Characteristics of Global Travel

Global travel has increased as reflected in Figure 2-1, showing numbers of international tourist arrivals from 1950 though 2005 and the projections until 2020. In addition to the marked increase in the overall number, there has also been a shift in areas visited by travelers, especially to areas in Asia. The 2006 figures from the World Tourism Organization showed the most rapid relative increase was to sub-Saharan Africa (UNWTO, 2008b). Travel between regions was increasing faster than travel within regions, and air transport was growing at a faster pace than ground and water transport. Figure 2-2 shows the breakdown by means of transport, with air travel accounting for 46 percent of transport (UNWTO, 2008). The shifts in destination mean more people will be traveling to low-latitude countries—areas with greater species richness (Guernier et al., 2004) and often characterized by poor sanitation and limited infrastructure, a milieu where risk for exposure to common and previously unidentified microbes may be higher. In looking ahead, it is unclear to what extent the current dramatic changes in the global economy will affect numbers of travelers or favored destinations. Political instability and disease outbreaks can also influence travel destinations, sometimes abruptly.

FIGURE 2-1

International tourist arrivals by region (in millions), 1950–2020. SOURCE: Reprinted with permission from UNWTO (2008a).

FIGURE 2-2

Inbound tourism by means of transport. SOURCE: Reprinted with permission from UNWTO (2008a).

A vivid example of the rapid increase in travel is the outline of lifetime tracks by David Bradley (1989; figure also reproduced in Cliff and Haggett, 2004), who recorded the life travel over four male generations in his own family (Figure 2-3). The linear scale for the spatial movement increases by a factor of 10 for each generation (Cliff and Haggett, 2004). Similar findings were noted in a study which showed that spatial mobility, taking into account all forms of transportation, of the French population between 1800 and 2000 increased 1,000-fold (Grubler and Nakicenovic, 1991). Using numbers from the U.S. Census Bureau and the United Nations World Tourism Organization (UNWTO, 2008c), one can calculate that the world population between 1950 and 2007 increased 2.6-fold, whereas the international tourist arrivals increased 35-fold. Although individuals on average are traveling much greater distances, González and colleagues found that individual human trajectories showed high degrees of temporal and spatial regularity, with individuals returning to a few highly frequented locations (González et al., 2008).

FIGURE 2-3

Life travel over four male generations in the same family. SOURCE: Adapted from Bradley (1989) and reprinted from Cliff and Haggett (2004) with permission from Oxford University Press.

Population Size, Density, Location, and Proximity to Animals

This increase in human movement is not occurring as an isolated event. Today, the global human population is the largest ever recorded—and so is the population of food animals. About half of the people on Earth live in urban areas, the largest fraction ever (Wilcox et al., 2008). These human hosts provide expanded opportunities for viral replication and mutation events. Most of the population and projected growth are in low-latitude urban areas—regions that are home to sprawling megacities, many surrounded by vast slum areas that lack clean water and sanitary facilities (Figure 2-4). Animals such as dogs, chickens, cows, and rats live in and near human living quarters, which have been assembled from whatever materials can be found. Individuals who inhabit these areas often work in major metropolitan areas, so they have regular contact with large, dense human populations in high-rise buildings and other built environments. Residents of periurban slums may also visit family in rural areas, thus potentially providing a link from rural to urban populations and to the rest of the world. Changes in the environment, including extreme weather events, can favor the appearance of some infections and can also displace populations (IOM, 2008).

FIGURE 2-4

Juxtaposition of urban slums and modern buildings in São Paulo, Brazil. SOURCE: Image courtesy of Zema Fontoura.

Animals have been the origin of many of the recently identified emerging infectious diseases, including HIV/AIDS, H5N1 avian influenza, severe acute respiratory syndrome (SARS) (Jones et al., 2008; Wang et al., 2008), and swine-origin H1N1 influenza A (Novel Swine-Origin Influenza A (H1N1) Virus Investigation Team, 2009; Smith et al., 2009a). Populations of animals have expanded rapidly, in large part to accommodate the desire for more animal protein in the diet, which has coincided with the economic resources to buy it. In China, for example, at the time of the 1968 influenza pandemic, the size of the human population was 790 million, the pig population 5.2 million, and the poultry population 12.3 million. By 2005, while the human population in China increased less than two-fold, the pig population increased about 100-fold to 503 million and the poultry population increased 1,000-fold to 13 billion (Osterholm, 2005). Concentrated animal feeding operations, where large numbers of genetically similar animals are raised in concentrated areas—so-called factory farms—are becoming increasingly common in the United States and other countries (Pew Commission, 2008). Unfortunately, little systematic surveillance of influenza and other potential pathogens in swine populations is routinely done, a shortcoming highlighted by the emergence and spread of a reassortant influenza in 2009 (Smith et al., 2009a).

Roles of the Traveler

Human travelers can easily carry person-to-person transmitted infections to any part of the world. An example is the human immunodeficiency virus (HIV), which was introduced to all areas of the world almost exclusively by travelers (Perrin et al., 2003). Recently, swine-origin H1N1 has spread globally, its movement hastened by global air travel. Although drug-resistant forms of tuberculosis can emerge in settings with inadequate and inappropriate treatment regimens, humans also transport and transmit tuberculosis, including multidrug resistant (MDR) and now extensively resistant (XDR) forms of tuberculosis, in geographic areas far from the point of acquisition (Jassal and Bishai, 2009; Oeltmann et al., 2008). Although tuberculosis is an old disease and is present worldwide, the incidence of infection and the levels of drug resistance vary enormously by population and geographic area. In 2008, the overall incidence of tuberculosis in the United States was 4.2 cases per 100,000 population (CDC, 2009). The rate in persons who were foreign born was 10 times higher than the rate in persons who were U.S. born, reflecting the vast differences in rates of tuberculosis around the world. A number of countries have annual incidence rates exceeding 300 per 100,000. Figure 2-5 displays the geographic spread of XDR tuberculosis between December 2006 and June 2008.

