Textbook of Influenza

By Robert G. Webster, Arnold S. Monto, Thomas J. Braciale and Robert A. Lamb

The Textbook of Influenza is a comprehensive resource covering all aspects of influenza, from the genetic and molecular biology of the virus through to clinical aspects of the disease and the latest drug developments and treatments. This new edition has been completely revised and reflects the integration of disciplines concerning the emergence, evolution, pathogenesis and control of influenza viruses in the field of human and veterinary public health.

Textbook of Influenza examines the lessons learnt from the latest pandemic and provides the current state of knowledge for many yet unresolved issues related to virus origin, spread, pathogenesis and disease severity to better prepare for future pandemics. It covers the background to recent advances in influenza genomics and reverse genetics which have allowed the identification of virus virulence factors and the analysis and reconstruction of influenza viruses such as the 1918 Spanish flu strain.

This new edition is divided into eight key sections, containing chapters co-written by international experts from both the clinical and scientific communities, covering:
• Influenza Perspectives
• Structure and Replication
• Evolution and Ecology
• Epidemiology and Surveillance
• Immunology
• Vaccines and Vaccine Development
• Clinical Aspects and Antivirals
• Public Health

Textbook of Influenza is for all those working in the area of influenza including clinical and basic scientists, immunologists, molecular and structural virologists, public health officials and global pandemic control planners.

• Language: English
• Publisher: Wiley
• Release date: Jan 6, 2014
• ISBN: 9781118636831

Robert G. Webster

Dr. Robert Webster has worked in the field of Virology for over 30 years, first in New Zealand and then at the John Curtin School of Medical Research in Australia. He has spent the past 25 years at St. Jude Children’s Research Hospital in the Department of Virology & Molecular Biology. In addition to his position as Chairman at St. Jude, Dr. Webster is Director of the U.S. Collaborating Center of the World Health Organization dealing with the ecology of animal influenza viruses. He has served on numerous national and international advisory boards and is a Fellow of the Royal Society. Dr. Webster has published extensively on influenza in areas covering the origin of pandemic strains, genetic variation, structure and function of virus and, in conjunction with Graeme Laver, was responsible for the development of influenza subunit vaccines.

Book preview

PART 1

Influenza: Perspective

Section Editor: Robert G. Webster

1

Human influenza: One health, one world

Daniel B. Jernigan and Nancy J. Cox

National Center for Immunization and Respiratory Diseases, Influenza Division, Centers for Disease Control and Prevention Atlanta, GA, USA

Introduction

Influenza viruses know no boundaries, circulating within species and occasionally jumping between them, causing infections around the globe. The impact of influenza is also wide-ranging, and the growing interconnectedness and complexity of the world presents an increasing challenge to influenza prevention and control. As people and the animals that support them increase in numbers and interactions, the opportunities for virus adaptation and cross-species transmission increases as well. An undetected exchange of viruses among humans and animals in a rural village may eventually manifest as a global pandemic. Within this interconnected context, opportunities are also emerging for coordinated, collaborative, innovative, and integrated efforts to focus new technologies and approaches in a shared response to the global challenges of influenza.

Global impact of influenza

Influenza causes significant human illness and death each year. The actual global impact of seasonal influenza is difficult to determine due to incomplete surveillance data; however, the World Health Organization (WHO) has estimated that around 1 billion cases of seasonal influenza infection occur each year, with around 3–5 million cases of severe illness, and 300 000–500 000 deaths [1]. The global financial costs of seasonal influenza are also not well known. In the United States, where data collection is robust, estimates of the costs of influenza have been reported to be an average of $10.4 billion per year for direct medical costs and $87.1 billion for the total economic burden of annual influenza epidemics [2].

Influenza infection in birds and other animals also has a substantial impact. Since the detection and reporting of highly pathogenic avian influenza (HPAI) A (H5N1) epidemics in 1997, the virus has spread globally with hundreds of millions of birds dying from illness or culling. Costs for the international response since 2003 have been estimated to be at least $2 billion [3]. The presence of ongoing outbreaks in South-East Asia has damaged trade potential and has shifted exporting of poultry away to other nonaffected regions. For resource-challenged communities in HPAI H5N1 endemic regions, the loss of protein nutrition from decimated domestic flocks and the drop in income from lost poultry sales is significant. Swine are also infected by influenza viruses, but the burden of illness and death for swine is considerably less than that seen for HPAI viruses in domestic poultry. Influenza illness in swine is often considered by pork producers to have considerably less of an impact on production than on the potential downside for pork consumption and global export which might be prompted by public concerns about swine influenza infecting people. From a human health perspective, the greatest impact of influenza from swine and birds is their important role as sources of novel influenza viruses capable of causing pandemics.

In the last 100 years, there have been four instances where influenza viruses with genes originating from avian or swine reservoirs emerged (either with or without reassortment with human influenza viruses) with sustained, efficient, human-to-human transmission, spreading around the world [4]. The impact from these pandemics has been substantial, most notably with the 1918 H1N1 virus. It has been estimated that, globally, there were around 50 million human deaths due to infection with the 1918 pandemic virus [5]. Estimates of the 2009 pandemic were substantially lower, with 151 700–575 400 deaths globally [6]. If a severe pandemic were to arise from the H5N1 virus, the World Bank has estimated a cost of up to three trillion dollars [7]. For past pandemic viruses, it is not known exactly where or when the interchange of virus occurred from swine or bird to humans, but the subsequent impact from these pandemics was extensive. What are the factors that contribute to emergence of these viruses? What can be done for early detection and prevention?

