Challenges

Health-related disaster research poses challenges ranging from study design problems to logistical issues. Methodologically, there is an inherent problem of a small number of cases, and rarely an opportunity to standardize data collection efforts to allow for a comparative case approach, much less meta-analyses. Inasmuch as prospective research designs are difficult to develop, the field must generally rely on observational and quasiexperimental research designs. This presents its own challenge: If each disaster case is “exceptional” in its own way, as a complex interaction of hazard, geography, response system characteristics, and vulnerable populations, then it is difficult to identify meaningfully consistent variables from which to generalize and theorize.

There is also a considerable challenge in the field to develop methods for testing the empirical basis of disaster practices, such as laboratory simulations, computer modeling, and operational research, which could then be incorporated into training and educational curricula.

A number of other “process” issues confront public health disaster research. Foremost is the need for rapid funding mechanisms and the training and deployment of research field teams, particularly if there is an interest in obtaining data during the response or short-term recovery phases. Community-based disaster research is often labor intensive. Furthermore, there are issues of institutional review board processes, including both time factors and the tendency of many institutional review boards to regard all disaster victims as a vulnerable population for whom any research may be viewed as a “risky and burdensome enterprise” requiring special attention and consent.Reference Collogan, Tuma, Dolan-Sewell, Borja and Fleischman16 Of course, these issues have less salience if the research is conducted during a predisaster period or considerably after the fact.

Untilled Research Fields

The fields of disaster medicine and public health preparedness may be defined broadly or narrowly circumscribed. The widest landscape would encompass research that examines the social forces that predispose some people and groups to be more acutely vulnerable to specific hazards, and would also consider the policies that could prevent the hazards from occurring or that could mitigate the consequences of any specific threat. Furthermore, the consequences of a disaster do not end once the environment has been stabilized and all critical injuries have been attended to. A community's recovery has significant health and medical components, in terms of the chronic health conditions of the population, the availability and quality of health services, the capacity to promote healthful activities and sustain preventive behaviors, and the overall ability of the affected community to reconstitute its health system and allied services.

A narrower definition of the field, however, would focus on preparing health sectors for various disasters, identifying event-specific screening and treatment modalities, and designing systems of care for responding to and treating the affected populations. Using this calculus, prevention and mitigation are generally beyond the scope of disaster medicine and public health preparedness, as are issues of long-term recovery.

Although it is critical that public health and medical systems maximize their ability to prepare and respond to disasters, we believe that the broader view is essential to developing knowledge and interventions that lead to enhanced population resiliency and survival. Research should be considered within an ecological framework that incorporates multiple levels, from the individual through the organizational, social, and political realms, an agenda similar to that proposed by the National Institutes of Health in its report on integrating social and cultural dimensions into multilevel analyses of health.17 Figure 1 illustrates some of the research areas in a matrix comparing units of analysis (individual, organizational, social, political, legal) for each hazard phase. This listing is not intended to be comprehensive, but rather to stimulate thought as to what research questions could be posed within each cell. For example, within the prevention and mitigation phase, in which the potential consequences of any disaster could either be averted or greatly buffered, a research team may consider whether enhancing self-sufficiency among vulnerable individuals would lead to greater resiliency, emergent leadership, and quicker recovery in the face of a disaster. A political analysis could explore the ethics of mass vaccination in the context of a high-risk/low-probability event.

FIGURE 1

Untilled or undertilled research areas, by hazard phase and analytic unit

Within the preparedness phase, there is a considerable need for outcomes research. Although federal campaigns exhort individuals to “Get a kit, make a plan, and be informed,”18 there is little empirical evidence to suggest which elements, if any, of the kit, plan, or information are associated with improved outcomes or a reduced burden on the response system.

Within the response phase there is an emerging literature about nonpharmaceutical interventions such as school closures and social distancing in response to a pandemic influenza outbreak, but less emphasis on the unanticipated consequences of such policies. Long-term school closures could lead to major workforce shortages, nutritional concerns if students are unable to access free breakfast and lunch programs, and an increase in youth risk behaviors if many are unsupervised. For legal and ethical frameworks during the response phase, some compelling topics include appropriate modifications of standards of care, resource allocation principles, revised triage protocols, and cross-border access and treatment concerns.

The recovery phase after a disaster is composed of short-term and long-term recovery components, which lack a clear demarcation between them. From a population perspective, there may be a defined role for ancillary services such as case management to support medical home models similar to those that have evolved in domestic HIV/AIDS programs. Furthermore, from a political and legal standpoint, a valuable research initiative would be an examination of disaster relief policies (eg, disaster-relief Medicaid, Robert T. Stafford Disaster Relief and Emergency Assistance Act provisions, HIPAA waivers) and their contribution to enhancing medical care access and health outcomes.

Although the progression of stages illustrated in Figure 1 is linear, one can easily imagine these stages as a circular process. Long-term recovery activities may lead to new prevention and mitigation efforts. Policies clearly affect individuals and organizations. It would be useful to apply multilevel modeling techniques, such as hierarchical linear regression models, that could accommodate these multiple domains and effects.

