Natural and Human-Made Disasters
Every year, approximately 400 natural disasters occur worldwide. Added to these are 30–40 armed conflicts (1). Together, these and other emergencies imperil the health of hundreds of millions of people and substantially increase levels of morbidity and mortality. The future may bring more calamity to more places around the world; climate change is a scientific certainty, and with it comes an increased level of dangerous weather events in all coastal areas around the globe.
Natural events and human-made emergencies (e.g., armed conflict; climate change; and “development disasters,” such as those ensuing from flooding upstream of dam construction or excessive damage from earthquakes where structures have not been built to code) frequently occur in relatively remote, difficult-to-reach locations, often in the poorer countries of the world that are least able to cope. The tasks of field epidemiologists who participate in response efforts include (1) accurately determining the number of people affected, (2) calculating rates of morbidity and mortality, (3) assessing the health-related needs of the population, (4) establishing priorities for providing health services, (5) monitoring progress toward rehabilitation and recovery, (6) evaluating the results of emergency interventions, and (7) improving future responses by communicating the consequences of these emergencies.
The approach to the way supplies and services are delivered to emergency-affected populations has changed radically during the past 50 years. The application of epidemiologic principles to emergency response is generally considered to have begun during the massive international relief effort mounted during the civil war in Nigeria during the late 1960s. During that war, which resulted in widespread starvation, massive internal displacement, and high rates of mortality, epidemiologists developed methods to help determine the health status of the affected populations so that appropriate assistance could be delivered (2). Nutritional surveillance evolved over subsequent years, and, by the late 1970s, internationally approved guidelines for measuring nutritional status had been developed (3).
Toward the end of the 1970s, the genocidal practices of the Khmer Rouge regime in Cambodia resulted in a massive exodus of survivors to Thailand, where hundreds of thousands of people were given refuge in several large camps. These so-called “death camps” quickly became the sites of numerous outbreaks of disease, but the extent and principal causes of morbidity and mortality were measured in quantifiable terms only when epidemiologists from the Center for Disease Control (later Centers for Disease Control and Prevention), working together with colleagues from the International Committee of the Red Cross and a group of nongovernmental organizations (NGOs), instituted a formal disease surveillance system and conducted methodologically sound surveys (4).
Before the regular use of field epidemiology techniques, emergency response was guided mainly by the best intentions of relatively inexperienced medical and surgical teams with inappropriate skills and inadequate logistical support. Doctors would build makeshift clinics, throw open the doors, and provide services to people who were able to access them—in most instances, only a small proportion of the affected population. Available services frequently did not match the public health needs of the population. Planners and managers were in the unenviable position of directing major relief operations with little information to guide their efforts (5). However, as sound epidemiologic practices emerged and were more regularly applied, reasonably accurate denominators on which to calculate rates of illness and death were generated and a more disciplined approach to the delivery of humanitarian assistance in the health sector evolved.
Unfortunately, disasters that have needed more honed epidemiologic approaches have continued to occur regularly. Examples include repeated famines and conflicts (the two are not unrelated) in the Horn of Africa; cyclones and tsunamis leading to massive flooding in countries bordering the Bay of Bengal and elsewhere in the Indian Ocean; earthquakes and hurricanes in the Caribbean and Central America; and wars in the Balkans, the Middle East, and Central Africa. All of these required distinct responses, but eventually, because of the development and application of epidemiologic techniques, including more formal approaches to rapid assessment, surveillance, and impact evaluation, patterns of morbidity and mortality emerged. In addition, training programs were established that resulted in an emergency response workforce that was more knowledgeable, more sophisticated, and more capable of reducing illness and saving more lives in less time (Box 22.1) (6).