FIGURE 2-5

Countries with XDR tuberculosis cases in December 2006 and June 2008. SOURCE: Reprinted from Lancet Infectious Diseases,Jassal and Bishai (2009), with permission from Elsevier. Copyright 2009.

The traveler can serve in many roles (Wilson, 2003). Nonimmune travelers are at risk for a number of infections that exist primarily in tropical areas. Vaccines and drugs are available today to reduce the risk for many of these infections, such as yellow fever and malaria. Travelers were important historically in the spread of infections and remain so today, perhaps to an even greater extent (Colizza et al., 2006). They can also carry microbes with resistance genes, even if they are unaware of it. Travelers today continue to spark outbreaks of measles in populations that do not have high levels of immunity. It is easy to see how travelers could play a key role in the global epidemiology of infections that are transmitted from person to person, such as HIV, SARS, tuberculosis, influenza, and measles (Hufnagel et al., 2004), but they are also important in the spread of some vector-borne infections, as will be discussed below.

Receptivity to Introductions

Geographic areas and populations vary in their receptivity to introductions of potential pathogens that can cause human disease. Multiple factors are in play. The physicochemical environment may preclude the presence of a necessary mosquito vector or essential intermediate host. The physical environment may also influence transmission dynamics. For example, influenza has a strong seasonal pattern, especially in temperate regions. This seasonality is influenced by the humidity; recent studies suggest that absolute humidity is a more useful measure than relative humidity. The absolute humidity affects influenza virus transmission and virus survival. Absolute humidity can explain 90 percent of the variability of influenza virus survival, whereas relative humidity can explain only 36 percent of variation (Shaman and Kohn, 2009). Hence, travelers with influenza returning to temperate areas during hot, humid months are unlikely to spark epidemic spread (Lowen et al., 2008), though focal outbreaks in contained, air-conditioned spaces (e.g., air-conditioned nursing homes and barges) have been reported during hot, humid weather. Influenza outbreaks in the Southern Hemisphere occur during hot-weather months in the Northern Hemisphere; in the tropics influenza can circulate throughout the year. Analysis of H3N2 epidemics worldwide between 2002 and 2007, including those in temperate regions, suggested that they were seeded annually by viruses that had first appeared in East and Southeast Asia (Russell et al., 2008). These viral strains appeared to have evolved from other Asian strains.

Other local factors that affect the receptivity of individuals and a population to introduction of a new infection include housing, sanitation, and living conditions. Good nutrition can reduce the vulnerability to some infections or diminish their severity. Populations may be immune because of vaccination or prior infection. Human behavior and activities influence exposure to a number of infections (e.g., sexually transmitted infections). And finally, good surveillance and wider access to good medical care may reduce the burden of an infection in a population and allow it to be brought under control.

In addressing the question of how controllable an infection is that is directly transmitted from person to person, a key factor is the proportion of transmission that occurs before onset of symptoms or during asymptomatic infection (Fraser et al., 2004). Public health measures are most likely to be effective when little or no infection is transmitted during asymptomatic infection. Figure 2-6 displays four infections that are transmitted from person to person: SARS, smallpox, influenza, and HIV. Fortunately, the fever caused by infection with the SARS coronavirus preceded onset of transmissibility, meaning that with strict surveillance for symptoms and isolation of those with symptoms, it became possible to interrupt the transmission of this infection. Influenza, on the other hand, is more difficult to contain because transmission may begin before onset of symptoms, and infected patients may have little or no fever. Based on epidemiological analyses, Fraser et al. (2009) estimated the basic reproductive number of the swine-origin H1N1 influenza A virus in the range of 1.4 to 1.6.

FIGURE 2-6

How controllable is an infection? Plausible ranges for the key parameters R₀ and θ for four viral infections of public concern are shown as shaded regions. The size of the shaded area reflects the uncertainties in the parameter estimates. SOURCE: (more...)

Vector-borne Infections

Vector-borne infections can also be introduced into new geographic areas by travelers, though the vulnerable areas are restricted to those with competent mosquito or other arthropod vectors. Because important vectors, such as Aedes aegypti and Aedes albopictus, can be dispersed by trade (especially via ships), more human-inhabited areas of the world are infested with these potential vectors than ever before (Tatem et al., 2006). Aedes aegypti thrives in an urban environment, and today about 2.5 to 3 billion people live in tropical and subtropical areas infested with this mosquito.

Two vector-borne infections, dengue fever and chikungunya infection, which have expanded in distribution in recent years, illustrate multiple contributions to this dynamic process and spread. Dengue virus is causing more infections, including more cases of severe and complicated dengue fever, than ever before (Wilder-Smith and Gubler, 2008). Although multiple factors contribute, three forces described above are especially important: urbanization with major expansion of populations living in tropical and subtropical areas; population size; and rapid, frequent travel of viremic humans to areas infested with competent vectors. Lax vector control programs and urban settings that lack piped water (so residents must store water in their homes) and are littered with used tires, discarded plastic cups, and other trash with standing water that allow breeding of mosquito vectors exacerbate the problem. Today, the dengue viruses have a much larger host population in which to replicate, recombine, and mutate than ever before, given the size of the human population. Zanotto and colleagues (1996) found that the number of dengue lineages has increased roughly in parallel with the size of the human population over the past 200 years (Figure 2-7). More urban areas in tropical and subtropical regions now have a population size large enough (estimated to be between 150,000 and 1 million) to allow the ongoing circulation of dengue viruses. An increasing number of geographic areas are experiencing cocirculation of more than one dengue serotype, setting the stage for secondary infections and more severe disease.

FIGURE 2-7

Global population size and lineages of dengue virus over time. SOURCE: Zanotto et al. (1996).

