Disasters often set the stage for scientific inquiry that could help mitigate potential short- and long-term health effects among responders, as well as improve capabilities for responding to future events. However, the disaster environment presents many challenges for the conduct of research, including the immediate emphasis on critical response activities, limited access to incident leadership, inability to engage response personnel in scientific research activities not immediately pertinent to the event response, marshalling of necessary resources in an expeditious manner, and timely situational awareness of important occupational safety and health events during the response or recovery operation. The disaster environment is typically dynamic and often hazardous, chaotic, and highly charged with conflicting scientific opinions, political pressures, and disparities in knowledge or the capability to safely perform the required work.¹

Emergency response workers and others involved in response, remediation, and recovery efforts (referred to herein as “responders”) are often engaged in non-routine activities in uncharacterized, potentially hazardous environments and may encounter novel or unique exposures or experience unexpected health effects. As such, there is frequently a pressing public health need to conduct research to answer critical questions regarding the health impact of these exposures and/or determine the factors that are responsible for any adverse health effects experienced in both the short- and long-term. Issues related to conducting research during and following disasters have gained increasing attention following the World Trade Center disaster in 2001 and, more recently, the Deepwater Horizon disaster.^2,3 For example, the National Biodefense Safety Board, an advisory committee to the Assistant Secretary of Preparedness and Response within the U.S. Department of Health and Human Services (HHS), recently engaged on the topic of inclusion of scientific investigations as a component of disaster planning.⁴ A responder research study, however, can be costly and difficult to design; therefore, one must carefully consider whether a proposed or requested study will be able to produce useful, reliable results and is a prudent public health investment.

Providing the foundation for responder health research within the context of a disaster and subsequently determining the need to pursue a study requires careful consideration and planning before an event occurs. Unless the potential for such a study is anticipated, the ability to collect critical baseline data can be lost. Research studies need to address clearly stated, important questions or hypotheses, and be appropriately designed to maximize the likelihood of producing a meaningful study. However, the factors and decision-making processes relevant to determining whether to conduct responder health research, or how to best perform such research, have been relatively unexplored.

For the purposes of this paper, health studies conducted in conjunction with the response to emergency events are divided into four basic types: (1) those that involve routine or baseline health monitoring, health surveillance, industrial hygiene or environmental assessments, responder interviews/focus groups, and roster/registry activities (discussed in detail later), optimally planned in a generic way prior to an event (termed Baseline Activities); (2) those that investigate and respond to immediate health problems and exposures, and are designed to expeditiously provide useful and actionable information that directly affects the health and safety of current responders (termed Public Health Investigations); (3) those that are exploratory or preliminary in their approach, often to determine the need for or feasibility of a more comprehensive research study (termed Pilot Investigations), and (4) those that entail a systematic and rigorous investigation, typically require peer-reviewed protocols, usually extend well beyond the duration of the emergency, and are designed to develop or contribute to generalizable scientific knowledge (termed Responder Health Research). Public health activities to prevent or control disease or injury and improve health, or to improve a public health program, are not generally considered research, as the purpose of research is to generate or contribute to generalizable knowledge through a systematic and scientifically defensible process.⁵

Baseline Activities include health monitoring, health surveillance, documentation of exposures early in and throughout disaster events, and roster/registry activities. Health monitoring refers to ongoing and systematic clinical evaluation of individual responders' health status, while health surveillance is the term used for an ongoing and systematic population-based data collection approach. Baseline activities may include routine exposure assessment, which refers to systematic industrial hygiene or environmental measurements to evaluate workers' exposures to chemical, biological, or physical hazards. Baseline activities also include compiling a “roster,” which is a list with contact information for all responders scheduled to perform a duty. Alternatively, a “registry” may be implemented to collect, store, and manage information about affected populations at risk of exposure or who have developed a specific disease. The absence of rosters or registries, or sufficient, well-organized exposure data, makes subsequent responder health research studies difficult to conduct, and quite probably less accurate because identifying the universe of response and remediation workers and reconstructing potential exposures after a response has concluded can be very difficult if not impossible. While not usually considered research, per se, baseline activities nonetheless provide critical information that is fundamental to supporting any future investigation or study and should be standard practice in any emergency that entails workers entering uncharacterized or uncontrolled environments, regardless of whether a decision to conduct a responder health research study has been made. Additionally, baseline activities often inform the need for responder research, assist in hypothesis generation, and/or help determine the feasibility of the proposed research.