Influenza in a crowded, connected, and converging world

Increasingly crowded

Over the last two centuries, the world's population has increased dramatically. The number of people on earth reached 1 billion sometime around 1804 [8]. After 123 years of growth, the population rose to 2 billion in 1927. Since then, the numbers have increased substantially, reaching 7 billion in 2011 (Figure 1.1). If the population continues to grow at the current rate, there could be around 11 billion people on the planet by 2050. The exponential explosion in numbers has not been uniform across the globe. Importantly, over 80% of the population increase has occurred in less developed countries. Currently, about 60% of the world's population resides in Asia, 15% in Africa, 11% in Europe, 7% in Latin America and the Caribbean, and 5% in Northern America [9]. By 2050, 83% are projected to reside in Asia and Africa.

Figure 1.1. Global population and travel trends, 1961–2010. Data sources include: United Nations Food and Agricultural Organization for poultry and pig population estimates [40], United Nations Department of Economic and Social Affairs for human population estimates [9], and United Nations World Tourism Organization for tourist arrival estimates [23].

c1-fig-0001

Much of this growth has occurred in large, dense, population centers. These communities, where the population exceeds 10 million inhabitants, are referred to as urban agglomerations or megacities. In 1970, there were only two such megacities, but in 2011, there were 23 [10]. Those most recently added to the list of megacities, and most likely to accelerate in size, including Lagos, Dhaka, Shenzhen, Karachi, Mexico City, Cairo, and São Paolo, are in developing countries and in tropical and subtropical regions. Residents of these urban agglomerations are, on average, much poorer and younger than in developed settings. Many of these megacities are facing a number of infrastructure and societal challenges, some of which also have implications for the control of influenza in humans and animals. These challenges include healthcare and public health infrastructure limitations, air pollution as a significant contributor to respiratory diseases, concentrations of the poorest of the population living in very crowded conditions, and stressed agricultural supply chains attempting to meet a growing appetite for protein [11,12].

Consider Bangladesh for example. In 2010, there were around 150 million people, a number just under half the population of the United States, living on land the size of the state of Iowa [13]. The population of the capital, Dhaka, was 1.4 million in 1970 [10]. It expanded 11-fold to 15.4 million by 2011. In 2025, this megacity is estimated to increase by another 1.5-fold to around 23 million residents. Dhaka currently is one of the world's most crowded cities. On average, the population density of Dhaka is around 32 600 persons/km² (84 500/mi²) [14]. In some exceptionally dense areas, such as the slums of Kamalapur, the density rises to 111 325/km² (288 332/mi²) [15]. Many of the residents in the area are recent immigrants from the countryside, and 44% are under the age of 20 years. In this crowded context with many children, respiratory infections abound. Studies have shown that childhood pneumonia occurs frequently among young children in Dhaka, and influenza and other viral pathogens are identified as the primary cause of these episodes [16,17]. In settings such as these, with the combined contribution of multiple socioeconomic factors and a considerable burden of respiratory illness, seasonal influenza has been identified as an important, and preventable, cause of illness and death.

Dhaka is not alone. Similar megacities are emerging where the promise of employment and a better quality of life continue to draw people to these exceptionally crowded conditions. What does all this mean for influenza among humans? As population density increases, infection transmission potential increases [18]. With a growing number of crowded urban settings, large numbers of younger individuals, and a high frequency of interaction among the population, influenza viruses are given all the right circumstances for efficient and sustained transmission. This is annually evident for seasonal influenza with high attack rates among schoolchildren, college students, military recruits, cruise ship passengers, nursing home residents, and others in densely populated settings. This is also the case with pandemic influenza. Crowded conditions were identified as a significant contributing factor for the 1968 H3N2 pandemic emergence in Hong Kong [19]. Population density and urbanization have been associated with pandemic spread and increased mortality for the 1918 pandemic in the United Kingdom [20]. An increasing population density not only provides an opportunity for influenza to be shared within a community, it also provides opportunities for influenza to be carried to other communities through travelers.

Increasingly connected

The world is more crowded but, just as importantly, it is also more connected. A person can travel to almost any major urban center in the world within the incubation period of influenza [21]. People travel for various reasons, and people are traveling more than ever. For 2011, the projected number of airline passengers was a record-setting 2.75 billion, over 600 million more than in 2006 [22]. Businesses now routinely utilize suppliers and services from distant locations, requiring more global travel. Tourism is also on the rise. From 1950 to 2011, the world's population increased by 2.5-fold, but the number of tourist arrivals increased over 37-fold [8,23] (Figure 1.1). Most increases were in tourists from Asia, and most tourists visited a few highly frequented locations.

The interconnection between communities has an important role in influenza epidemiology. Whereas influenza transmission within communities is predominantly driven by children, transmission between communities is predominantly driven by those who travel frequently, predominately adults. Travelers have been identified as a major contributing factor to the annual cycle of seasonal influenza. In one study, researchers used antigenic and genetic analyses to demonstrate that influenza A (H3N2) virus epidemics for the years 2002–2007 were initiated by influenza viruses originating from East/South-East Asia [24]. The researchers attributed the high frequency of travel and trade from East/South-East Asia to Oceania, North America, and Europe as a likely contributor to the migration of influenza A (H3N2). A lack of travel and trade with South America was suggested to account for the lag in appearance of new viruses to that region.

Tourists have also been identified to contribute to the spread of seasonal influenza. In one report, older travelers from Australia flew to New York City for a cruise to Montreal, likely carrying the influenza A/Sydney/5/97 virus with them [25]. Although this virus had begun to circulate in the Southern Hemisphere, it had not been detected previously in the Northern Hemisphere. The travelers likely infected cruise ship crew members who then infected at least two subsequent cohorts of travelers on the cruise ship. This report demonstrated the potential for travelers to contribute to the global spread of influenza.