Spread of Chikungunya Virus Infections

Although the mosquito-transmitted alphavirus, chikungunya² (family Togaviridae), was first identified as a cause of outbreaks in Africa in 1953 (Lumsden, 1955) and has caused outbreaks at irregular intervals since then in Africa and Asia, the virus has caused multiple outbreaks in new geographic areas since 2004. Outbreaks in coastal Kenya in 2004 were followed by multiple explosive outbreaks on the islands of the Indian Ocean beginning in 2005 (Charrel et al., 2007; Josseran et al., 2006; Renault et al., 2007; Sergon et al., 2008). Subsequent outbreaks occurred in India, Thailand, Indonesia, Malaysia, and other Asian countries (AbuBakar et al., 2007; Mavalankar et al., 2008). In India alone, more than a million cases were reported. In some areas, extremely high attack rates of infection were observed, affecting more than half of the population (Sergon et al., 2008). Because the virus causes a severe, frequently incapacitating polyarthralgia³ as part of the illness, the infection has been devastating in areas with high attack rates (Simon et al., 2008). Unlike dengue fever, another acute febrile illness, chikungunya infection is frequently followed by persistent joint symptoms that can continue for months or even years.

More than 1,000 cases have occurred in travelers who have returned to non-endemic regions (Hochedez et al., 2006; Panning et al., 2007; Parola et al., 2006; Simon et al., 2008). A high number of travelers with chikungunya fever have been diagnosed in Europe, reflecting frequent travel destinations in Indian Ocean islands. Chikungunya fever has always been considered a disease of the tropics, but in the summer of 2007, 175 cases were laboratory confirmed as the cause of an outbreak of an acute illness in two villages in northeastern Italy (Rezza et al., 2007). The clinical attack rate increased with increasing age in this outbreak, as has been observed in other outbreaks as well. Investigation identified a visitor from India as the index case. Transmission occurred during the hottest months of the year and stopped when temperatures cooled. Chikungunya virus was found in Aedes albopictus, the presumed mosquito vector in this outbreak.

Textbooks describe chikungunya fever as a self-limited illness with fever, polyarthralgia, headache, and rash, but in the recent outbreaks excess mortality was also reported (Figure 2-8; Mavalankar et al., 2008). Deaths have been concentrated in older persons.

FIGURE 2-8

Increased mortality associated with chikungunya outbreak, India. SOURCE: Mavalankar et al. (2008).

Because virus is present in high concentration in the blood of infected humans (may exceed 10⁹ RNA copies/mL of plasma), potential risk exists for nosocomial transmission. At least one instance is described of transmission to a health care provider in France after blood exposure (Parola et al., 2006). The virus could also be transmitted by transfusion of blood donated during the period of viremia. Investigators have modeled the potential risk for transfusion in an area with an outbreak (Brouard et al., 2008). In Reunion, where a massive outbreak occurred in 2005–2007, blood collection was interrupted for several months during the epidemic. Based on sentinel surveillance, knowledge of the duration of viremia in chikungunya infections, and the frequency of asymptomatic infections, the authors estimated that the risk of viremic blood donation was 1,500 per 100,000 donations during the peak of the epidemic and was a mean of 132 per 100,000 donations over the entire course of the outbreak. Returning travelers who donated blood during asymptomatic viremia could also be a potential source of infection in new geographic areas.

A mutation in the chikungunya virus may partially explain the intensity of recent outbreaks (Schuffenecker et al., 2006). Ae. albopictus, originally found in Asia, has become widely dispersed beyond Asia and is now found in many parts of the Americas, Europe, Africa, and the Pacific Islands. Although in the past the mosquito had not been considered a particularly efficient vector, it was identified as the primary vector in some of the recent chikungunya virus outbreaks (Reiter et al., 2006). A mutation has been identified in a viral gene encoding the envelope protein of the virus (A226V) and was found in more than 90 percent of viral isolates in the latter part of the outbreak in Reunion and in isolates from the outbreak in Italy. Of note, presence of the mutation is associated with enhanced susceptibility of Ae. albopictus to infection with chikungunya virus and to more rapid dissemination into the mosquito salivary glands. This means that a mosquito can become infected when exposed to a lower level of viremia, which may give the mutant virus enhanced survival benefit (Tsetsarkin et al., 2007). Mosquitoes vary in their susceptibility to infection and competence to transmit viruses (Tesh et al., 1976). Recent studies of Ae. aegypti and Ae. albopictus strains from Florida, using a chikungunya virus isolate from the Reunion outbreak, demonstrated that the mosquitoes were susceptible to infection and capable of transmitting the virus (Reiskind et al., 2008). Large areas of the Americas are infested with Ae. albopictus and Ae. aegypti. The potential for introduction of chikungunya virus is greatest in areas with high temperatures, abundant mosquitoes, housing or work places that allow mosquito-human contact (e.g., absence of window screens and air conditioning), and large volumes of travelers arriving from areas of Africa or Asia experiencing outbreaks (Charrel et al., 2008; Wilson, 2009).

Other Routes of Transmission

Dengue infection (which can lead to dengue hemorrhagic fever) can also be transmitted by blood transfusion (Tambayh et al., 2008) and by blood exposures in health care settings (Chen and Wilson, 2004, 2005). This may occur more frequently than is documented because occasional events of nosocomial and transfusion-related transmission could be difficult to distinguish from background infections in areas where infection is endemic and epidemic. Such events are more likely to be documented in nonendemic regions (Chuang et al., 2008). Recent studies have identified dengue RNA in blood donated by residents in Puerto Rico, Brazil, and Honduras (Linnen et al., 2008; Mohammed et al., 2008).

Medical Tourism

A reason for travel that is cited by increasing numbers of international travelers is medical treatment abroad (Reed, 2008). An estimated four million people travel internationally each year for medical care that often involves surgical procedures. A few countries attract the bulk of these travelers. Thailand, for example, currently receives about a million patients per year, and India, Singapore, and Malaysia are expected to attract similar numbers by 2012 (Smith et al., 2009b). Those who seek medical treatment abroad often undergo major procedures, such as heart bypass graft surgery, heart valve replacement, hip replacements, and bone marrow and liver transplantation, procedures that frequently require support by transfusions of blood and blood products. Some international institutions are seeking accreditation by the Joint Commission International⁴ to increase patient confidence in the quality of care. Still, in addition to ethical and legal issues, there are concerns about unrecognized risks to these medical tourists because of geographic differences in disease risks. Even if locally donated blood is tested for any evidence of infection, as would be carried out in the United States, other risks may exist in these areas, such as locally endemic and epidemic diseases, like dengue and chikungunya (Wilder-Smith et al., 2009). At present, blood donated in these regions is not screened for these infections.

