Summary

INTRODUCTION

Biological warfare agent (BWA) detectors are designed to provide alerts to military personnel of the presence of dangerous biological agents. Detecting such agents promptly makes it possible to minimize contamination and personnel exposure and initiate early treatment. It is also important, though, that detectors not raise an alarm when the situation does not warrant it. Such “false positives” may result in unnecessary precautionary measures and could impact a mission’s success. For biological agents, establishing an appropriate level of sensitivity for an alarm is especially difficult because biological agents are diverse and vary widely in infectivity and virulence. Thus, detectors ideally should be capable of discriminating between highly infectious or virulent agents—to which they must be very sensitive—and innocuous ambient environmental microorganisms or low-risk ones—for which an alarm would be undesirable.

It is a challenge, then, to choose what physical characteristic detectors should measure, since the effect of interest is inherently biological activity. Ideally, a single characteristic, common to all hazardous agents, and accurately quantifiable, would be most useful. Such a characteristic would make it straightforward for a test and evaluation program to compare the accuracy and sensitivity of different detectors. The U.S. Department of Defense (DOD); Joint Program Executive Office for Chemical and Biological Defense (JPEO CBD); Joint Project Manager, Nuclear, Biological and Chemical Contamination Avoidance (JPM NBCCA); Product Director, Test Equipment, Strategy and Support (PD TESS)1 sought a standard unit of measure that could be used for biological material independent of the state of the material and independent of agent type. PD TESS asked the National Academies to conduct this study, addressing the use of measurement in the test and evaluation of aerosol detectors and evaluating whether the standard unit currently in use, Agent-Containing Particles per Liter of Air (ACPLA,) is the most appropriate measure and what alternatives exist. As the name suggests, ACPLA is a measure of the number of particles in a liter of air that contain a viable biological agent. Formally, the term applies to bacteria, but it can be modified to refer to viruses or toxin molecules. The question considered in this report is whether ACPLA is an appropriate unit of measure for use in the evaluation of aerosol detectors and whether a better, alternative measure can be developed.

This report discusses several problems with the use of ACPLA. First, ACPLA does not distinguish between “active” and “inactive” biological agents. (While the term “active” can have different meanings, for the purpose of this report the term is used as shorthand for the form of the biological agent that can cause harm—live bacteria, infective viruses and active biotoxins. The term is more general than “viable”, which is specific to bacteria and refers to the capacity to grow in culture, as opposed to the ability to cause harm.) Second, ACPLA does not distinguish between virulent and innocuous strains of bioagents. Third, even as a simple measure of quantity, ACPLA falls short because it does not distinguish between particles containing a single

1

The abbreviation PD TESS will be used to indicate this organization.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
Summary INTRODUCTION Biological warfare agent (BWA) detectors are designed to provide alerts to military personnel of the presence of dangerous biological agents. Detecting such agents promptly makes it possible to minimize contamination and personnel exposure and initiate early treatment. It is also important, though, that detectors not raise an alarm when the situation does not warrant it. Such “false positives” may result in unnecessary precautionary measures and could impact a mission’s success. For biological agents, establishing an appropriate level of sensitivity for an alarm is especially difficult because biological agents are diverse and vary widely in infectivity and virulence. Thus, detectors ideally should be capable of discriminating between highly infectious or virulent agents—to which they must be very sensitive—and innocuous ambient environmental microorganisms or low-risk ones—for which an alarm would be undesirable. It is a challenge, then, to choose what physical characteristic detectors should measure, since the effect of interest is inherently biological activity. Ideally, a single characteristic, common to all hazardous agents, and accurately quantifiable, would be most useful. Such a characteristic would make it straightforward for a test and evaluation program to compare the accuracy and sensitivity of different detectors. The U.S. Department of Defense (DOD); Joint Program Executive Office for Chemical and Biological Defense (JPEO CBD); Joint Project Manager, Nuclear, Biological and Chemical Contamination Avoidance (JPM NBCCA); Product Director, Test Equipment, Strategy and Support (PD TESS)1 sought a standard unit of measure that could be used for biological material independent of the state of the material and independent of agent type. PD TESS asked the National Academies to conduct this study, addressing the use of measurement in the test and evaluation of aerosol detectors and evaluating whether the standard unit currently in use, Agent-Containing Particles per Liter of Air (ACPLA,) is the most appropriate measure and what alternatives exist. As the name suggests, ACPLA is a measure of the number of particles in a liter of air that contain a viable biological agent. Formally, the term applies to bacteria, but it can be modified to refer to viruses or toxin molecules. The question considered in this report is whether ACPLA is an appropriate unit of measure for use in the evaluation of aerosol detectors and whether a better, alternative measure can be developed. This report discusses several problems with the use of ACPLA. First, ACPLA does not distinguish between “active” and “inactive” biological agents. (While the term “active” can have different meanings, for the purpose of this report the term is used as shorthand for the form of the biological agent that can cause harm—live bacteria, infective viruses and active biotoxins. The term is more general than “viable”, which is specific to bacteria and refers to the capacity to grow in culture, as opposed to the ability to cause harm.) Second, ACPLA does not distinguish between virulent and innocuous strains of bioagents. Third, even as a simple measure of quantity, ACPLA falls short because it does not distinguish between particles containing a single 1 The abbreviation PD TESS will be used to indicate this organization. 1