One notable watershed occurred in the wake of the Rwanda genocide of 1994, when more than 500,000 refugees fled that country to then-Zaire, with many settling in a few camps near the northern tip of Lake Kivu. Within weeks, an estimated 45,000 refugees had died of cholera, despite the presence of hundreds of nongovernmental organizations, United Nations agencies, military medical contingents from at least nine Western countries, and many other public health officials (7). The collective failure to respond effectively to this situation clearly underscored the need for the emergency relief community to develop indicators for a successful intervention and to work to achieve those indicators in every emergency. This need led to development of the Sphere Project and its accompanying Handbook (Humanitarian Charter and Minimum Standards in Humanitarian Response) that remains obligatory reading for persons working in this field (8). In addition to establishing standards in key areas (shelter, food security, food aid and nutrition, water and sanitation, and health services, and the cross-cutting areas of gender and protection), the Sphere Project has provided opportunities for epidemiologists and other public health experts to agree on a relatively standardized approach to emergency relief. A fourth edition of this essential Handbook will be published in Fall, 2018.
More recently, notable humanitarian crises resulting from natural disasters have included a massive earthquake in Haiti (2010); flooding that displaced 20 million people in Pakistan (2010); several typhoons in the Philippines, including Typhoon Haiyan/Yolanda in 2013; and the ongoing (2017) severe drought in the Horn of Africa. Human-made emergencies commanding the attention of the international humanitarian community have included ongoing conflicts in South Sudan, Central African Republic, and throughout the Middle East. Although the peer-reviewed literature addressing responses to such disasters remains relatively sparse, field epidemiologists preparing to respond to future crises should be encouraged to learn from these case studies.
The principal objectives of epidemiologic field investigations and response in emergency settings are to
- Establish the magnitude and distribution of the public health consequences of the event.
- Assess the size and health needs of the affected population.
- Help provide and promote epidemiologically derived data as the principal basis for resource allocation.
- Help guide implementation of public health programs to minimize postemergency morbidity and mortality.
- Monitor progress of the relief effort.
- Evaluate the effectiveness of the relief effort.
The field epidemiologist is a core member of the emergency response team. Increasingly, the international response to emergencies is organized in a command-and-control manner, in accordance with the Incident Command System (see Chapter 16) or similar systems approaches (9). Knowledge of the organizational structure of the relief effort and identification of the decision-makers is important, as are being a team player and understanding the roles of other team members. In the face of tragedy, many unseasoned hands will adopt an “act first– think later” approach and view the methodical collection and analysis of data as a frivolous, time-wasting activity. In these instances, the field epidemiologist must be an affirmative voice of reason—strongly advancing an evidence-based approach to health interventions that maximizes benefit to the affected population.
Although no cookbook approach exists to emergency response, flexibility and sound judgment are hallmarks for the successful use of field epidemiology. Accordingly, a flexible framework of steps for the epidemiologist includes
- Determining the impact of the event on the public’s health by establishing rates of illness and death with an optimal attainable level of accuracy (note: “the perfect should not be the enemy of the good”). In doing so, it is, of course, essential to focus on the determinations of both numerators (cases and deaths) and denominators (total population and, wherever possible, age and sex breakdowns).
- Initiating disease surveillance as quickly as possible, beginning with a minimum amount of data to collect and augmenting as deemed appropriate and feasible.
- Identifying personal, household, and environmental risk factors for elevated rates of illness and death.
- Advocating for the early initiation of essential public health interventions and disease-control programs on the basis of knowledge of the actual and potential distribution of diseases in the population.
- Arguing forcefully that health actions of lesser priority be deferred.
- Becoming an essential member of the health response team by attending appropriate meetings; working with public health officials and other responders from different organizations, including government officials; and providing frequently updated reports about the situation to those who have a need to know.
This last point (i.e., providing situation reports) is critical; in emergency response, “consequential epidemiology” needs to be practiced (10). The contribution of epidemiologists reflects their ability to provide timely and accurate data in a way that decision-makers can easily understand, analyze, and use for action. The use of those data should enable effective implementation of appropriate public health measures. Conversely, collecting and providing potentially useful information that decision-makers do not act on might be viewed, in part, as a failure of field epidemiology, as is the implementation of health interventions that relevant data do not support. Thus, epidemiologic skills are necessary but not sufficient: equally critical are the abilities to communicate effectively, advocate successfully, and provide strong leadership in support of the policymakers directly responsible for consequential actions.