Travelers as Sentinels

Travelers can serve as sentinels and couriers and should be an integral part of the global surveillance system (Wilson, 2003). Although travelers to an area do not experience all of the same infections seen in a local population, their infections reflect those present in the area. Returned ill travelers often have access to medical care in facilities where diagnoses can be confirmed, bacteria cultured, malaria species identified, viruses isolated, specific serologic tests carried out, and sensitivity testing done. Under some circumstances, sequencing of an organism or molecular studies carried out in a research laboratory can yield insights that can be helpful in the understanding, prevention, control, or treatment of the disease in the area where it was acquired. Infections in travelers can also provide a global alert about infections in an unexpected location or with an unusual resistance pattern. In some instances, the diagnosis of an infection in a returned traveler can be the first indication of the presence of an infection in a particular geographic area or at a specific site. During their visits, travelers sample the biome of an area simply by staying in the area, and they can be a courier in which the microbes survive, replicate, and can later be sampled and examined.

Study of infections in travelers has also provided new insights about transmission mechanisms that might not have been identified in endemic areas where laboratory support is less robust. Examples have been documented of transmission of dengue to health care workers by exposure to blood of infected travelers in temperate areas where dengue is not endemic (Chen and Wilson, 2004).

The GeoSentinel Surveillance Network,⁵ a worldwide communication and data collection network of travel and tropical medicine clinics started more than a decade ago, systematically gathers information on ill international travelers and migrants and has been able to provide early alerts about unusual infections or infections in unusual locations or populations. This global network includes staff at 42 travel and tropical medicine clinics with sites on all six continents. As of early 2009, data from more than 100,000 clinical visits had been entered into the database. Despite the limitations of data gathered by this type of surveillance network (Leder et al., 2008), analyses have yielded useful insights, including showing how the spectrum of disease varies depending on the place of exposure among ill returned travelers (Freedman et al., 2006).

Falciparum malaria infections in travelers returned from the Dominican Republic signaled the reappearance of malaria in parts of the island and led to changes in recommendations for chemoprophylaxis (CDC, 2005). The increasing resistance of salmonellae to quinolones documented in returned travelers with typhoid fever has influenced initial treatment choices in patients with severe typhoid fever. An index case of schistosomiasis in a traveler returning to Israel from a luxury safari trip to Tanzania prompted an investigation leading to the identification of 22 cases of acute schistosomiasis (Leshem et al., 2008). Subsequent investigation revealed that 81 percent of those exposed at a specific site became infected. Early diagnosis of the index case enabled clinicians to study the usefulness of different diagnostic tests and observe the clinical course of acute schistosomiasis in nonimmune persons who had been infected during a single short (mean duration, 40 minutes) exposure to an unchlorinated, freshwater pond. Dissemination of information about the outbreak through electronic networks (ProMED and GeoSentinel) made it possible to inform the global community about the outbreak more than a year before a paper about the outbreak appeared in a peer-reviewed medical journal.

Analysis of dengue cases from the GeoSentinel database over a 10-year period revealed that a surge in dengue infections in returned travelers could herald an increase in dengue-endemic countries in identified regions (Schwartz et al., 2008). In 2002, an increase in cases led GeoSentinel staff to post an alert on ProMed (Freedman et al., 2002). In several instances, an increase had been evident before official surveillance data were available from specific countries, which reinforces the importance of sentinel surveillance in travelers.

A different kind of network, the Boston Area Travel Medicine Network (BATMN), is in the early stages of collecting serum samples on selected travelers before and after travel to begin to assess exposures to infections in different geographic regions.

Conclusion

Travelers play a critical role in the movement of microbes globally. In an increasingly interconnected world with a growing, increasingly urban population in low-latitude areas, new risks exist and disease-causing microbes and resistance genes can move even more rapidly than in past decades. Travelers can also serve as an important sentinel population. Studying them can help to characterize the global microbial traffic.

Causes of Infectious Diseases Due to War

Among the major causes of infectious diseases due to war are:

• Adverse effects on medical care and public health services,
• Damage to the health-supporting infrastructure and the environment,
• Forced migration,
• Diversion of resources, and
• Biological weapons.

Each of these causes deserves further discussion.

Forced Migration

There are approximately 12 million refugees and 22 to 25 million internally displaced persons globally. They suffer from loss of sociocultural support systems and reduced access to safe food and water, inadequate medical care and public health services, and inadequate clothing and shelter, and they are at increased risk for many infectious diseases. For example, in 1980 at a camp for Cambodians in Thailand, the leading diagnoses in the emergency ward of a 1,000-bed field hospital were predominantly infectious diseases (Table 2-2). In 1994, approximately 1 million refugees from Rwanda fled to Zaire in less than one month; many of them died from cholera or dysentery soon after arrival in refugee camps there. In 2005, mortality among internally displaced persons and others in Darfur, Sudan, was mainly due to infectious diseases (Table 2-3; World Health Organization and Federal Ministry of Health, Sudan, 2005).

TABLE 2-2

Leading Diagnoses, Emergency Ward, Khao-I-Dang Camp for Cambodians, Thailand, 1980.

TABLE 2-3

Cause-Specific Mortality Among Internally Displaced Persons in Camps, South Darfur, May–June 2005.

