at local, state, and federal levels is passive. These rely on systems of disease reporting from health providers and are notorious for their poor sensitivity, lack of timeliness, and minimal coverage. They are inexpensive to implement, but the quality of information is greatly limited, and most are not well suited to the needs of modern disease surveillance, including that needed in the case of a biologic terrorist event. The single exception appears to be electronic laboratory reporting (described below), which can give useful and timely data for epidemiologic purposes.
The Centers for Disease Control and Prevention (CDC) oversees a large number of passive infectious disease surveillance systems. These systems are based on voluntary collaboration with state and local health departments, which in turn depend on physician-initiated reports of specific diseases or information from state health laboratories regarding bacterial or viral isolates. The best known system is the National Notifiable Disease Surveillance System, which the CDC describes as the backbone of collaborative reporting procedures involving clinicians, state, and local health departments, and the CDC. Clinicians, hospitals, and laboratories
in each of the 50 states, the District of Columbia, and the territories are required (by their own laws) to report cases involving any of a list of approximately 50 diseases. The list is compiled and periodically revised by a collaboration of state and CDC epidemiologists (federal agencies cannot legally dictate to states which diseases should be reported). It currently includes several of the diseases this report focuses on as possible bioterrorism agents: anthrax, botulism, brucellosis, plague, and eastern and western equine encephalitis.
Each state and territory has its own list of reportable conditions, most of which overlap with the national list. Certain other medical conditions likely to be caused by bioterrorists may be reportable conditions in selected states, but this is not consistently true across the nation.
The reliability of passive surveillance systems is often quite low, especially if a physician or hospital fails to make the initial report or does not do so in a timely manner. While many states may have legal penalties that can be brought to bear against a provider who does not report, such penalties are almost never imposed. Neither are there any real incentives to report. In the case of an illness due to an exotic biological agent, reporting to proper public health authorities may be more likely to occur than with an illness caused by a common pathogen. If a terrorist uses a common pathogen, it may be difficult to determine the mode of transmission, considering the normal background disease incidence.
Little or no federal funding is provided to state and local health departments to support the surveillance of general communicable diseases. The ability of the states to support infectious disease surveillance has declined in recent years. A 1993 survey indicated that 12 states had no professional position dedicated to surveillance of foodborne and waterborne diseases (Osterholm et al., 1996). Although most states do have some infectious disease surveillance capacity, it is most often supported by categorical (i.e., disease-specific) funds, for example, Acquired Immune Deficiency Syndrome, vaccine-preventable diseases, sexually transmitted diseases. These funds generally cannot be used to support noncategorical communicable disease activities.
CDC has begun an Emerging Infections Program (EIP), under which grants are awarded to state or local health departments for improving epidemiological and laboratory capability. One of the basic benefits the EIP provides for state-based communicable disease surveillance is more trained professionals to follow through with investigations that might not have been initiated because of limited resources. This is one effort designed to ensure there are enough trained public health epidemiologists to maintain a working surveillance system, follow up on cases obtained from that system, and conduct the necessary investigation to develop evidence for causation.
In seven states supported by EIP, active population-based surveillance for selected diseases (foodborne diseases, opportunistic infections in inner-city HIV patients, community-acquired pneumonias, febrile and diarrheal illnesses in migrant farm workers, and unexplained deaths in young adults) is under way. Expansion of these or similar efforts to all fifty states, if accompanied by an aggressive telecommunications effort to make the resulting data widely and easily available, would be a good start towards remedying the domestic surveillance shortfalls identified by a previous IOM report (Lederberg et al., 1992). Although certainly not adequate in itself, this remedy could serve as the model for improving the eroded public health surveillance and investigation infrastructure necessary for preparing the country to adequately respond to a biological agent release.
Active surveillance, which requires staff to actively search for and identify new cases, provides more timely and accurate information than the commonly used passive systems. To conduct active surveillance, state and local health departments must have sufficient numbers of adequately trained epidemiologists. These scientists collect, compile, analyze, and interpret the epidemiologic data to determine the source of the infectious agent. Additionally, capacity is required to conduct the field investigations that are dictated by the surveillance data or by case reports.