Travel has also contributed to the spread of emerging pandemic influenza viruses. Various factors during the 1918 pandemic demonstrate the role of inter-community connections in maintaining the progression of the pandemic. Large-scale movements of military recruits within the United States, and between Europe and the United States, carried the emerging pandemic strain to other naïve populations [26]. Introduction of the virus to crowded military barracks in the United States and in Europe had a devastating impact on the troops. In the Pacific, ship passengers introduced pandemic influenza to remote islanders with substantial impact; Western Samoa suffered the loss of 19–22% of its population [27,28]. Isolated locations in North America were greatly impacted by introduction of the virus [26]. Native Americans had high case fatality rates, reportedly up to 9%. Remote areas of Alaska worked hard to prevent infection from arriving transport ships, but asymptomatic influenza in sailors on departure from Seattle or Vancouver allowed for introduction in port towns throughout the Alaskan peninsula with devastating outcomes. Even infrequent interaction between fur traders and remote Canadian communities was associated with introduction and spread of pandemic H1N1 in 1918 [29].

Almost a century after the 1918 pandemic, the 2009 pandemic H1N1 (A(H1N1)pdm09) influenza virus emerged and its spread was accelerated by international travel. Soon after the detection of the first two recognized cases of A(H1N1)pdm09 in southern California in the United States, additional cases were rapidly reported from Texas, Chicago, Arizona, and New York [30]. In the first reported case series, 18% of patients with A(H1N1)pdm09 in the United States had traveled to Mexico within 7 days of illness onset [31]. Subsequent travel from Mexico, and onward, especially from the United States, led to the spread of A(H1N1)pdm09 to multiple other countries. Researchers have used statistics of airline travel between Mexico and other countries for the 1 March to 30 April 2008 timeframe to see if the locations of the most frequent destinations predicted sites of earliest recognition of A(H1N1)pdm09 in 2009 [32]. During that time period, 2.35 million people traveled from Mexico to over 1000 cities globally. Countries that had received more than 1400 travelers from Mexico were at a significantly elevated risk of importation of A(H1N1)pdm09. Over 80% of travelers from Mexico flew to the United States and Canada. The three most common destinations were Los Angeles (221 494 people), New York City (126 345), and Chicago (111 531).

For many US college students, and families with children in school, travel to Mexico for spring break provided the opportunity for the virus to infect travelers who then returned home and disseminated the novel virus within their communities. One example from early in the 2009 pandemic is from the University of Delaware [33] where a small number of students traveled to Mexico for spring break and were infected with A(H1N1)pdm09. After returning to school, these travelers introduced the virus to their naïve classmates and provided the virus an opportunity to spread rapidly through the campus over the subsequent two weeks. While the initial number of introductions from travelers was low, transmission of infection increased considerably due to significant person-to-person interaction at Greek Week. This fraternity and sorority event had a number of social and athletic activities which were apparently quite favorable to virus transmission.

Periodically, some events serve to combine extreme crowding and travel, such as the World Cup football match or the Olympic Games. These events have the potential to be a perfect storm for influenza transmission. One of the most dense and traveled mass gatherings is the Hajj, the annual Muslim pilgrimage to Mecca in Saudi Arabia [34]. At its most concentrated point during the one-week event, the crowd compresses to a density of around seven people per square meter as pilgrims circle the kaba sharif. Respiratory infections are common, with influenza frequently identified [35]. In the past, travelers returning from the Hajj have initiated outbreaks of cholera and, most recently, meningococcal meningitis. For past influenza pandemics in 1957 and 1968, the number of visitors was considerably smaller (215 000 and 318 000) and few traveled by air. In 2009, only months after the emergence of A(H1N1)pdm09, over 2.5 million people from 160 countries converged on holy sites in and around Mecca. To minimize the impact of A(H1N1)pdm09, Saudi public health officials screened all arriving travelers for fever and symptoms, and isolated ill persons, established laboratory testing and disease surveillance, utilized treatment centers and clinical algorithms, and encouraged persons at high risk of severe illness to postpone their pilgrimage for a year [36]. Over the summer of 2009, some Muslim countries recommended against attending the Hajj completely. Following the event, surveillance identified no increased illness rates among Hajj attendees and no evidence of significant acceleration of the pandemic attributed to the Hajj [36,37].

Convergence: poultry, pigs, people, and pandemics

The world's population is increasing in megacities, and, for many of these urban residents, the potential for wealth is also increasing. The greater buying power of these residents has led to a rising demand for meat. From the mid-1960s until the mid-1990s, the global consumption of meat rose by 150% [38]. By 2030, the appetite for meat is anticipated to rise an additional 44%; poultry consumption is expected to have the greatest increase. Greater consumption of livestock protein improves overall nutrition and is an efficient food source for urban dwellers [39]. When one considers which source of protein is most efficient for supplying the urban appetite of megacities in the developing world, poultry and pigs are the clear winners; cattle have space and feed requirements that make them less practical. Poultry and pigs, referred to as monogastrics, can be raised in very animal-dense settings and can be more easily transported to urban markets. One additional important characteristic of both poultry and pigs is their role as predominant animal reservoirs of influenza.

To keep pace with the rising population and growing demand for meat protein, the numbers of monogastrics have risen as well. According to the Food and Agriculture Organization of the United Nations (FAO), in 2010, there were an estimated 21.5 billion poultry (91% are chickens), up from 4.3 billion in the early 1960s [40] (Figure 1.1). That equates to around three birds for every person on the planet. There are around 1 billion pigs as well. Over the 40 years from 1967 to 2007, the amount of available protein from poultry meat (grams per person per day) increased 3.4-fold [41]. For pig meat, it increased 1.6-fold. As the population and demand for meat continues to increase in megacities in the developing world, the infrastructure and biosecurity needed to support livestock commerce will need to be drastically altered.