Biological Weapons

There is a long history of the sporadic use of biological weapons during war or preparation for war. Examples have included contaminating drinking water with microorganisms, hurling of plague victims into a walled city, infecting blankets with smallpox virus, placing dead animals in water sources, infecting horses with glanders bacteria, and testing anthrax bombs on a deserted island (Metcalf, 2002). In the 1950s and the 1960s, the United States and the Soviet Union each developed a large infrastructure for research and development of both offensive and defensive biological weapons. In 1972, the Biological Weapons Convention was signed, which banned the development, production, stockpiling, or acquisition of biological weapons and their means of delivery, except for peaceful purposes. Although there is no formal verification regime for the Convention, 162 nations have now signed or ratified it.

The Centers for Disease Control and Prevention has developed three categories of diseases caused by biological agents, based on the severity of these diseases and the presumed likelihood that they could be caused by bioweapons. Category A includes anthrax, botulism, plague, smallpox, tularemia, and viral hemorrhagic fevers; category B includes brucellosis, disease caused by epsilon toxin of Clostridium perfringens, food safety threats, glanders, melioidosis, psittacosis, Q fever, ricin toxin, staphylococcal enterotoxin B, typhus fever, viral encephalitis, and water safety threats; and Category C includes emerging infectious diseases, such as those caused by Nipah virus and hantavirus.

Preparedness for bioterrorism should be placed more appropriately among U.S. national priorities for prevention and control of infectious diseases. Table 2-4 shows the incidence and mortality for selected causes of infectious disease in the United States from 2001 to 2004, demonstrating that the numbers of incident cases and deaths for AIDS, hepatitis C, and hospital-associated infections were much higher than those for bioterrorism.

TABLE 2-4

Incidence and Mortality for Selected Causes, United States, 2001–2004.

The U.S. National Counterterrorism Center Report for 2007 reported on approximately 14,000 terrorist attacks worldwide, accounting for approximately 22,000 deaths and approximately 44,000 wounded people. Armed attacks and bombings accounted for the vast majority of fatalities. However, none of these reported attacks were due to biological agents (U.S. National Counterterrorism Center Report, 2008).

What Needs to Be Done

The following measures need to be better designed and implemented to address the problems related to armed conflict and infectious disease:

1. Surveillance for infectious diseases, and also for factors that are known to increase the risk of armed conflict;
2. Evaluation of prevention and control measures for infectious diseases related to war, and for measures to help prevent armed conflict;
3. Protection of medical care and public health services and maintenance of their neutrality during war and the aftermath of war;
4. Elimination, or the prevention of increases, of disease vectors during war and the aftermath of war;
5. Epidemic preparedness and responsiveness to outbreaks, especially in less-developed countries;
6. Diagnosis, treatment, and prevention of infectious disease;
7. Protection of the health-supporting infrastructure and the environment during war and the immediate aftermath of war;
8. Reduction of forced migration during war and its aftermath, and meeting of the basic needs of refugees and internally displaced persons and protection of their human rights;
9. Control of biological agents and strengthening of the Biological Weapons Convention; and
10. Creation of a world without war by addressing the underlying causes of war, controlling weapons, and strengthening the infrastructure for peace.

Ultimately, eliminating infectious disease caused by armed conflict will require the elimination of armed conflict.

Cross-Border Trade and the Spread of Infections

In 2008, total global merchandise trade was valued at more than $15.8 trillion (WTO, 2009). Although the annual rate of growth in global trade dropped from 8 percent in 2006 to 6 percent in 2007 and continued its decline in 2008, the international exchange of goods and services has continued to increase, albeit at a slower rate. The following case studies examine the risks that emerge in a world in which people and products are continually crossing borders.

Far-Flung Distribution

The 2006 multistate outbreak of Escherichia coli O157:H7 across the United States (Grant et al., 2008) linked to spinach grown in California aptly demonstrates how cross-border trade can expand the scope and complexity of outbreaks. Between August 5 and September 5, a total of 84 cases were detected in 20 states. Only one of the 84 cases was in California; the other 83 were spread across the country, from Oregon to Wisconsin to New York to Tennessee. Consequently, there was no ability at the local level to detect the outbreak or identify its source. It was not until investigators conducted the molecular epidemiology across affected states that the cases could be linked and the source identified.

Trade across borders (state or national) creates a new, very direct dissemination of infection and a new challenge for public health, both at the local and the global levels. As Figure 2-9 illustrates, trade and travel are the mechanisms by which local outbreaks become pandemics. And as the blue arrow indicates, it is the growth of transnational trade and travel that enhances the risk of the transnational spread of disease. The E. coli outbreak described above shows that, when agents enter the cross-border trade flow, local public health authorities at the source may not, in fact, be in a position to perceive and address disease clusters.

FIGURE 2-9

Trade and travel are key to global dissemination of disease.

The Perfect Storm: H5N1

The influences of the global economy, in particular increased urbanization and intensified agricultural processes, have contributed to the brewing threat of an avian influenza pandemic. These dynamics are particularly evident in areas of Southeast Asia, where poor families are moving, along with their animals, from rural areas to crowded periurban settings in pursuit of economic opportunity. The families maintain their food animals in backyard farms with poor sanitation and water supply—an opportune blend for the emergence of bird flu. Further compounding the risk of spread is the nearby introduction of intensive poultry agriculture operations that raise thousands of birds from hatching through slaughter. Between 1980 and 2000, poultry exports from Far East Asian countries leapt from less than 200,000 tons of exports to nearly 1,800,000 tons (Figure 2-10; Kimball, 2006). These burgeoning poultry operations generate a variety of potential biosecurity risks such as crowded coops with large volumes of waste—risks that they do not always have the resources or incentive to mitigate. Added to the mix are free-ranging ducks that serve as asymptomatic reservoirs for the H5NI virus.

FIGURE 2-10

Poultry exports from Far East Asian countries from 1961 to 2002. SOURCE: Based on data in FAOSTAT 2003 and reprinted from Kimball (2006) with permission from Ashgate Publishing.