A national effort to improve active reporting is the expansion of the Sentinel Surveillance Networks, supported by the EIP program. In this effort, selected reporting relationships are established with medical specialty groups like the Infectious Disease Society of America, the International Society of Travel Medicine, and a group of about 100 emergency departments called the Emerging Infectious Disease Network. These specialists are likely to treat persons with infectious diseases and are periodically contacted about any specific or unusual conditions.
Further research into the potential of electronic disease reporting from (and to) physician's offices is also warranted. Large medical practices, public health clinics, and managed care settings often have patient medical records in electronic form, or other electronic documentation that may identify a case of a reportable (or potential biological terrorism-related) illness. It would be helpful to determine how to develop the capacity to have certain records and information automatically sent to proper health authorities. The legal authority for release of certain information already exists in most state health codes and some local health ordinances. Issues of confidentiality and security must be addressed at the same time as the technological constraints are tackled. An important group of professionals that should not be overlooked in such a surveillance effort is the veterinary medicine community. They are often familiar with a number of the biological agents (anthrax, plague, brucellosis, tularemia, and the
equine encephalitides are all far more common in animals than in the U.S. population), and their incidence and prevalence in local livestock and wildlife.
The CDC is currently examining the development of an emergency department (ED)-based surveillance system called Data Elements for Emergency Department Systems (DEEDS). The DEEDS system is designed to standardize electronic emergency department reporting across clinical systems of care. The system could be the core of a reporting process which identifies diseases of public health importance on a daily basis. Systems such as this could be strengthened to provide real-time surveillance systems of communities at risk. Such a system would not only identify established diagnoses but could also report the actual laboratory request for certain diagnostic assays. These tests would include those done to look for unusual diseases with potential for biological terrorism (i.e., a culture for anthrax). Other systems that track the utilization of specific antibiotics or vaccines (i.e., botulism) could also serve as an early warning system.
The DoD Global Emerging Infections Surveillance and Response system (DoD-GEIS) is designed to conduct antibiotic resistance surveillance at the six DoD tropical medical research units and coordinate emerging infections surveillance for the three services. The U.S. Air Force Global Surveillance Program is an attempt to integrate several surveillance efforts in a way that tracks emerging diseases. Systems such as this could be utilized to track diseases of concern for biological terrorism.
Internationally, projects such as the Canadian Bacterial Disease Network (CBDN) link university and government laboratories in an effort to track diseases of importance. Health Canada, Canada's federal health department, is developing a system to scan the Internet for evidence of new disease outbreaks, utilizing a new search engine technology designed to constantly search a large number of predesignated sites, popular as well as scientific. The system, called the Global Public Health Intelligence Network (GPHIN), will identify disease reports and link them to the World Health Organization's information system.
Numerous other surveillance systems track disease transmission and antimicrobial resistance both nationally and internationally (Appendix B), but there are few meaningful links among them. Mechanisms to integrate and link these systems should be developed.
Detecting and characterizing an outbreak caused by a covert release of a biological agent can be difficult, or it may be startlingly obvious. A reported case of anthrax in an area of the country where anthrax is never reported or in an individual with no obvious risk factors for the disease would raise the suspicions of the public health epidemiologist. Although intentional infection would not necessarily be the first explanation investigated, a process of elimination or additional case reports would eventually
lead to serious consideration of this possibility. The time it takes to reach this point is the rate-limiting step in societal response. It is, then, a critical infrastructure resource and expertise problem of national importance. Without a sufficient number of adequately trained epidemiologists at the local and state level, there may be significant delays in identification and response. In the case of a biological event, lost time may quickly translate into lost lives.