In its early stages, the emergency relief environment is always chaotic. However, every responder has the same essential needs: food, water, shelter, transportation, communication, and a place to sleep. Thus, the field epidemiologist’s first priority is to arrange to meet these basic needs. This is important because the more independent one can be, the less others will have to divert attention from their work to provide assistance.
Hiring staff is another early priority, especially in international emergency relief. Because field epidemiology is a population-based discipline, the epidemiology team should include members who know the local language, geography, and customs. Therefore, recruiting and retaining people who can be relied on to be effective liaisons with the local communities is a high priority. Although English-speaking translators are highly valued, because they do not always represent the community and are unlikely to be professionally trained, information they provide should be carefully assessed and verified.
Establishing Rates of Illness, Injury, and Death
In most emergency relief settings, accurate measurement of the size of the affected population and its current health status is missing and difficult to establish. For the field epidemiologist, though, it is critical to determine a reasonably precise denominator on which to base the calculation of rates, such as crude, age-, sex-, and disease-specific death; prevalence of moderate, severe, and global acute malnutrition in the affected community; incidence of high-priority conditions; and access to use of health services. Determining rates is essential for comparing population groups and prioritizing public health interventions. A variety of methodologic options can be used to calculate population size, ranging from the more basic, such as extrapolating from the number of people in a sample of dwelling units, to the more sophisticated, such as using aerial photography and/or satellite imagery. The field epidemiologist needs to consider the context in which the relief effort is occurring to select the best method—one that provides reasonably accurate numbers in a culturally and contextually sensitive way.
In 1980, in one of the many emergencies on the Horn of Africa, women were observed to be wearing no jewelry, a sign that all valuables had been sold to purchase food that had become available at exorbitant prices. There was one exception, however: almost all women wore a thin string around their necks with a small, spoon-shaped pendant attached to it. The significance of this oddity eluded field epidemiologists assessing the health status of the population until a visiting ophthalmologist mentioned that this population suffered from an unusually high prevalence of trachoma. The spoon-shaped device, it was learned, was used to remove inverted eyelashes, an action that helped relieve the irritation and pain associated with the scratched and ulcerated cornea that are a feature of this disease. Their ubiquity was a testament to the importance of the disease—and keen observation was the key to diagnosing this public health problem.
Source: R. Waldman, unpublished data.
Field epidemiologists play a key role in the earliest stages of any relief effort. In addition to an appreciation for quantifiable data and for how and when to collect it, the “shoe leather” component of epidemiology is valuable in and of itself for conducting an initial rapid assessment. A wealth of information can be gleaned from observation during a walk-through of the affected area if one knows what to look for and how to employ basic qualitative techniques.
- If commodities are being sold or traded in the marketplace, then their price, compared with preemergency prices, indicates their availability or scarcity.
- Black markets spring up quickly in postdisaster settings, and the willingness of people to make major sacrifices to pay for essential commodities indicates dire need.
- Indicators such as the amount of and type of jewelry being worn can be meaningful (Box 22.2). The absence of traditional adornment in a society in which it is customary might signify food insecurity and that everything of value already has been sold.
Interviews with community leaders, transect walks through affected areas, and results from a constellation of methods that frequently are grouped as participatory rapid appraisals can be useful even before the analysis of survey data that might provide more accurate information but at the cost of timeliness. Of paramount importance for the field epidemiologist is reaching the disaster location as quickly as possible, visiting all affected areas and population groups, and helping the relief community gather, collate, and assess the value of all information. Postemergency settings are dynamic, but ultimately decisions about public health and health service delivery must be made from day 1 on the basis of existing evidence (11).
As valuable as nonquantitative data might be, the lack of routinely collected health information means that, as soon as is feasible, surveys will need to be conducted. A precise sampling frame will be difficult to establish at first, and careful judgment is needed to ensure that samples drawn from the population are representative. However, in most circumstances, a less than optimally representative systematically chosen sample will be superior to a convenience sample, especially if the results are to guide the equitable distribution of commodities and services.