Ultimately, these communities crowded with people and poultry are approaching an ecological tipping point that could someday result in an influenza A (H5N1) pandemic among humans. There are, in fact, two distinct opinions on this issue: (1) large industrial operations provide enhanced biosecurity and backyard poultry should be eliminated to reduce risk, and (2) the large-scale operations are themselves risky. While definitive studies are lacking, a careful review of the Thai experience suggests that the second opinion is more aligned with the evidence (Graham et al., 2008). Already, H5N1 outbreaks have occurred in poultry throughout Asia. As of September 2009, 442 cases of human H5N1 and 262 deaths had been documented in 15 countries, including China, Hong Kong, Thailand, Vietnam, Cambodia, and Indonesia (WHO Global Alert and Response, 2009). To date, the H5N1 virus has had limited ability to move from poultry to humans and even less ability to move from person to person. The rapid global spread of the recent H1N1 virus linked to pigs provides sobering evidence of the pandemic potential once an agent that originates in animals develops the capacity to spread among humans.

Bovine Spongiform Encephalopathy and New-Variant Creutzfeldt-Jakob Disease

The emergence of BSE and new-variant Creutzfeldt-Jakob disease (nvCJD) provides another example of the potential consequences of the pressures that result from global trade. In this case, the chain of events arguably began when the United Kingdom entered the World Trade Organization (WTO) in 1993 (Gibbs and Shaw, 1996). As a condition of joining the WTO, the United Kingdom lifted some 13 percent of the tariffs charged for beef imports, thus opening their beef market to greater global competition. This in turn resulted in a consolidation of the previously fragmented beef industry. As part of that transition, beef producers changed their rendering practice and, to some extent, their animal husbandry practices.⁸ Specifically, they changed their rendering process from a so-called batch procedure to a lower-temperature vacuum-extraction process. This process cost less and enabled continuous production, but unfortunately it had a hidden downside as well—it did not deactivate the prion, a new pathogenic agent with a 10-year incubation period. This long incubation period enabled extensive circulation before any problems were detected. During this period, meat and bone meal that had been rendered in Great Britain from cattle infected with BSE were shipped around the world in animal feed.

In the mid-1990s, a new series of cases of vCJD began to appear in the United Kingdom. The frightening disease, which claims victims’ lives in 18 months and has no known treatment, was linked through case-control studies to consumption of beef from the United Kingdom. While scientists continue to debate the causal link between prion disease in cattle and prion disease in humans, the global markets reacted swiftly to the possibility. UK beef exports plummeted in 1996 (Figure 2-11) after trading partners began embargoing beef shipments under the WTO’s Sanitary and Phytosanitary (SPS) agreement. The SPS agreement allows importers to issue “urgent notifications” that they will suspend import of a particular product when human safety is at risk (WTO, 2000). Thus, in the world of global food trade, where safety concerns can amount to economic devastation, incentives run against reporting potential risks.

FIGURE 2-11

Superimposed epidemics. Long incubation period of up to 10 years allowed extensive circulation of meat and bone meal-infected product in the global market prior to identification of risk. SOURCE: Reprinted from Beisel and Morens (2004).

Antimicrobials in Farm Animals

Global economic pressures have contributed to other risky practices in animal husbandry, including the use of antibiotics as growth promoters, which fosters the development of antibiotic-resistant pathogens in both animals and humans. It is estimated that, in the United States alone, some 12 to 70 million kg of antibiotics are administered annually in this way (Aarestrup and Pires, 2008). Evidence suggests that the use of antibiotics in animal feed may be even higher in developing economies that carry heavier disease burdens. In some parts of Asia, integrated fish farming facilities feed fish farm-animal waste that is laced with antibiotics to suppress the level of microbes. Humans in turn consume the fish, along with the antibiotics, potentially perpetuating the spread of antibiotic-resistant pathogens

The European Union has banned the use of antibiotics as growth promoters in food animals (Silbergeld et al., 2008). A similar global ban prohibiting the application of antibiotics for farm animals outside of therapeutic use would be well advised.

The use of antimicrobials as growth hormones and in feed is another illustration that underscores the fact that incentives in global trade often are misaligned with the interests of biosecurity. Add to that the far-reaching and rapid distribution system for food products and the result is a pandemic risk that cannot be ignored.

Beyond Food: Xenotransplantation

Another growing facet of global trade in services that poses a significant microbial risk is xenotransplantation, or the transplantation of animal parts into humans. Implanting animal organs or valves in humans introduces yet another level of cross-species exposure to pathogens. A variety of global pressures are driving this trade, including increases in diabetes and other chronic conditions detrimental to organ health, a lack of effective therapies for degenerative conditions, and, in countries such as the United States, an upcoming surge of baby boomers entering old age. As a result, the number of people seeking organs far exceeds the number of organs available for transplant and in many countries, people seeking transplants face long waiting lists.

Some players in the global marketplace have responded to the growing demand for organs and alternative therapies by venturing into experimental areas such as xenotransplantation and stem cell therapies that are strictly regulated or even prohibited in other countries. Stem cell procedures and “precursor” stem cell procedures are promoted, often via the Internet, on the global medical marketplace. They include various donor species, such as rabbits and pigs.⁹ The majority of these procedures are offered without systematic protocols, primate studies, or regulation.

In the United States, the transplantation of tissues from other mammalian species is highly regulated. For instance, although the pig has been extensively explored as a potentially promising donor source, the United States has prevented the licensure of pig organs (with the exception of heart valves) because pigs carry in their genetic material endogenous retrovirus known as “PERV” (Porcine Endogenous Retrovirus), which potentially could infect its human host post-transplantation. In light of such concerns, the World Health Assembly in 2004 passed a resolution on xenogenic transplantation that urges member states to establish regulation and surveillance mechanisms and to collaborate on global strategies to prevent infections (Resolution 57-18 2004, section II). Although the United States was a key promoter of this resolution, it did not invest in World Health Organization (WHO) programming to support the initiative. Consequently, xenotransplantation regulatory progress has lagged on a global level. The WHO collaborating center at Geneva University attempts to keep a dataset on xenotrans-plantation therapies;¹⁰ however, the information is not exhaustive.