Formal training of epidemiologists occurs in schools of public health, medical and veterinary schools, and in on-the-job-training in health departments around the nation. The number of epidemiologists who are prepared for field public health work is limited. The applied public health sector competes poorly with academia and industry for new epidemiology graduates. In addition to those trained at medical schools and schools of public health, the CDC trains a cadre of Epidemic Intelligence Service officers (EIS), who are available to assist state and local epidemiological response. The EIS was created during the Korean War in response to fears about biological weapons and the perception that state and local public health resources were inadequate to deal with disease outbreaks (Langmuir and Andrews, 1952). Now, nearly 50 years later, facing a threat from these same weapons in the form of biological terrorism, our nation still finds itself understaffed and underprepared.
Current public health epidemiology staff often lack access to authoritative guidelines and other information regarding treatment and control measures. In medium to small jurisdictions, the person serving as the epidemiologist may have little formal training in the concepts and methods of field epidemiology and surveillance and have no background in biological or chemical threats. Most large states and cities employ trained epidemiologists, but few of those individuals have any sophisticated knowledge about biological or chemical weapons. There are national experts who have a significant amount of knowledge regarding biological agents and the consequences of their release, but there is little interaction between these experts and the front-line health department epidemiologists.
No hard data exist on the current knowledge of state and local health departments regarding chemical and biological terrorism. From that information, specific educational and training efforts could be developed and implemented for all state and local health departments. These solutions might include the development of training meetings, Internet-based training, video conferences, or other information exchange technology aimed at the education of local and state health department staff. Syllabi might include not only the medical perspectives of the diseases of interest, but also information regarding modes of transmission, laboratory considerations, working with emergency responders, educating physicians and other health care providers about these conditions and reporting
them to the proper health agency, psychological aspects, and the special considerations of intentional release of these agents. Additionally, since health departments will be requested to disseminate information about the agents, the circumstances of the release, and control measures, information packages for public use will need to be developed. The Centers for Disease Control and Prevention and national organizations, such as the Council of State and Territorial Epidemiologists (CSTE), the National Association of County and City Health Officials (NACCHO), and others may be well suited to develop and facilitate this training.
In the last decade, advances in information technology have greatly accelerated the speed at which business and commercial transactions and information exchange occurs. However, these advances have, in large part, not reached local, and in some cases, state health agencies. The capacity of state and local health departments to communicate electronically with each other is severely limited. Fewer than 50 percent of local health departments have any capacity for Internet connectivity (electronic mail) (National Association of County and City Health Officials, 1997).
One current system of communicating news and information about worldwide emerging diseases utilizes the Internet and electronic mail capabilities. A project of the Federation of American Scientists, ProMED (Program for Monitoring Emerging Diseases) was established to provide communication among sentinel stations around the world capable of detecting unusual outbreaks of infectious diseases or toxic exposures, including those that might result from a biological attack. Scientists from around the world, including many national and global infectious disease experts, participate in a daily exchange of information regarding infectious disease cases and outbreaks. Currently there are more than 10,000 participants in this communication system, which provides a forum for discussing disease occurrence. While ProMED may not be directly useful in emergency notification of authorities of an unusual, potential biological agent release, it could serve to quickly bring experienced scientists together electronically for discussion of the situation. This would improve the response capabilities of public health departments. Additional research into the use of ProMED or other Internet-based information sources (e.g., Outbreak, and Communicable Disease Prevention and Control) appears warranted.
Laboratory Capacity and Surveillance
The variable and often substantial delay between exposure to a biological agent and the onset of clinical signs and symptoms, as well as the possibility of person-to-person transmission, makes rapid and accurate diagnosis important, even if treatment of the earliest patients cannot be
guided by laboratory findings. Protection of healthcare workers and treatment of delayed victims of the attack and secondary victims infected by contact with an early victim will be much enhanced if exposure can be confirmed and treatment started prior to symptom onset.