OCR for page 1
2 organism or molecule and those that might contain hundreds or thousands of units of the bioagent. Finally, the health effects of particles containing bioagents can be dramatically affected by the size of the particle—small particles can be drawn deeply into the lungs where they are more harmful, while large particles may only reach the nasal passages, or not be inhaled at all. ACPLA makes no particle size distinction. The problem with a unit of measure that does not include such parameters can be illustrated by imagining the response to two different threats of a detector measuring ACPLA: in one scenario, an average liter of air contains 100 particles, each carrying one avirulent bacterium. The detector would give a reading of 100 ACPLA. In the other scenario, an average liter of air contains 1 particle carrying 100 extremely virulent, live bacteria. The detector would need to be sensitive down to the level of 1 ACPLA to detect this attack. If the detector were set to sound an alarm at, say, 10 ACPLA, the detector would alert military personnel to the first attack—which is, in fact, harmless—and fail to alert to the second attack—which might be extremely dangerous. Even if the detector were sensitive enough to detect 1 ACPLA, it would still sound an alarm in both situations, where only one warrants taking precautionary measures. The challenges described above are just some of those associated with the measurement of biological species. The complexity of the bioagent threat is such that there is only one relevant characteristic shared by all agents of interest: the capacity to interact with the human body and potentially cause harm. Thus, a unit of measure that considers the health hazard posed by a given concentration of aerosolized biological agent in air would allow comparison across all agent types against a characteristic that would have real utility both for test and evaluation of detectors and in the field. ACPLA is easily understandable and measurable; thus, it is straightforwardly incorporated into a system of detector requirements and evaluation, but it fares poorly in the more complicated task of providing a tool to measure the actual hazard posed by a biological attack or a system’s ability to detect that hazard. ACPLA focuses attention on a generic characteristic (quantity of agent containing particles) that cannot be related, even relatively, to health hazard. Instead, it would be more useful to adopt a framework of measurement that makes it possible to evaluate relative hazard by including agent identity and activity, particle size and infectious dose. The new measurement framework would be more complicated than ACPLA. Not all of the information needed to compare the health hazard of different agents is readily available. Even with imperfect knowledge, however, the new framework could be implemented with current technology. More importantly, implementing the new framework would serve to focus future development efforts on detectors that measure parameters relevant to health risk in both military and civilian settings. CHARGE TO THE COMMITTEE At the request of PD TESS, the National Research Council was asked to evaluate current units of measure for biological aerosols and, if necessary, determine a standard unit of measure that can be used for biological material independent of the state of the material (aerosol or aerosol resuspended in liquid) and independent of agent type (bacteria, viruses, or toxins). The committee addressed the following questions: • Is there a single unit of measure that is appropriate for use in the evaluation of aerosol detectors?