Global Problem: Global Solutions

The foregoing cases have demonstrated how our increasingly global economy, with growing international travel and trade (including trade in services such as transplantation), has ultimately made virtually any emerging microbial risk global in nature. In the examples of foodborne E. coli and BSE, we see the globalization of direct infectious risk. In the instance of the overuse of antimicrobials in food animals, we see the globalization of antimicrobial resistance. With medical travel for organ transplants, we see traveling patients become potential vectors for the spread of disease.

While microbial risks have been globalized along with commerce, the corresponding health and protective measures for the most part have not. The second edition of the IHR (2005), which took effect in 2007, provides some important safeguards to help limit the international spread of infectious disease. The IHR require countries to conduct surveillance for and report to the WHO a “public health emergency of international concern,” that is, an event “that may cause international disease spread.” If WHO determines such a threat exists, as it did with the recent H1N1 outbreak, it may issue recommendations to curb the spread of disease, such as quarantine or travel restrictions for affected or potentially affected individuals. As the experience with H1N1 demonstrated, WHO must carefully balance the threat of disease spread with the potential economic consequences of any travel or trade restrictions in order to minimize disincentives for countries to report potential threats. While WHO Director-General Dr. Margaret Chan raised the “Pandemic Alert” level to 6 (the highest), WHO actively discouraged trade and travel restrictions after determining that they would not be effective in curbing the spread of the influenza virus and could needlessly result in significant economic repercussions. Instead, WHO focused on identifying and treating individuals with infection and urged those individuals with illness or symptoms to avoid travel and contact with others. This did not stop some countries from instituting their own travel restrictions. Several nations banned flights to Mexico, and China quarantined more than 70 travelers from Mexico (Browne, 2009). Despite the moderated response, Mexican authorities estimated $2.2 billion losses to the nation’s economy as a result of the outbreak, including more than a 40 percent drop in tourism revenue (Llana, 2009).

The revised IHR represent an important step toward a more coherent global response to microbial threats. However, the full implementation of the core capacities for public health competency will not be in place until 2011, and resources will be needed to achieve that benchmark. Another promising initiative has been the efforts of the Asia Pacific Economic Cooperation (APEC) to create a community of interest in health across the 21 APEC economies, including a communications platform to enable members to collaborate in the event of public health emergencies. The APEC effort actively seeks to bring in the vibrant private sector through collaborative projects in biopreparedness led by its Health Working Group. The full engagement of the private sector as a stakeholder would bring a key “driver” of global trade into the discussion and implementation of public health protection.

In conclusion, there are some key questions raised by these case studies for which critical information is missing: (1) What exactly is the nature of emergent influenza risk of industrialized poultry and swine practices in poor and medium-income settings and can these be mitigated? and (2) What surveillance of trade product and practice would be useful to inform global public health? Moreover, at a global level, the IHR process requires support to ensure capacity building to enhance global access to surveillance, laboratory, and epidemiological investigation capabilities; and community building to improve cross-border communications and collaboration.

Our discussion of xenotransplantation highlights some of the risks entailed in the growing practice of medical tourism, or the movement of patients across international borders, issues that we explore in greater depth elsewhere (Hodges and Kimball, in preparation).

Finally, as we move into the pandemic phase of H1N1 and the Northern Hemisphere moves into its next influenza season, it is timely to reflect seriously on the risk our ever-globalizing trade, travel, and food production poses to population health and biosecurity.

Real-World Challenges

The FDA has encountered several issues linked to the globalization of the food supply. The most pervasive is the nondeliberate or accidental contamination of food. For example, in 2003, a massive hepatitis-A outbreak of more than 1,000 cases across multiple states was linked to green onions imported from Mexico. In 2008, an outbreak of Salmonella Saintpaul caused by contaminated hot peppers imported from Mexico resulted in 1,450 documented cases, which probably represents a fraction of the actual number of people who got sick.

We have also dealt with the deliberate contamination of food for the purpose of economic gain. This opened a new arena for regulatory consideration, because, unlike bioterrorism, these deliberate acts are not intended to cause harm. One such case occurred in 2007, when it was discovered that the chemical melamine had been added to wheat gluten by Chinese manufacturers (FDA, 2008a). Because melamine is high in nitrogen, it increased the apparent protein content of the wheat gluten (which is gauged by nitrogen content), and therefore its price. The use of this adulterated wheat gluten in pet foods manufactured in the United States resulted in animal deaths due to kidney failure. Unfortunately, lessons were not learned from this experience, and within a year, the FDA was investigating melamine-contaminated dairy products from China, where some infants allegedly died after ingesting formula containing melamine, and more than 50,000 were reportedly hospitalized in China for urinary problems (FDA, 2007; WHO, 2009).

The following descriptions of the investigations of foodborne Salmonella Saintpaul in 2008 and melamine in pet food in 2007 illustrate the complexities involved in determining the sources of contaminants once they enter the globalized food system.

Salmonella Saintpaul in Peppers

The 1,450 cases confirmed in the 2008 Salmonella Saintpaul outbreak occurred in 43 states and the District of Columbia, and they were linked to multiple food types. The outbreak came to the attention of the FDA via an alert from the Centers for Disease Control and Prevention (CDC) in early May 2008. Within two weeks, tomatoes had emerged as the likely vehicle, so we initiated a traceback to determine the origin of the contamination. This led us to Florida and Mexico, where we began inspecting tomato farms and their supply chains.

When we approached the Mexican government about this issue, they were very cooperative and met with us daily. The FDA visited several farms in the state of Sinaloa, Mexico, but failed to find tomatoes contaminated with Salmonella, and as the epidemiological investigation continued, other food sources were implicated. In July, we began tracing back peppers, as illustrated in Figure 2-13.

FIGURE 2-13

Salmonella Saintpaul outbreak and distribution. Partial view of the traceback and distribution of peppers from Mexico, July 16–22, 2008. SOURCE: FDA (2008b).