In each case of suspected exposure, appropriate diagnostic samples from blood, serum, stool, saliva, or urine are needed for laboratory identification of the specific agent. Franz et al. (1997) list the following diagnostic assays: (1) Gram's stain for anthrax and plague; (2) serology, including enzyme-linked immunosorbent assay (ELISA), agglutination, immunofluorescent assay (IFA), hemagglutination inhibition, and antibody (AB) ELISA for anthrax, plague, Q fever, tularemia, viral encephalitis, viral hemorrhagic fevers, botulinum toxin, and staphylococcal enterotoxin B; (3) culturing for brucellosis, plague, and tularemia; (4) Wright-Giemsa stain for plague; (5) virus isolation for smallpox, viral encephalitis, and viral hemorrhagic fevers; (6) electron microscopy for viral hemorrhagic fevers; and (7) polymerase chain reaction (PCR) for identifying the genetic material of smallpox and hemorrhagic fever viruses. These assays may take anywhere from 2 hours to 30 days to complete, and, in the case of smallpox and the hemorrhagic fevers, demand Biosafety Level 4 procedures (e.g., controlled-access laboratory, change to laboratory coveralls and shower on exit, work conducted in fully enclosed, separately ventilated biological safety cabinet [CDC and the National Institutes of Health, 1993]). Even research facilities with this level of protection are not common.
Although many hospitals and commercial laboratories have the necessary equipment and expertise to perform these and similar assays, these diseases are extremely rare in the United States, and so these laboratories rarely perform these assays. Most laboratories will thus not be prepared to immediately conduct the specific analytical test needed to identify the agent, even when the attending physician is astute enough to ask for the test. Veterinary diagnostic laboratories may be more likely sources for a rapid confirmatory assay, since many of the biological agents are significant sources of disease in domestic animals (nonhuman species may even be the first victims of a biological attack, unintentionally or as part of a effort to avoid discovery). Veterinarians and veterinary laboratory workers, moreover, are likely to have been vaccinated against many zoonotic diseases, and used to working with these agents. In no case, however, will a test be done unless a suspicious physician requests it. A more likely scenario is a round of tests for more common pathogens, followed by or concurrent with some symptom-based treatment. Continuing deterioration of the patient's condition or an unusually large number of affected patients may then lead to involvement of a state health department laboratory and state epidemiologist.
Although the capabilities of these laboratories vary widely among the
states, all have working relationships with the CDC and can call upon CDC and other federal and university laboratories for help. Because these organisms and diseases are seldom seen in the United States, there are few experts in these diseases even at the CDC (additional evidence of resource erosion), and the CDC may need to call upon USAMRIID if one of these diseases is suspected. Although this channeling of samples from the initial round of victims to a single expert organization will help in identifying an outbreak, ensure that medical and laboratory personnel are protected, and facilitate rapid diagnosis of those delayed or secondarily infected patients, the process by its very nature is quite slow and will provide very limited benefit to the first victims. Research into the utility of developing a network of regional laboratories (including veterinary laboratories) capable of rapid diagnostic testing may be useful. Ideally, such a network would involve strengthening diagnostic expertise and laboratory capability at all major medical centers, but given the expected frequency of cases involving biological warfare agents, a more realistic goal might be a regional approach based on state and local public health laboratories.
Knowledge regarding the laboratory recovery of a pathogen or positive diagnostic assay is a key element that public health epidemiologists will need in their investigation. While, in general, an epidemiologic investigation can be initiated without laboratory confirmation, final confirmation of the exact nature of the pathogen will be needed. In some cases, the laboratory report may serve as the initial notification that there is a human illness associated with a microorganism that is known to be a potential terrorist agent. Such a report does not make a diagnosis, nor does it suggest how the pathogen was acquired by the patient. However, it could serve as an early notification system for public health, thereby improving the chances of responding quickly to additional cases and saving lives.
Currently, biological samples are taken from the patient and delivered to the laboratory for analysis. After identification through culture or diagnostic assay, the results are sent back to the treating facility for use in medical management of the patient. Most communicable disease surveillance systems rely on telephonic or weekly written reports for subsequent transmission of this information to local or state health officials. Even if a suspect atypical pathogen is identified and interested health care staff make special efforts to report, the time lag before the health department knows of the culture may exceed 3 days.