OCR for page 1
3 • What are the possible alternatives to the use of ACPLA and what are the advantages and disadvantages of their use? • Are different measures appropriate in different circumstances? • Is there a robust way to convert between various units of measure? RESPONSE TO THE CHARGE Evaluation of existing units Some of the limitations ACPLA were discussed above. Even the real advantages of ACPLA—its relative simplicity and clear quantifiability—are compromised by the fact that, in practice, ACPLA is measured in many different ways. Some instruments may determine the identity and amount of live agent by culturing concentrated air samples on solid agar medium. Other instruments measure such characteristics as the particle count or sample fluorescence. ACPLA does not explicitly make a distinction between active and inactive agent, and instruments that measure different characteristics may provide different ACPLA assessments of the same sample. For example, an instrument that counts colony-forming units (CFUs) would be measuring live bacteria, while an instrument that counts agent genome equivalents using polymerase chain reaction (PCR) would be measuring both live and dead bacteria. Each could be used to infer the number of agent-containing particles, but they could give different measures of concentration. Finally, because ACPLA does not account for the dramatic differences in the health risks posed by different agents, the recipient of an ACPLA count must know what agent is detected, and understand the health risks posed by the specific strain in order to determine the proper course of action. Therefore, ACPLA alone does not provide enough information to determine the truly relevant information—whether a health threat exists. Another possible unit, agent concentration in irreducible units (e.g., spores), suffers from similar issues as ACPLA. This unit, unlike ACPLA, does reflect the total amount of agent present, but like ACPLA, fails to measure activity or particle size, which can significantly change both the lethality and the infectivity of the attack. Another alternative measure is the number of colony-forming units (CFU) of bacteria, plaque-forming units (PFU) of virus, or mass of toxins per unit volume of air. These measures also lack the detail needed to fully assess hazard. Important factors, such as particle size, are again lost with these units. These measurements, in the case of bacteria and viruses, do provide useful information on biological activity—a necessary condition to determine whether an agent is capable of producing an adverse biological outcome—but they reflect fundamental biological differences of the different classes of agents and are not directly interconvertible. After a survey of the literature to identify units describing the concentration of biological aerosols, the committee concluded that no single unit of measure is available that could be used directly to compare the health hazard posed by different biological agents and the many forms in which they might be encountered in an attack. An integrated framework that allows comparison of hazard

OCR for page 1
4 The committee concluded that in order to be useful and comparable across all biological agents and detection systems, measurements of biological agents must ultimately be related to health hazard. The capacity to cause harm is a quality shared by all pathogens, including those not associated with biological warfare, such as SARS, avian influenza, or the toxins associated with red tide. While this report focuses on biological warfare agents, the concepts presented can be applied to a broad range of airborne pathogens and biological toxins. Two critical factors determine the probability that a BWA aerosol exposure will produce an adverse health outcome—the hazard posed by the agent and the physiological responses of the individuals exposed to it. A cloud of smallpox virus is a hazard, but it may pose no health risk if all of the exposed individuals are effectively immunized against smallpox. The hazard posed by an agent is determined by its physical and biological characteristics: identity and strain of agent, activity, virulence, and particle size. For the purpose of developing, testing and evaluating bioaerosol detectors, it is the hazard that would be measured. This report proposes a unit—Biologically Active Units/Liter of Air or BAULADae— that quantifies the hazard posed by a particular bioaerosol where Dae is the particle aerodynamic diameter. The BAULADae unit can then be embedded in a full health risk framework that places that hazard in the context of the likely physiological impact on the particular population in question. The parameters that are included in BAULADae are summarized in Figure S.1. FIGURE S.1 A framework for evaluating the health hazard posed by aerosolized biological warfare agent exposure. Proposal of BAULA (Biologically Active Units per Liter of Air) as a new framework for evaluation of health hazard The proposed unit of hazard—Biologically Active Units per Liter of Air (BAULADae)—