A commonly used survey method is two-stage cluster sampling, first developed by the World Health Organization to measure vaccination coverage rates (12). The logistical demands of this method are far less than for either simple random sampling or systematic random sampling because relatively few clusters need to be visited to obtain statistically valid results with a reasonable degree of precision. Although sample sizes can be relatively large, the advantages of using this method usually outweigh the disadvantages. Nonetheless, two distinct disadvantages should be noted:
- Cluster sampling is not well suited for measuring characteristics that are not homogenously distributed in the population. For example, if malnutrition is clumped in certain areas, then cluster sampling might miss it entirely or, conversely, overidentify it, resulting in skewed, nonrepresentative values for the population as a whole.
- Cluster sampling can be difficult to explain to decision-makers.
Finally, a frequently overlooked problem with surveys is that nonsampling error is likely to be more important than the disadvantages of any sampling method. Surveyors need to be carefully trained to understand the objectives of the survey and the importance of collecting accurate and unbiased information. When people affected by an emergency have lost their possessions or suffered other shocks, they can be eager to please those they perceive to be in a position to help them by providing answers they think the surveyors want to hear, resulting in a sincere, but inaccurate, picture of reality. For example, people might not report household deaths because they fear having their rations decreased. Therefore, the field epidemiologist needs to be aware of the many real and potential biases in obtaining accurate information from an emergency-affected population and must take steps to ensure that none of the epidemiologic activities inadvertently contributes to further deterioration of the situation. For epidemiologists, as for clinicians, “do no harm” is an important rule.
Organizing Priority Interventions
The main goals of emergency relief are to save lives and restore individuals and communities to their preemergency conditions. Although individual-and population-directed health interventions are important in many settings, other types of interventions might take precedence. In the book, Refugee Health, the medical relief organization Doctors Without Borders suggested 10 top priorities in disaster response (13). Of the top five, only one—measles vaccination—is a health-specific intervention, and its importance might have diminished since publication of that book as more countries have achieved high measles vaccine coverage rates through routine health services. (In situations of protracted conflict, however, where primary healthcare services have been unavailable to the population for some time, vaccination coverage levels can fall dramatically. As a result, measles outbreaks have occurred increasingly throughout the Middle East and in migrant populations in Europe.) The other priorities are initial assessment; water and sanitation; food and nutrition; and shelter and site planning. Although these are clearly related to public health, in most international emergency responses they are considered to be distinct from the health sector.
Some humanitarian interventions address basic needs of the emergency-affected population slowly and even inadequately. For example, in the area of nutrition, field epidemiologists have been called on to identify, diagnose, and design appropriate interventions for rare conditions (e.g., scurvy, pellagra, and beriberi) while simultaneously implementing surveillance for acute moderate and severe malnutrition. Although relief team members who are experts on specific problems understandably will focus on those problems, the field epidemiologist needs to address the overall spectrum of the relief effort and promote the most appropriate interventions, regardless of the sectors to which the interventions might belong.
Public health surveillance is a critical element of disaster response, and its establishment usually becomes the responsibility of the on-site epidemiology team. In humanitarian settings, epidemiologists attempting to implement effective surveillance might have to address several challenges, including
- Balancing speed and accuracy in adverse conditions.
- Integrating multiple sources of sometimes conflicting data while determining which are credible and which are not.
- Soliciting others to participate in the surveillance effort when they might not assign it the same priority the epidemiologist does.
- Assisting decision-makers in using surveillance data to take action.
Rapidly established, well-monitored, and widely used surveillance systems have been instrumental in preventing deaths as, for example, in the aftermath of the Asian tsunami of December 1994, when on-scene, experienced epidemiologists helped conduct effective surveillance.
Conditions targeted for surveillance vary in relation to specifics of the setting. In most developing countries, at the start it may be sufficient to target a simple surveillance system toward syndromic presentations and easily recognizable conditions, such as acute lower respiratory illness (a proxy for pneumonia), acute watery or bloody diarrhea (cholera, dysentery), fever with or without stiff neck (malaria, meningitis), and measles. In other settings—especially in middle-and higher income countries—the focus might be on measuring the needs of chronically ill persons who might be cut off from their medications or procedures; in these situations, such conditions might be more prevalent than common acute communicable diseases. In all settings, surveillance should focus on the most vulnerable segments of the population (e.g., infants, children, older persons, women, destitute and underserved persons, and persons with special needs). To ensure they are not neglected, epidemiologists should disaggregate data to facilitate identification of health problems in these groups.