For urban settings in developing countries, where refrigeration is not widely available, poultry and other meat sources must be transported into cities through wholesale and retail live animal markets. Large cities require a constant source of birds from producers in and around the city at large-scale facilities as well as small village farms. This poultry commerce has led to significant increases in urban and peri-urban livestock density and has generated a complicated network of poultry suppliers [42]. Take Jakarta, Indonesia, as an example. It has a population of almost 10 million people who consume around 1 million birds each day [43]. Birds originate from multiple locations on the islands of Java and Sumatra and are carried to Jakarta traveling on over 500 trucks making multiple runs back and forth daily. These birds are handed from wholesaler to retailer, and are made available at over 200 live bird markets in and around the city. A similar situation exists in many other locations around the globe, including Dhaka, Mumbai, Cairo, Lagos, São Paolo, Guangzhou, and many others. In these resource-limited settings, it is challenging to prevent the emergence and spread of avian influenza.

A number of factors have been associated with the emergence, spread, and maintenance of highly pathogenic avian influenza (HPAI) H5N1 viruses in domestic poultry [44]. The cycle of infection first starts with wild migratory birds. Waterfowl are a natural reservoir for low pathogenic influenza viruses, and rivers and lakes, where wild birds congregate with domesticated ducks and geese, promote the exchange of avian influenza viruses. Low pathogenic H5 and H7 viruses may become highly pathogenic as they replicate in domestic birds and, in turn, transmit HPAI to migrating birds that can spread avian influenza viruses across vast distances where domestic poultry, such as chickens, ducks, ostriches, and other domestic fowl, can be infected along the flyways. Once introduced into domestic flocks, other manmade factors lead to further disease: (i) poultry kept in dense conditions where viruses are easily shared among birds; (ii) national highway networks which allow for rapid transport of birds during incubation and infectious periods; (iii) live bird markets which serve as nodes where transmission of HPAI can occur and spread onward; (iv) communities with a high density of homes where people live in close proximity to one another and to flocks of poultry; (v) low literacy rates which are associated with decreased recognition of HPAI and lower rates of reporting infected flocks to authorities; and (vi) practices in low-income communities that depend almost entirely on poultry as a source of protein where diseased poultry are used for their own consumption or are sent to markets.

People and poultry exist in close proximity; however, human infection with avian influenza viruses appears to be rare. Influenza seemingly circulates in parallel, but separate, human and avian epidemics. The same is not the case with pigs. Influenza viruses from birds and people converge in swine as a common basin or mixing vessel where adaptation and reassortment are a frequent occurrence. This convergence is most exemplified at a genetic level with the A(H1N1)pdm09 influenza virus [45]. Immediately after detection of the first two recognized cases of pandemic influenza in April 2009, genetic sequencing revealed that the eight gene segments of the virus were from a mix of avian, human, and swine origin from both North America and Eurasia. The mixed lineage of the virus, evident at a molecular level, is a direct reflection of the co-mingling of poultry, pigs, and people at a macro level in farms and live markets throughout the world. Recent experience with a newly recognized H3N2 variant (H3N2v) influenza virus in the United States further reveals the interconnection of animals and humans and the importance of surveillance in swine [46].

In August 2011, routine surveillance for influenza in Pennsylvania and Indiana in the United States detected two unassociated human cases of infection due to an influenza A (H3N2)v virus with genes from avian, swine, and human influenza viruses [46]. Over the subsequent year, hundreds of additional infections from many states were detected primarily among young children who had frequent and prolonged contact with swine, notably at county and state agricultural fairs. This virus was first detected in swine in November 2010 but had not been detected yet in humans. The virus had two interesting features: (i) the hemagglutinin gene was similar to seasonal H3 genes from viruses that circulated among humans in the early to mid-1990s; and (ii) the matrix (M) gene was the M gene of the 2009 H1N1 pandemic virus. Using available sequence data from swine and human influenza surveillance, the natural history of the virus could be deduced. Human H3N2 viruses had been introduced into the swine population in the 1990s. Gene exchange among viruses in swine produced a reassortant virus between (i) the human H3N2 virus which provided the HA and NA genes; and (ii) a common swine influenza virus containing a group of genes referred to as the Triple Reassortant Internal Gene (TRIG) constellation which provided: the PB1, HA, and NA genes from human influenza origin; the NP, M, and NS genes from swine influenza origin; and the PB2 and PA genes from avian influenza origin. This TRIG-containing virus circulated among pigs since at least 1998. Following the introduction of the 2009 pandemic virus into the swine population in 2009–2010, a further reassortment occurred with the swine H3N2 virus accepting the M gene from the A(H1N1)pdm09 virus to produce the new H3N2v virus which caused hundreds of infections in humans in direct contact with swine in the United States. The H3N2v virus has been shown to be transmissible in ferrets similar to human seasonal viruses [47]. It also appears that people born after the late-1990s have low levels of or no detectable serum antibodies that would be expected to provide cross-protective immunity to the virus. Will this virus take off like the 2009 pandemic H1N1 virus, rapidly traveling to large urban centers around the globe? Even though there appears to be a greater amount of cross-protective immunity in the population than for the 2009 H1N1 pandemic virus prior to its global spread, close monitoring of influenza in human and swine populations is needed to detect any change in virus transmission characteristics, and appropriate swine and human control measures that should be implemented.