In most large laboratories, tracking of specimens and results is done electronically via computer. Recent efforts, spearheaded by the National Center for Infectious Diseases, at the CDC, and the CSTE have focused on the possibility of electronic reporting of laboratory results. This beginning effort has demonstrated that many large laboratories are capable of downloading assay findings in formats that are usable by state health officials.
Originally designed to reduce the workload of reporting results to over 50 jurisdictions, electronic laboratory reporting may serve as an important step towards getting important information into the hands of the epidemiologist quickly. Additional research towards full development, and then national implementation of electronic laboratory reporting will improve the public health response to a biological release. These solutions will also strengthen the national disease surveillance infrastructure in general, a glaring need detailed in the IOM report Emerging Infections: Microbial Threats to Health (Lederberg et al., 1992).
An important new laboratory development is the ability to sequence different parts of microbial genomes. By identifying distinct features of different genes, it is possible to identify not only microbes of interest but specific strains and thus more precisely track infectious disease outbreaks. This "fingerprinting" technique will be useful as a sentinel indicator that a new strain has entered a community and in distinguishing natural from intentional releases by identifying microbial or viral strains foreign to the normal community flora or by matching new outbreak pathogens with pathogen strains from suspect terrorist groups or intelligence sources. Such a library of genetic fingerprints would also have enormous value not only in the effort to track ongoing outbreaks, but also in predicting antigenic shifts for vaccine production..
The beginnings of a national network of state health laboratories utilizing this type of fingerprint technology and sharing fingerprint libraries is under way. Known as PulseNet, this system, officially initiated in 1998, links the CDC, U.S. Department of Agriculture, and the U.S. Food and Drug Administration to a network of state laboratories using pulsed-field gel electrophoresis to look for characteristic DNA patterns of organisms implicated in foodborne infections The system is currently capable of identifying patterns of E. coli 0157:H7 and tracking them across 12 states. Expansion of this system to other states and additional pathogens is planned and could form the backbone of the national network suggested above. Including more states is relatively straightforward; including more pathogens is far more involved, and probably depends not only on identification of stable and accessible DNA and RNA sequences that are unique to particular pathogens or classes of pathogens, but also on simpler, faster, user-friendly microbiological assays. Chapter 6 will address the possibilities of such improvements and their application to both patient diagnostics and environmental monitoring and testing.
Although the rapid onset of symptoms in a large number of victims may make it obvious when a chemical attack has taken place, poison control centers (PCC) can serve as the basis for a surveillance system for
recognizing covert or multiple-site poisonings with chemical agents or biological toxins. The PCC is also the logical place to turn in that scenario for advice on treatment.
There are no studies defining the theoretical or actual skills of poison information specialists and physician staff with regard to chemical or biological warfare agents, but these individuals are well prepared to provide advice to emergency personnel in the field and at hospitals on the basis of signs and symptoms. The PCC may also serve as a coordination point, even when the ''incident" is confined to a narrowly circumscribed locale, for victims may be dispersed to a number of different medical facilities. The Tokyo incident involved release of sarin in five trains on three different subway lines; 278 medical facilities received patients in the following 48 hours (Sidell, 1996). The value of a medical information coordinating center in such a situation cannot be overestimated. Unfortunately, poison control centers have also faced years of declining resources, with many centers performing few of the essential tasks to recognize, much less contain, a developing epidemic. This trend must be stopped or the nation may face a shortage of medical toxicology specialists who can provide critical treatment information to treating physicians. Additionally, linkages to the public health system do not exist in many states and should be encouraged.
At the national level, the Agency for Toxic Substances and Disease Registry (ATSDR), instituted a hazardous substances emergency events surveillance (HSEES) system in 1990 (U.S. Department of Health and Human Services, 1993, 1994b, 1995a). State health departments in selected states collect and transmit information on the circumstance and health outcomes surrounding hazardous materials releases. The information is generally provided well after the event. The information in HSEES is made publicly available for use in locating, training, and equipping Hazmat teams, first responders, and employees as well as guiding follow-up epidemiology. However, the time constraints associated with an incident of chemical terrorism will not permit a significant real-time surveillance role for the HSEES.