Emergency relief almost always occurs in emotionally charged environments. Although the need for highly coordinated action is universally recognized (some have suggested that “poor coordination” should be recorded as a cause of death on death certificates), many responders might want to coordinate but not “be coordinated.” The most common scenario is for a health cluster to be established at the onset of the relief effort. Government officials, representatives of the World Health Organization, and a designated person from a nongovernment organization usually are assigned joint responsibility for chairing cluster meetings and overseeing their functioning. In large disasters, such as the Haiti earthquake of 2010, several hundred responders regularly attended health cluster meetings, many seeking guidance on how to respond effectively (14).
The epidemiologist, for better or for worse, frequently is thrust into a position of responsibility and authority because most responders will not be familiar with the published medical and/or public health literature and few will be able to view the chaos through the objective lens of unbiased data. Epidemiologists responding to an emergency for the first time might be unfamiliar and even uncomfortable with the amount of respect they are accorded.
Humanitarian response settings are the emergency rooms of public health. Lifesaving, irreversible decisions frequently are made in the early phases of the relief effort. A fundamental task of the field epidemiologist is collection and circulation of essential data on the health and nutritional status of the affected population as accurately as possible in the shortest possible time. The purpose of these data is to help first responders prioritize the interventions most likely to limit excess preventable death. The environment is often chaotic, uncoordinated, and characterized by logistical and resource constraints, but the epidemiologist needs to be calm, assertive, and able to convey the power of accurately collected and analyzed data. Ultimately, however, successful contribution to a disaster response will be measured not on the basis of the elegance of the epidemiologic investigations, but rather as a function of how many lives are saved (15).
- Centre for Research on the Epidemiology of Disasters. http://www.cred.be
- Brown RE, Mayer J. Famine and disease in Biafra: an assessment. Trop Geogr Med. 1969;21:348–52.
- De Ville de Goyet C, Seaman J, Geijer U. The Management of Nutritional Emergencies in Large Populations. Geneva: World Health Organization; 1978.
- Glass RI, Cates W Jr, Nieburg P, et al. Rapid assessment of health status and preventive-medicine needs of newly arrived Kampuchean refuges, Sakeo, Thailand. Lancet. 1980;1:868–72.
- Sommer A, Mosley WH. East Bengal cyclone of November, 1970. Epidemiological approach to disaster assessment. Lancet. 1972;1:1029–36.
- Toole MJ, Waldman RJ. Prevention of excess mortality in refugee and displaced populations in developing countries. JAMA. 1990;263:3296–302.
- Goma Epidemiology Group. What happened in Goma, Zaire, in July 1994? Lancet. 1995;345:339–44.
- The Sphere Project. Humanitarian Charter and Minimum Standards in Disaster Response. 3rd ed. Geneva: The Sphere Project; 2011.
- Southeast Alaska Petroleum Response Organization. What is the incident command system? http://www.seapro.org/pdf_ docs/ICS.Overview.pdf
- Field Epidemiology Manual Wiki. Field epidemiology manual. https://wiki.ecdc.europa.eu/fem/w/wiki/field-epidemiology
- Checchi F, Warsame A, Treacy-Wong V et al. Public health information in crisis-affected populations: a review of methods and their use for advocacy and action. Lancet. 2017;390:2297–313.
- Henderson RH, Sundaresan T. Cluster sampling to assess immunization coverage: a review of experience with a simplified sampling method. Bull World Health Org. 1982;60:253–60.
- Médecins Sans Frontières. Refugee Health—An Approach to Emergency Situations. London: Macmillan; 1997.
- World Health Organization. Global health cluster guide. http://www.who.int/hac/global_health_cluster/guide_glossary_of_key_terms/en/
- Waldman RJ, Toole MJ. Where is the science in humanitarian health? Lancet. 2017;390:2224–6.