Public Health Investigations are designed to provide actionable information that will have a direct impact on the health and safety of current responders. Public health investigations involve medical/epidemiologic investigations, industrial hygiene assessments, and/or biological monitoring to assess individual exposure levels and health effects. The focus and timing of these investigations are often to resolve questions arising during the course of the response to help mitigate exposure or disease and identify early indications of significant adverse health effects among responders. Because of the time elements inherent in these studies, they primarily address acute effects, conditions, or concerns and are not designed to detect delayed or chronic effects. Typically, these investigations are not considered research because they are not designed to produce generalizable knowledge, although information derived from the investigations can sometimes inform future responses. As with baseline studies, public health investigations may serve to generate or explore hypotheses and stimulate research ideas. Examples of public health investigations include cross-sectional surveys (“snapshot” surveys of workers that collect information about health issues, risk factors, and exposures at the same point in time), cluster investigations (evaluations of reported conditions and illness above the expected level), exposure assessment investigations, case series (tracking individuals by examining similar exposures and health outcomes), and site-specific investigations, such as NIOSH Health Hazard Evaluations.⁶

Pilot Studies are typically small-scale studies designed to collect information on the feasibility and value of conducting a more comprehensive responder health research study. A pilot investigation also may test, inform or improve the design of an eventual larger research project. Such studies can provide valuable insights; for example, should any research component be identified as missing in the pilot study, it can be added to the larger research study to improve the chances of a clear outcome. Additionally, if insufficient rationale is present to justify a larger research study, pilot research studies can sometimes provide this justification. Pilot studies can be particularly useful in identifying design issues (e.g., a planned procedure or assessment is impractical in the field) before more comprehensive research is conducted. Furthermore, pilot studies may also be considered when data are needed expeditiously because it may not be available later once a larger research study is designed and approved.

Responder Health Research Studies are designed according to a detailed research protocol to test explicit scientific hypotheses about associations between defined exposures (including “psychological exposures”) in the work environment and adverse health outcomes or analogous questions. These studies entail rigorous and/or complex scientific methodology, create generalizable knowledge, typically have peer-reviewed protocols, and usually extend beyond the duration of the emergency. These studies can address acute, sub-acute, and/or chronic effects and conditions. Examples of such studies include (1) cohort studies (designed to follow over time a defined group of individuals who are exposed in varying degrees to a factor that is hypothesized to influence the probability of the occurrence of a given disease or other outcome), and (2) case-control studies (designed to compare differences in exposures and other risk factors among those with an illness/condition (cases) and those without the illness/condition (controls)).

Upon review of the response activities related to the Deepwater Horizon oil release in the Gulf of Mexico in 2010, a need was identified for a systematic decision-making process to determine if post-disaster responder health research should be initiated. In response to this need, the Director of NIOSH convened a work group consisting of senior NIOSH scientists experienced with disaster response and representing the disciplines of epidemiology, occupational medicine and psychiatry, and industrial hygiene, and that included former directors of the NIOSH emergency preparedness and response program, medical staff involved in the World Trade Center Health Program, and epidemiologists involved in industry-wide studies. Drawing on direct experience in multiple large emergency/disaster responses, including the terrorist attacks of September 11, 2001, anthrax events, severe acute respiratory syndrome (SARS), Hurricane Katrina, and the Deepwater Horizon oil release, the workgroup collaborated to devise a comprehensive rationale for determining when to conduct responder health research studies.

An extensive literature search, including MEDLINE,^®EMBASE,^® Cumulative Index to Nursing and Allied Health Literature (CINAHL^®), Health and Safety Science Abstracts, and NIOSHTIC, yielded no publications directly addressing this topic. Search terms including “emergency responder, disaster worker, emergency worker, first responder” and “long-term study, post-disaster study, surveillance, health monitoring” yielded numerous studies that investigated the health status of emergency response workers in various response scenarios; however, none of these studies addressed the thought processes or criteria involved in the decision to conduct such a study, and no general treatments of this issue were identified. The Agency for Toxic Substances and Disease Registry (ATSDR), however, has developed guidance for determining if an environmental assessment should be conducted in a non-emergency situation.⁷ The National Biodefense Science Board recently released a report discussing the need for disaster research and the significant barriers to conducting disaster research, but this report does not specifically address the decision process on whether to proceed with research.⁴

After defining the types of responder health studies and activities pertinent to a disaster response, the work group discussed previous emergency response studies as well as current trends in emergency response research methodologies. While considering various criteria potentially necessary for determining the need for a study, it became clear that developing a focused hypothesis (or hypotheses) based on varying pre-existing information would be a critical component of an effective decision-making process. The work group then discussed factors that should be evaluated to assess whether a study was warranted, including issues such as feasibility and research significance. Using these considerations, the group conducted a table top exercise (described later in the paper) to test the validity of this decision-making process.