Global interconnectedness requires global coordination and response

Global challenges for surveillance

The prevention and control of influenza is impossible without ongoing monitoring of human and animal influenza viruses. Virologic surveillance for human influenza is critical for (i) situational awareness of circulating viruses for directing clinical management; (ii) detection of novel influenza viruses which may indicate emerging and potentially pandemic threats; (iii) acquiring representative influenza viruses for use in vaccines; (iv) monitoring influenza virus resistance to antiviral drugs; and (v) evaluation of the antigenic properties of circulating viruses to detect antigenic drift from currently available vaccines. While it is important to know these characteristics of influenza viruses, many countries have not implemented comprehensive human influenza surveillance programs. A number of challenges face public health authorities: lack of sustainable funding for surveillance programs; training and maintaining epidemiology and laboratory staff; and justifying influenza surveillance in the context of other public health priorities. For example, many countries in sub-Saharan Africa prior to 2007 had no human influenza surveillance at all; many concluded that influenza only rarely was present and causing disease. With additional resources as a part of global pandemic preparedness, surveillance was initiated and revealed unexpectedly high levels of influenza illness and associated child disease burden [48].

Influenza surveillance in humans is patchy, with some regions of the globe covered much more than others. Animal influenza surveillance, on the other hand, is either poor or nonexistent in most countries. Monitoring influenza viruses in poultry and swine is important for determining causes of illness, targeting control measures, developing appropriate animal vaccines to prevent illness and loss of production, and identifying and sequencing genes of emerging influenza viruses with pandemic potential. The number of gene sequences from avian influenza surveillance has not kept pace with the global increase in the population of poultry [49]. Importantly, sequences from swine influenza viruses were difficult to find in 2009 when the pandemic influenza A(H1N1)pdm09 virus emerged. Gene sequences from the first two recognized human cases were posted as soon as testing and analysis had been completed, and comparison of the novel influenza sequences to those from thousands of recent and historic human influenza viruses immediately demonstrated that the illnesses from California were due to a reassortant virus that previously had not been detected in humans. Comparison with available animal virus sequences revealed clear similarity to sequences from swine viruses, but a significant divergence from the relatively few swine influenza sequences that were available for analysis. This significant evolutionary distance between the new pandemic virus and its closest relatives was attributed to a lack of swine surveillance and to the paucity of posted swine influenza sequence data [45]. Although contributions of animal influenza virus sequences have improved recently, most information in publicly available databases are not geographically representative and are often only deposited years after collection of the specimen. However, sequences from swine influenza viruses in the United States have increased substantially during 2011–2012 in response to the emergence of H3N2v [50].

Surveillance of animal influenza viruses is challenging on a number of levels. First, there are resource challenges. For many countries, there may be difficulties prioritizing resources for animal influenza surveillance over other needs. Establishing and maintaining specimen collection systems and laboratory testing infrastructure requires funding and technical expertise that may be difficult to find. Second, competing interests can be a challenge. In many countries, there are longstanding differences in the mission and priorities of human health authorities and agriculture authorities which often lead to constrained communication and limited collaboration. For instance, Ministries of Health (MoHs) prioritize limiting human illness from zoonotic infections and detecting novel influenza virus infections early to prevent, or at least prepare, for potential pandemics. Ministries of Agriculture (MoAgs), however, often promote swine and poultry production for feeding the population or to increase sales and exports. Differences in missions and priorities between these governmental agencies may be difficult to harmonize.

A third challenge for surveillance of influenza in animals is the designation of a virus as intellectual property [51]. As the world gets smaller, and as companies become more global, the protection of proprietary business information and adherence to international patent law is paramount. As major exporters of emerging technologies, Europe and North America have pushed for greater enforcement of intellectual property rights in other countries. Pirating of movies and computer software are considered theft of intellectual property which should be stopped. What about animal influenza viruses? Are they similar to software? In the United States, and in many other countries, the answer is yes. Patents can be granted for genes, proteins, engineered animals, bacteria, nonhuman influenza and other viruses, and just about anything in the life sciences except humans. This is most relevant to influenza prevention and control in the manufacture of vaccines. A novel animal-origin influenza virus may be detected first in South-East Asia, but production and profit from vaccines using the virus may occur among manufacturers in Europe. Whose property is the virus? It is good to have pre-pandemic vaccines, but should not the country where the virus originated also benefit? Alternatively, a veterinary laboratory in the United States or elsewhere may identify a novel H2N3 among swine and consider the virus as a business trade secret with manufacturing potential for vaccinating swine. Profits from use of the virus will allow continued laboratory testing and identification of other viruses; but should not public health have access to the virus and use of the genetic information it contains? When viral genes are considered to be intellectual property, there may be an incentive to delay or withhold public notification of the discovery and details of the virus. Innovation and discovery are the soul of biotechnology and should be protected; however, there is something about influenza viruses that requires a different and more dynamic approach. The influenza virus continually reinvents itself as it travels around an increasingly connected world, and global systems for surveillance and information sharing need to be just as flexible, dynamic, and nimble.

Global regulations for detection and control

Influenza viruses do not respect jurisdictional boundaries. International efforts for mandating common approaches for detecting and responding to influenza epidemics and pandemics are challenging. Starting with cholera in the mid-1800s, global regulations for control of human infectious diseases had various different iterations; however, it was not until 1951 that a formal set of International Health Regulations (IHR) were developed [52]. The IHR were subsequently updated in 1969 and remained unchanged for years. Following eradication of smallpox, the regulations essentially only applied to the traditional quarantinable diseases of cholera, plague, and yellow fever. Over the subsequent decades, the world changed, becoming more crowded and more connected. Notably with the recognition of emerging infections in the 1990s and with severe acute respiratory syndrome (SARS) and avian influenza in the early 2000s, the existing IHR were in need of an update. Global control and response measures needed to be rapid, and existing systems for detecting and reporting important events were inadequate.