OCR for page 1
5 is a particle aerodynamic diameter (Dae) dependent measure of biologically active units per liter of air, such that a biologically active unit is the amount of agent required to have a certain probability of causing a negative health outcome. BAULADae incorporates the key information needed to estimate health hazard. Because different threat agents pose different hazards, biologically active units would be measured and calculated differently for each threat agent. For example, two agents present in the same quantity—as determined by CFU—but with differences in virulence would result in different values of BAULADae although their concentration as reported in the current unit, ACPLA, would be the same. Normalizing agent concentration to health hazard enables direct comparisons between all types of agents. It is important to note that all of the factors in BAULADae cannot currently be accurately quantified for every biological agent. For example, the dose required to sicken 50 percent of a particular exposed population is not known for every agent. Nor is it known, for every agent, how health hazard varies as a function of location in the respiratory tract. Extrapolation of animal data to human effects has uncertainties. In addition, obtaining non- human primate data of adequate precision and accuracy is extremely challenging but required due to the low natural occurrence of many of the pathogens of military concern. While the information used to calculate the health hazard may not be ideal (for example, LD50 levels determined in animal studies), even reasonable estimates of health hazard would improve the utility of the information given by detectors compared to a unit that does not even account for activity, much less virulence. The BAULA framework illustrates what type of information would be needed if detectors were able to provide warning not only of agent presence, but also of biological activity that can be related to health hazard. If it can be agreed that this is the ultimate goal for BWA detectors, then the BAULA framework provides an appropriate standard for reaching this goal. Initially, referee systems, and subsequently detector systems, can be developed with the new unit as their reference point, which will guide research and development and assist in identification and prioritization of knowledge and data gaps. Until these gaps have been filled—and some of these data (e.g., non-human primate LD50) may never be obtained—it will be necessary to make appropriate assumptions to effectively use the unit and the framework. Application of the BAULA Framework to Detector Assessment The effective evaluation of detectors requires an effective set of test protocols and requirements.2 During a test, the referee system, which does not have to evaluate the aerosol in real time and may consist of several different instruments, is used to provide an accurate measure of what the detector is being exposed to during testing. Because BAULADae, like ACPLA, requires assessment of the physical properties of the aerosol, existing referee equipment can be used to quantitate the physical characteristics of the BWA aerosol. BAULA also requires biological activity to be assessed as a surrogate indicator of infectivity. The information acquired from these measures will then need to be related back to a chart providing population level lethal dose rates to determine BAULADae. Detector developers and testers operate at the interface of the physical and life sciences; both physical measures (e.g. particle size, presence of an agent) and biological measures (e.g., 2 For the purposes of this report, the term “detector” is used to describe equipment that is under evaluation or used in the field, “referee” corresponds to the description provided here, and “instrument” is used generically to describe the equipment and components in either detector or referee systems.

OCR for page 1
6 viability, virulence, host susceptibility) are needed to fully characterize the hazards of bioaerosols. Currently, and for the foreseeable future, there are insufficient scientific data to have an absolute performance measure for all detectors against all biological agents. Despite this limitation, there is sufficient scientific evidence to allow effective testing, particularly through the comparison of relative performance. These comparisons can be used to link to policy and requirements as well as to prioritize investments for improving detector performance. As an example, consider two candidate detectors and their performances detecting the bacteria Bacillus anthracis and Francisella tularensis. (Dennis et al. 2001; Inglesby et al. 2002) If both agents are equally likely to be used by an adversary, and data suggest that Francisella causes infection at a lower dose, then the detector that is more sensitive to Francisella would be preferred. Over time, this framework could help direct detector research and development to focus on developing measurement techniques to monitor the specific physical and biological characteristics that have the greatest impact on the hazard posed by BWA. It should be noted that real-time detectors themselves will not likely directly measure BAULADae. Because detectors can be designed to measure fundamentally different qualities of the BWA (e.g., CFUs, PFUs, quantity of nucleic acid, immunofluorescence), their outputs are not directly convertible to BAULADae, but can be inserted into the BAULA framework. Referee systems will quantitate the BAULADae level to which the detector is exposed. Reducing the challenge amount until the detector can no longer discriminate challenge from background noise would define the detector’s minimum detectable concentration. A major consideration in adapting current testing procedures to the proposed new framework is that it requires differentiation of active and inactive agents. Current detectors and referee systems measure either active agent (in the case of instruments that measure CFU, for example) or all agent (for example, instruments that measure immunofluorescence) but do not distinguish active from inactive agent. This limitation is currently circumvented by making the conservative assumption that all agent detected is active. However, as technology progresses and instruments are developed that can distinguish between active and inactive agent, it will be important to take into account both whether the detector and referee systems are measuring the same thing and how much of the sample is remaining active throughout the testing process. Loss of activity during the testing process could cause discrepancies between detector results and referee instrument measurements. For example, if the sample preparation and aerosolization processes result in a 50 percent loss in activity, detector A (that senses both active and inactive agent) may appear to have twice the sensitivity of detector B (that detects only active agent) if the referee system is based only on active agent detection. This may translate to an inaccurate assessment that detector A is more capable of detecting the amount of agent needed to pose a health hazard than it actually is. Current testing protocols often require that the aerosolization process produce an aerosol with fairly narrow particle size distributions. Accurate assessment of health hazard would require the generation of particle sizes that reflect different health hazards; the eventual goal would be to develop detectors capable of characterizing these physical characteristics of the aerosol. For now, improvements in the referee system are needed to obtain the particle size data identified in the equation. In the long term, use of the BAULA framework may facilitate the development and deployment of detectors that can determine the sample’s particle size distribution. There is significant value in implementing the use of BAULADae in the near term, even though the scientific basis for quantitatively linking all biological agents to health hazard is currently incomplete. With appropriate documentation of standard operation procedures,