This paper summarizes the Work Group's recommendations for determining when a responder health research study should be initiated, including the critical factors that need to be considered when making the decision to carry out research.

Because of the many complexities associated with both the decision to conduct responder research and the actual conduct of responder research, analyzing a prioritized set of potential disaster scenarios could help identify research opportunities. For example, it could be useful to evaluate research questions and knowledge gaps likely to arise and be amenable to study during a variety of incident types, including those foreseen in Federal Emergency Management Agency (FEMA) National Planning Scenarios.¹¹ Careful analysis in the planning stages enables the appropriate application of multidisciplinary expertise to identify scientific gaps, formulate well-defined hypotheses, and design potential studies within the context of the life cycle of a disaster and the requisite phases of emergency management. Planning includes pre-identification of subject matter experts in a variety of scientific disciplines and pre-determination of data needs and systems for obtaining the necessary data (e.g., exposure assessment, worker rosters, etc.).

To facilitate the decision process, the organization or agency contemplating research should, as part of its pre-planning efforts, create a science planning team of experienced subject matter experts from diverse scientific disciplines (e.g., occupational medicine, industrial hygiene, toxicology, epidemiology, psychology, emergency response). The team should be substantially separate from those primarily responsible for coordinating the immediate response. Once a disaster commences, the science planning team should be rapidly convened to review, refine, and prioritize any previously developed generic research plans to identify potential research needs, including a statement of rationale and importance. Scope of the effort (including approximate numbers of staff and potential costs) should also be estimated. This information should be reviewed and a determination made as to whether the research should be pursued. In some cases, the level of research interest may inform determinations. A pre-determined level of research interest may be reflected in an agency or institution's research agenda.¹²

The science planning team, which can be supplemented by specific subject matter experts depending on the type of disaster, should be able to evaluate the information available and come to a consensus on whether there are valid, testable, and relevant hypotheses or appropriate theoretical foundations, and whether a responder research study can feasibly be conducted. The team should also be required to periodically reassess the need for conducting responder research because of the rapidly evolving nature of disasters and worker exposures.

Disaster events, particularly those where responder health research is anticipated, should routinely include collection of a core set of baseline data, such as exposure data, rosters of exposed individuals, and baseline health status. Failure to begin collecting critical data as soon as possible can result in data that are unavailable, lost, or cannot be reconstructed, which can significantly affect study feasibility. Research plans as described above should include provisions for anticipatory data collection, including standard instruments for emergency or generic OMB clearance (in the case of federally sponsored research) and advance preparations for IRB approval, even if it is ultimately decided not to move forward with a research study.

As part of planning efforts for research initiatives, strong interagency communication and cooperation will be needed among federal, state, and local governmental agencies (e.g., health departments, workers compensation bureaus, labor departments, Federal Emergency Management Agency) in the jurisdiction of the disaster. These discussions may be expedited prior to an event through discussions with relevant umbrella organizations, such as the Council of State and Territorial Epidemiologists, Association of State and Territorial Health Officials, and National Association of County and City Health Officials.

Finally, discussions with other public safety, public service, and nongovernmental organizations, including labor unions and educational institutions, are necessary to establish expectations, roles, and responsibilities related to research that may be considered during or following a disaster. In order for research projects—especially those requiring data collection from a large number of people—to go forward efficiently, it is imperative that workers and other participants understand the importance of the work, trust that it is being carried out for their or for society's benefit (and not solely to fulfill institutional needs), and understand that the results will be shared with them when the data are collected and analyzed. Moreover, the affected workers themselves will be an asset in many projects, supplying specific knowledge and skills, including appropriate language and cultural sensitivities, which can strengthen data collection efforts.

The decision process on whether to proceed with a responder health research study can be broken down into various components that should be formally recognized as critical to determining whether a research study of responder health is warranted. As described in this paper, the decision process should be formalized to ensure the best possible judgment and recommendation, and may benefit from advanced planning in anticipation of possible events. Several critical gatekeeper factors are provided that must be satisfied if a research study is to be initiated. A determination to conduct a study without sufficient information to develop the exposure-outcome relationship or without an initial hypothesis is not generally recommended and may be subject to ethical concerns. In the current austere fiscal environment, the recommended framework should ensure that research that is most needed and justified will be identified and prioritized.