In 2005, after years of development and revision by Member States, WHO announced the new IHR which differed dramatically from the 1969 update. The new regulations attempted to address the dynamic nature of emerging infections and the existing resource and political challenges of detection and reporting. One important difference in IHR 2005 was the development of a public health emergency of international concern (PHEIC). In bright distinction to the previous, limited list of three quarantinable pathogens, IHR 2005 required reporting of an extraordinary event which is determined: (i) to constitute a public health risk to other States through the international spread of disease; and (ii) to potentially require a coordinated international response [53]. A PHEIC could be an event within a country's boundaries, but also could be an event that has had spillover into a country from abroad. This part of the new regulation helped to address the issues of emerging infections in an increasingly interconnected world. In addition to the broad event of a PHEIC, the IHR 2005 lists certain diseases for which a case may need to be reported (e.g., cholera, plague, yellow fever). For both the diseases in that list, as well as for the PHEICs, the IHR 2005 provided a decision tool to help Member States determine whether or not an IHR report should be sent to the WHO. The algorithm requires a Member State to ask the following questions: (i) is the public health impact of the event serious; (ii) is the event unusual or unexpected; (iii) is there a significant risk for international spread; and (iv) is there a significant risk for international travel or trade restrictions. If two or more of the questions are answered as yes, then a report of either the PHEIC or the listed disease is required. One important exception to this new framework, however, is that four diseases are required to be reported within 48 hours of their detection without use of the decision algorithm and with no Member State consideration as to whether an IHR is necessary. Those four diseases are smallpox, wild-type poliomyelitis, SARS, and human influenza caused by a new subtype.

What does all this mean for influenza? For one thing, it means that influenza is a central concern for global public health disease control. If a human infection with a new subtype, such as H5N1 or H9N2 occurs, cases must be reported quickly to remain in compliance with the new IHR. But with the addition of the PHEIC concept, there are further implications for the reporting of human infections with other novel influenza viruses. For example, in April 2009, the first two recognized cases of infection with the A(H1N1)pdm09 virus were identified in the United States. This virus did not represent a new subtype of influenza in humans, but was a significantly drifted subtype for which much of the world's population had little to no immunity. Under the expanded definition of an emergency as defined in a PHEIC, the first cases of the 2009 pandemic clearly constituted a public health risk requiring a coordinated international response. For this reason, it was reported immediately as a PEHIC to WHO as required by the IHR. Thus, IHR 2005 provided a framework for reporting and communication that facilitated notification of the world's public health community of the emerging pandemic very early and throughout the course of the event.

Regulations for reporting animal influenza viruses follow a different approach. The World Organization for Animal Health (OIE) maintains a list of notifiable animal diseases to ensure transparency and enhance knowledge about zoonoses around the globe [54]. The FAO enforces international regulations for food security, and requires countries to report the animal diseases on the OIE list [55]. For poultry, this translates to reporting any infection due to avian influenza A H5 and H7 viruses with high or low pathogenicity. When these viruses are detected, there are immediate measures that should be taken, including culling the flock. This focus on H5 and H7 in poultry aligns well with the IHR and human health priorities for preventing illness in people exposed to these birds and for monitoring for the emergence of potentially pandemic viruses. For swine, on the other hand, there is no notifiable influenza disease on the OIE list. Influenza viruses in pigs may cause mild illness or no illness at all. Herein lays a predicament. Human public health authorities want to know when viruses of concern are appearing in swine populations so that pre-pandemic vaccines can be made and diagnostic tests can be optimized for detecting the new viruses. More importantly, human public health authorities are required to submit IHR reports of novel influenza infections to WHO. From the agriculture side, there is no requirement to look for swine influenza, and there is no required intervention if it is detected. Surveillance for swine influenza viruses may, in fact, only serve to frighten the public and lead to decreased pork sales and exports. This disproportionality makes coordinated pandemic preparedness and response very challenging. Any attempt to improve monitoring of novel influenza viruses in swine must address this disparity in regulated reporting between human and agriculture authorities. Unless there is a coincident robust system for detecting swine influenza in pigs, even upgraded regulations to improve reporting of cases are of little value.

Global network for surveillance

Following the discovery of the viral cause of influenza in the 1930s, researchers in London, New York, and elsewhere developed laboratory methods for culturing and characterizing the viruses. As the tools for detecting influenza improved, the potential for monitoring influenza through virologic surveillance was realized. Outbreaks of respiratory disease could be shown to be caused by influenza, and with viruses available for study, vaccines were then possible. Successful efforts to develop an inactivated influenza vaccine in the early 1940s by the US military led to establishment of a small US network of laboratories for isolating the viruses [56]. In 1947, significant antigenic drift in circulating influenza viruses rendered the influenza vaccine ineffective and raised the specter of another pandemic. This event was, in part, the inspiration in 1948 for the WHO Interim Commission to establish the World Influenza Centre in London and propose the organization of a laboratory network that would be coordinated by WHO to monitor influenza viruses infecting humans. By 1949, 38 countries had WHO-designated regional influenza laboratories and it was proposed that all countries eventually might have their own National Influenza Centers (NICs). In 1962, this nascent network of laboratories had grown to include two Collaborating Centers (CCs) for influenza (at the National Institute for Medical Research in London, United Kingdom and the Centers for Disease Control in Atlanta, United States) and 59 NICs. A third CC for the Ecology of Influenza in Animals (Memphis, Tennessee) was initiated in 1976, and by 1984, 108 NICs had been established worldwide. This Global Influenza Surveillance Network (GISN) included five CCs for Influenza (with the addition of CCs at the Victorian Infectious Diseases Reference Laboratory in Melbourne, Australia, and the National Institute for Infectious Diseases in Tokyo, Japan) and 134 National Centers in 2010. In 2011, a sixth CC was established at the Chinese Center for Disease Control and Prevention in Beijing and the name of the WHO network was changed from GISN to the Global Influenza Surveillance and Response System (GISRS) following discussions among Member States at the World Health Assembly. As of July 2012 there were 6 WHO CCs for influenza and 140 NICs in 110 WHO Member States (Figure 1.2).