OCR for page 1
7 methods, and reagents, many of the measurements needed to assess BWA exposure in BAULADae units can be accomplished with current detection technologies. For instance, particle sizing and counting; collection rate; sampling time; concentration; and CFU, PFU, and toxin mass can all be implemented today. Adoption of BAULADae as the standard approach to determining hazard will highlight the necessity to develop accurate factors for converting CFU (or other direct measurements) to biologically active units, thereby providing a focus for where research would be most productive. Overall, although the proposed BAULADae unit is more complex than the current ACPLA unit, BAULADae has the advantage both of providing a valid basis for comparison across all biological agents and, when incorporated into the broader risk framework, of providing an assessment with true operational value: the risk to health of those exposed to biowarfare agents. RECOMMENDATIONS Recommendation A: A unit of hazard should be adopted as part of the evaluative framework. The committee recommends the unit be a particle size (Dae) dependent measure of biologically active units per liter of air (BAULADae). Recommendation B: In support of an overall DOD detector evaluation philosophy that relates health risks and aerosol exposure, the committee recommends using the relationship in Figure S.1 to provide a framework for relating health hazard and aerosol exposure and to identify the information provided by different types of detectors. Recommendation C: Standardize testing procedures for evaluating detectors. C.1 Aerosol challenges need to be well characterized, including Dae, BAULADae, BAPLADae (size-resolved number of Biologically Active Particles per Liter of Air), and material rendered inactive during the process of aerosol dissemination and transport. C.2 Challenge biodetectors with aerosols of defined size distributions. At least three challenge aerosols with different median aerodynamic diameters (Dae) should be used in chamber or component testing of detectors. These should be chosen to represent deposition in the three regions of the respiratory tract. C.3: The standardized unit of hazard, BAULADae, and the broader evaluative framework for health hazard should be adopted as DOD-wide standards, including use in T&E and procurement. The method for implementing the unit and framework should be documented, externally peer reviewed, and published. Revisions and updates should follow similar vetting processes so that calibration, referee instruments, and testing reagents are standardized and variation is identified. Recommendation D: Survey the literature to better understand the transport and inhalation of particles to improve selection of appropriate Dae (median aerodynamic diameter) challenge aerosols for test and evaluation. Balance this knowledge with intelligence of possible threats. Specifically examine whether to include testing of particles smaller than 1µm and larger than 10µm. Recommendation E: Maintain ability to learn from unanticipated events by archiving data.

OCR for page 1
8 E.1. Archive parameters, methods, measurements, and test conditions during T&E of detectors. E.2. Consider archiving raw data collected by deployed systems to guide future development of detectors through performance evaluation. Recommendation F1: For priority DOD scenarios, an evaluation of uncertainty in the science, measurements, and situational awareness should be conducted so that resources can be invested in reducing the largest uncertainties that impact decision making. Recommendation F2: An applied science program should be designed and executed to obtain information that will both improve the accuracy of the equation and help evaluate threats and their inherent uncertainties. This will require information about different biothreat agents, instrumentation types, and scenarios important to DOD.