As Figure 2-13 illustrates, this was an extremely complicated process, and it begs the question as to how such episodes of contamination, which are occurring with increasing frequency, can best be prevented and addressed. The investigation led us to a distributor in McAllen, Texas, where we found positive pepper samples. They had received their jalapeno peppers from a packing facility in Nueva Leone, Mexico, so we returned to an entirely different part of that country to inspect pepper farms. On July 30, we found the outbreak strain, Salmonella Saintpaul, in irrigation water and on remaining peppers on a farm in Tamaulipas, Mexico.

Melamine in Pet Food

This investigation began with reports of sick pets from consumers, and also from a pet food company whose research animals developed kidney failure following routine taste tests (FDA, 2008a). The only associated change in pet food formulation was the source of the wheat gluten it contained.

Once this was recognized, we soon determined that scraps from affected pet food manufacturers were being used to produce food for livestock, which could in turn introduce the adulterant into foods consumed by humans. Thus, melamine was traveling up the food chain, beginning with food ingredient manufacturers, and onward to feed mills, to poultry farms and hog farms, and from there to chicken and pork in supermarkets. A joint risk assessment was conducted between the FDA and the U.S. Department of Agriculture (USDA) to determine if the levels in poultry and pork were of a concern to public health. The assessment concluded the levels were not a public health concern. We also discovered that not only was this wheat gluten coming into the United States, but a U.S. company was also importing it to Canada. We alerted the Canadian authorities, who found that the gluten was being used there to make fish feed that was being shipped back into the United States.

The complete U.S. distribution chain of the melamine-contaminated wheat gluten is shown in Figure 2-14. It makes the point that contaminants—including pathogens—connected to food may be disseminated through vast and complex systems that profoundly affect international trade and economic relationships, as other workshop participants noted.

FIGURE 2-14

Sequence of events—different routes of melamine-contaminated wheat gluten into the United States. SOURCE: Food and Drug Administration.

Time for a New Approach

In light of these complexities, the FDA is attempting to change its approach by becoming more proactive in addressing food safety by addressing the whole supply chain. One important route to this goal is to conduct targeted, risk-based inspections, but these may be difficult to identify. Risk-based inspections are dependent on having adequate data, analytical capabilities, and inspectors. Currently such inspections are conducted at ports of entry, where it is decided whether a given product made in a foreign country and shipped to the United States will be inspected or tested for chemical or microbiological contamination, based on the various factors such as the nature of the product, its origin and destination, and the past inspection history of similar products. This procedure, euphemistically called a “snapshot,” examines only one point in the distribution chain: importation. In the future, our goal is to gain greater assurance for the complete supply chain—from grower to manufacturer to shipper to importer to distributor to retailer—of imported food products. This is going to require a multifaceted approach that includes more inspections.

There are significant challenges and scarce resources for pursuing inspections of overseas growers and manufacturers, among them the sheer numbers (approximately 200,000, as previously noted) of registered foreign food facilities. There is no way, despite some peoples’ wishes, that we could inspect and sample every import at the port of entry. Rather, we must try to optimize inspection based on risk. We cannot simply test or inspect our way to safe food and having appropriate preventive controls throughout is a key step.

Risk-Based Inspections

To pursue a risk-based approach to food safety, we need leverage with foreign governments and with industry in order to gather information on the status of foreign growers, manufacturers, and foreign governments. Using these data, we must conduct effective analyses to inform risk-based decisions at the port of entry. Finally, rather than determining whether to examine food products on a case-by-case basis, we need to build a decision-making system.

We are currently using a prototype of such a system, which we call PREDICT, to maximize the use of FDA data in making risk-based decisions. PREDICT software allows the FDA to make risk-based decisions for import inspections and has the capacity to receive new datasets as other data streams come online. A pilot project, conducted in the Port of Los Angeles, employed PREDICT to make inspection decisions regarding seafood, for which clear inspection standards had already been established. In this trial, PREDICT prompted inspectors to detect contaminated foods more frequently than did the existing inspection system (known as OASIS). The FDA is currently attempting to expand PREDICT, recognizing that this system is only as good as the data provided to it.

Increased Foreign Presence

Another way the FDA is supporting a more proactive stance on addressing threats to the globalized food supply is by establishing a greater foreign presence. To that end, we recently opened three offices in China—in Beijing, Shanghai, and Guangzhou—staffed by about a dozen permanent FDA employees. Although their staffs are too small to conduct significant numbers of inspections, these foreign offices will be able to strengthen relationships with the Chinese government that will improve our ability to exchange information and deal with contamination issues as they arise, and potentially before they become problems for the United States. In order to situate FDA personnel throughout the world, we are opening similar outreach offices in New Delhi and Mumbai, India; in central South America; and in Europe.

Use of Third-Party Data

The food industry extensively inspects and tests foods that are imported into the United States. Can the FDA make use of this information? Recognizing that this is a potentially contentious issue (see below), our goal is to examine the process surrounding third-party certifications and determine how the FDA can use that information to better protect the public and make maximal use of resources. One aspect of this is to use standards for third parties that provide information supplied to the FDA, whether it is derived from state agencies, foreign governments, or private third parties. We then want to be able to use that information to inform the risk-based inspection process. Currently, the PREDICT model at ports of entry operates primarily on U.S. compliance data. Incorporating inspection data and information from overseas and from industry into our analyses should improve our ability to make risk-based decisions.

Conclusion

The regulatory challenges involved in providing safe food to the United States will increase as globalization of the food system continues. The complexities of the food supply are enormous, and there is considerable economic benefit in making food distribution as rapid and efficient as possible. It is therefore critical that we abandon the practice of focusing our efforts on ports of entry and instead embrace a new approach focused on the whole supply chain that attempts to understand foodborne threats in foreign countries, anticipates their potential to spread globally, and uses risk-based inspections to detect them before an outbreak occurs in the United States.