Figure 1.2. Global Influenza Surveillance and Response System, World Health Organization. Established in 1952, the network currently comprises six WHO Collaborating Centers, four WHO Essential Regulatory Laboratories, and 140 institutions in 110 WHO Member States, which are recognized by WHO as National Influenza Centers, in addition to ad hoc groups established to address specific emerging issues [57]. © WHO 2012. All rights reserved.

c1-fig-0002

In an increasingly connected world, a highly connected influenza laboratory network is necessary. From its inception in 1947, the proposed design for the global network of influenza laboratories was for NICs within countries to collect respiratory specimens from clinical encounters, isolate the viruses in culture, and send the grown viruses of interest to the CCs [56]. This basic model has persisted since that time. Many NICs have been added in the last 5 years, especially an encouraging increase of NICs in Africa [57]. The capability of NICs varies considerably; some have high-throughput capacity and others maintain only limited laboratory testing. WHO recognizes this disparity and works through CCs to provide training, laboratory test guidance, and quality assessment in order to have a more standardized laboratory approach across the network. One way WHO has attempted to achieve lab standardization is through distribution of common protocols and reagents. Since the 1970s, standard reference reagents for antigenic testing have been distributed to each WHO-recognized NIC by the WHO CC at Centers for Disease Control (CDC) in the United States. With the availability of molecular testing using real-time reverse transcription polymerase chain reaction (PCR) assays, CDC also began providing molecular test kits to all NICs in 2008. NIC summary reports for virologic surveillance initially were sent weekly by post to WHO in Geneva, and reporting migrated over time to telex, fax, and finally the internet. Currently, data are collected weekly through the WHO FluNet web site and are made publicly available. The amount of testing and reporting of information increased significantly in 2009 with the emergence of the A(H1N1)pdm09 virus and the availability of PCR reagents. By use of the internet, and through training and use of standardized laboratory testing, GISRS was able to rapidly ramp up to provide near real-time tracking of the 2009 pandemic virus. Since then, over 3 million specimens were tested and reported by NICs, with 1.4 million in 2009, 730 000 in 2010, and 930 000 in 2011.

Viruses and specimens first tested at NICs are sent to one of the five CCs that characterize human influenza viruses. NICs are asked to forward representative virus isolates as well as any low-reacting viruses based on antigenic testing. These viruses represent a subset of all tested specimens and are further characterized at the CCs both antigenically and genetically. Collection of specimens and testing at NICs along with comprehensive laboratory testing at the CCs provides critical information for influenza prevention and control: tracking of antiviral resistance, detection of drifted or novel viruses, measurement of the degree of match between circulating and vaccine viruses, and identification of the best vaccine viruses for upcoming vaccine manufacturing. The WHO recommended vaccine viruses are provided to manufacturers who then produce vaccine for use for the Northern and Southern Hemisphere seasons. The WHO GISRS network of NICs and CCs serves to connect across great distances such that a virus causing illness in an isolated part of the world, such as the Solomon Islands, can become a vaccine preventing illness around the globe.

In the past, viruses were freely distributed and shared among members of the WHO surveillance network and with researchers, vaccine manufacturers, and commercial companies with no restrictions. With the emergence of avian influenza A H5N1 in 2003, and with concerns about the availability of vaccine if a pandemic were to occur, many WHO Member States requested a dialogue to re-evaluate the way the WHO global surveillance system managed the sharing of influenza viruses and the benefits accruing from their use in vaccines and other products [58]. The heart of the issue was that some developing countries were concerned about potential difficulties securing H5N1 vaccine developed from their own viruses. Here was the convergence of people, poultry, pandemics, and property. Countries participating in the WHO surveillance network contributed avian influenza viruses isolated from their citizens, but believed they were receiving no benefit in obtaining affordable vaccine to prevent illness in their countries during a pandemic. Through extended international negotiations, the World Health Assembly addressed this concern and, in 2011, approved the Pandemic Influenza Preparedness (PIP) Framework [58]. The agreement was designed to achieve a fair, transparent, equitable, efficient, effective system for the sharing of H5N1 and other influenza viruses with human pandemic potential and to provide access to vaccines and sharing of other benefits. In particular, the framework requires users of PIP biologic materials which includes human clinical specimens, virus isolates of wild type human H5N1 and other influenza viruses with human pandemic potential; and modified viruses prepared from H5N1 and/or other influenza viruses with human pandemic potential developed by WHO GISRS laboratories, these being candidate vaccine viruses generated by reverse genetics and/or high growth re-assortment [58] to agree to benefit sharing with the countries that provide those materials. In an increasingly connected world, where influenza viruses are shared and transformed into vaccines, the PIP Framework attempts to maintain a lasting connection between virus and benefit sharing for resource-challenged countries. Implementation of this global agreement is only beginning but WHO's GISRS laboratories are instituting new systems for virus tracking. In addition, Industry, Ministries of Health, NICs, CCs, and WHO are working to understand and adhere to the framework. The PIP Framework, IHR, and expanding global laboratory networks are some of many changes occurring with influenza surveillance and control. Many other diagnostic, analytic, and information technologies are available or on the horizon with potential benefits for monitoring outbreaks of concern and response to them.