Aids for Clinical Diagnosis Based on Signs and Symptoms
Emergency medical personnel, both at the scene of a hazardous materials incident and at hospital emergency departments, have a wide selection of reference materials to call upon for guidance in patient management. These include traditional textbooks, handbooks such as the three-volume set of medical management guidelines prepared by the Agency for Toxic Substances and Disease Registry (U.S. Department of
Health and Human Services, 1994a), paper, CD-ROM (Poisindex, Drugdex, Emergindex), or on-line Material Safety Data Sheets. Some of these sources provide information on nerve and mustard agents, but most of these resources are organized by chemical rather than by symptom complex. That is, given some independent knowledge of the identity of the hazardous substance, one can readily ascertain the likely effects and appropriate treatment. Deducing the substance from the effects is far more difficult. Poison control centers are routinely faced with this problem and are a good source of help.
The Washington, D.C. MMST has addressed this difficulty by incorporating a symptom checklist tool called the "NBC Indicator Matrix" into their training (Defense Protective Service, 1996). First developed by the Defense Protective Service, which provides security at the Pentagon, it is a paper-and-pencil checklist of symptoms. A system for scoring and processing the results leads to a suggested agent or agents. For Hazmat incidents at the Pentagon, and by definition, for incidents which lead to a request for help from the MMST, first responders are highly likely to turn to such a tool. At other locales, they may need some reason to suspect terrorism to consider using the matrix. The matrix may have some utility even in its present form, especially in cases in which exposures are mild. Difficulties seem likely when victims are critically ill or have pre-existing illnesses, or in incidents involving more than a single agent.
With the exception of some of the toxins (botulinum, SEB, T-2 mycotoxin), and possibly the hemorrhagic fevers, the initial signs and symptoms produced by the biological agents considered here are nonspecificfever, chills, fatigue, headache, muscle or joint pain, a cough or chest pain. Blood in the excreta or petechiae (pinpoint-sized, hemorrhagic spots in the skin) may lead an astute clinician to consider a hemorrhagic fever, but few U.S. practitioners are likely to recognize the other diseases associated with biological weapons on the basis of signs and symptoms alone. Correct diagnosis will almost certainly depend on perception of an unusual epidemiologic picture by public health epidemiologists. This is an area where pre-incident intelligence could have a major impact in reducing the number of casualties.
The development of interactive computer-based diagnostic systems that enhance the potential for early recognition and analysis of the unique aspects of rare diseases, including the manifestations of disorders produced by biological and chemical agents, would be a substantial advancement. An integrated system that utilizes natural disease rates and clinical probabilities, based upon signs and symptoms, and laboratory findings could enhance an early warning system prior to a clinician's decision on a particular diagnosis or disease. One model, the Global Infectious Disease and Epidemiology Network (GIDEON), uses a Bayesian matrix for compatible
diagnoses. This particular system, which focuses on unusual diseases, is limited by the weight it places on country of disease origin and requires a database that might not be initially available.
An improved system with refinements in diagnostic decisionmaking that offers the untrained or inexperienced clinician assistance in considering a chemical or biological exposure would be of great value. For any system this would necessitate a complex, multiple search mechanism that includes early signs and symptoms of atypical disorders caused by biological or chemical terrorists agents.
The committee recognizes that the first of the following recommendations is a recommendation, not for research itself, but for the prerequisites for productive R&D. The committee strongly believes that no research effort, no matter how important or sophisticated, will be productive until the nation rebuilds the public health infrastructure to a level at which the results of appropriate research can be properly applied. This infrastructure improvement would have enormous value to the average citizen on a day-to-day basis and would generate significant health benefits beyond readiness for terrorist events.