Fundamental to all of these potential research topics is the questioning of assumptions, a critical reluctance to accept the received wisdom. Is the way in which we have organized our response systems to respond to emergencies, using such hierarchical structures as the Incident Command System, the most adept at dealing with complex emergencies? Russell Dynes, the founding director of the Disaster Research Center, has long noted the differences between “military model” response structures, such as that embodied by the Incident Command System and the National Incident Management System, and more nimble “problem-solving” models, which are closer to collaborative public health strategies.Reference Dynes and Tierney19 Louise Comfort, a veteran disaster researcher, has noted that “building networks of organizations committed to a process of continual inquiry, informed action, and adaptive learning is a more flexible, robust strategy than the standard practice of establishing greater control over possible threats through administrative structures.”Reference Comfort20 It is likely that a blending of the models may be most effective. Research, if done properly, could inform practice.

One recent advance has been the promotion of an Utstein Framework for evaluation and research of disaster-related health events,Reference Sundnes21 modeled on cardiopulmonary resuscitation research. This field, like disaster research today, had been plagued by a lack of standardized terminology.Reference Cummins, Chamberlain and Hazinski22 Much as an epidemiological case definition standardizes the clinical identification of an infectious disease, the Utstein Framework standardizes the case description of a disaster, allowing researchers to devise various methods of comparative case analyses, including post hoc regression models, or to develop meta-analyses. Even more innovative techniques could include a variation of Ragin's comparative case methodology, which uses Boolean algebraic truth tables to permit case comparisons rather than variable-level comparisons conventionally used in multivariate analyses.Reference Ragin23, Reference Ragin24 Depending on the particular research question there are also opportunities to capitalize on larger numbers of cases by focusing on health systems, public health departments, or governmental jurisdictions such as counties at which much emergency management is organized. Prospective studies could then use research strategies such as randomized cluster designs, whereas observational and retrospective studies could benefit from a number of sampling strategies.

Qualitative studies, in general, can offer much to the nascent research field. Oral histories and detailed ethnographic and participant-observation studies can provide rich detail on social and organizational processes for which there are no quantitative datasets. Qualitative research can inform theory generation; prove (or disprove) quantitatively deduced hypotheses; and engage the research, practice, and policy worlds through the strength of its narrative. Case study methods, for example, can be as rigorous as the most complex quantitative design. Reference Yin25–Reference Hamel, Dufour and Fortin27

Necessary Partnership of Medicine and Public Health

In some ways disaster medicine and public health preparedness in the 21st century mirrors that of the late 19th century, when physicians and public health officials found themselves constant allies rather than occasional colleagues.Reference Lasker28, Reference Starr29 The factors that made 19th century populations particularly vulnerable to infectious agents were inherently social: overcrowding, economic disparities, poor housing, poor nutrition, and inadequate safety regarding water and sewage systems. Similarly, many disaster threats faced today, whether caused by anthropogenic or natural hazards, have a social rather than clinical etiology, broadly encompassing population growth, economic imbalances, cultural tribalism, emerging infectious agents, or chemical or nuclear agents.Reference Arnold30 Many of the consequences of the devastation wrought by Hurricane Katrina—failed sewage pumps, rotting food stocks, and the need for mass vaccinations, mortuary services, primary medical care for special needs populations, and even durable medical equipment for chronically ill populations—fell to public health. The private medical community was the first among many partners that public health had to recruit for its response and recovery missions.

In testimony before the US Congress in 1996 former Federal Emergency Management Agency administrator James Lee Witt noted that “All disasters are political.” Although the direct provision of medical care during a disaster may not seem like a political act, the environment that precedes and succeeds that clinical encounter is informed by politics and policies. Policymakers shape the landscape that will influence the nature of mitigation, preparedness, response, and recovery for any domestic population or hazard event. If public health disaster research intends to be translational, then its scope should span the multiple layers of ecological units and hazard phases, and its primary audiences must include not only providers, planners, and researchers but also policymakers.

Study Limitations

The literature review conducted for this study captured a cross-section of relevant research articles in the medical and public health journals. In an effort to focus on the core public health disaster research literature, the authors elected to purposefully limit the search by language, time frame, and journal databases. Future studies should expand such reviews to non-English language journals to include the larger international literature. Also, a broader key word search within a longer time period will likely result in sufficient numbers to allow for trend analyses, particularly among categorical subgroups. The search would also be enriched by including other literatures, such as those of the social sciences and public administration. An analysis of citations could suggest the level of cross-disciplinary interest in the field.

The medical and public health journals from which this literature survey were drawn represent the most common sources of research and data for health-oriented practitioners and researchers in the field. The disaster literature has been scattered over journals in a variety of fields and disciplines, but are often not used by researchers in disaster medicine and public health because of the difficulty of identifying them. Until recently, there have been few journals covering the full spectrum of disaster research. This fragmentation of knowledge, and difficulties in integrating and applying results and methodologies from other disciplines has been a serious limitation in the development of the field of public health disaster research.