New opportunities in a changing world

New tools for global detection and surveillance

The recognition of an emerging pandemic and the global response in 2009 was vastly different from efforts for the 1968 H3N2 pandemic and even different from SARS in 2003. The 2009 response took advantage of new molecular technologies, instant information in connected virtual communities, and convergence of multiple data sources to improve forecasting and for focusing interventions. The 2009 pandemic was the first pandemic during the era of molecular diagnostics. Reverse transcription-PCR was first described for use in detecting influenza in 1991, and since then, the method has been gradually adopted and, with standardization of protocols and decreasing costs, has been applied to influenza surveillance. It is now the method of choice in academic medical centers and public health laboratories [59]. The CDC's WHO CC developed a PCR assay to detect influenza A and B, and influenza A subtypes A/H1, A/H3, and A/H5 to support public health surveillance and pandemic preparedness in the United States. The assay was one of the first seasonal influenza PCR assays to be approved by the US Food and Drug Administration, occurring in September 2008, and was subsequently provided to public health laboratories. By April 2009, when the first cases due to the emerging pandemic A(H1N1)pdm09 virus were recognized, 45 US public health laboratories were already using the devices for seasonal influenza surveillance [60]. As part of the pandemic response, an additional 60 US public health laboratories were rapidly added and many NICs in other countries were also equipped. PCR reagents were shipped from CDC to over 150 GISRS laboratories across the globe throughout the response. For many laboratories distant from the North American epicenter of the pandemic, PCR reagents arrived much sooner than cases of pandemic influenza, allowing for early detection and response.

Perhaps the greatest benefit from the molecular miracle has been the use of genetic sequencing to rapidly characterize the viruses for directing public health decision-making. Influenza virus sequence data had been increasingly available in GenBank and other public online repositories; however, with the emergence of A(H1N1)pdm09, sequence entries skyrocketed. The WHO CCs posted thousands of sequences during the first year of the response. Sequence submission from GISRS laboratories and other sources made possible the near real-time monitoring of the virus as it emerged and evolved. Early on, epidemiologic surveillance for hospitalizations and deaths demonstrated great differences in disease outcomes in North America; cases in Mexico appeared to be considerably more severe than in the United States. Fortunately, sequencing of the genes of the H1N1 pandemic influenza viruses demonstrated that the viruses circulating in the two countries did not differ from each other, and neither country's viruses carried any known genetic markers associated with increased disease severity. Genetic sequencing provided the capability to monitor for antiviral resistance, identify the emergence of new mutations associated with more severe outcomes, and track any changes that might represent increasing drift from chosen vaccine viruses. Genetic sequencing and monitoring through GISRS provided a platform for global tracking of the emerging virus, connecting the resources of the broad surveillance community together for pandemic response. Although the 2009 pandemic virus was first detected in humans, it contained genes most closely related to swine influenza viruses. Participation from MoAgs and animal health investigators was less critical during the pandemic response given that the virus was widely circulating in humans. Since then, especially with the increasing prevalence of H3N2v in humans and swine, joint monitoring of swine and human influenza viruses has improved with a considerable increase in the numbers of swine influenza sequences appearing in GenBank.

Looking forward, greater access to rapid sequencing capability will transform influenza diagnosis and surveillance. Given the relatively small genome of the influenza virus, routine sequencing of all eight gene segments will allow diagnostic devices to not only detect the presence of influenza virus, but will concurrently reveal additional information for clinical management, such as known markers for antiviral resistance. For human and animal surveillance, greater routine sequencing will allow public health officials to more easily detect and rapidly characterize emerging novel or antiviral-resistant influenza viruses. Sequence information can be combined with other laboratory and epidemiologic findings to help identify animal influenza viruses most likely to emerge as a pandemic in humans. An example of this convergence of information is the Influenza Risk Assessment Tool (IRAT) being developed by influenza experts which combines 10 evaluation criteria characterizing the properties of the virus, human host susceptibility and pathogenesis, epidemiology of infections, and the ecology of the virus [61]. This tool will be used to prioritize which viruses should be used to make vaccine viruses and whether pre-pandemic lots of vaccine should be produced.

At present, specimens from humans are collected in clinical settings, sent to state or provincial public health laboratories such as GISRS NICs for PCR testing and culture, and forwarded to WHO CCs for sequencing. Implementing sequencing at the hospital or NIC level, and possibly at the point of clinical care, will greatly expand the availability, timeliness, and comprehensiveness of influenza virus genetic data. Nonetheless, even with the promise of next generation sequencing capability, there are some challenges remaining. First, sequencing is not a replacement for culture. Functional viruses are needed for monitoring the antigenic characteristics of the virus and for performing neuraminidase inhibition assays. Second, although sequencing has dropped dramatically in cost and increased considerably in speed and throughput, the data storage and analysis requirements remain daunting. Current sequencing devices generate terabytes of data requiring dedicated servers to support analysis, often taking considerable time to complete. Use of off-site, cloud-based storage and analysis services will be necessary to allow sequencing to be accessible and cost-efficient for influenza detection and surveillance.

Instant and converging information

Just as the world of people, poultry, and pigs is becoming more crowded and connected, the world of information is undergoing similar dynamics. When influenza cases are reported or when genetic and antigenic characterizations of influenza viruses are posted online, cases and viruses essentially become information. Once available on public web sites, the experience from the 2009 pandemic shows that information itself goes viral, spreading and changing as it is transmitted. Information is now instantaneous, abundant, and interlinked. Typing influenza into a commonly used search engine provides about 260 million results in only 0.18 seconds. The 2009 pandemic was the first pandemic, and the first major infectious disease outbreak, to utilize the full potential of the internet and online communities [62]. From the first recognized cases in April 2009, influenza virus sequence information could be accessed virtually from anywhere, anytime, via the internet. Case counts, of varying accuracy, were posted daily on government web sites along with frequently changing recommendations for clinicians, public health officials, and the general public. Online news media sources reported the progression of the pandemic in a 24-hour news cycle. Publications of scientific findings were available online as soon as peer review was completed and were accompanied by continuous threads of commentary and dialogue among experts and others.

These new online tools allowed public health officials to push information as soon as it was available through multiple channels such as RSS feeds, Twitter, Facebook, ProMED, email groups, and other media. The general public was not only passively receiving information, it could also engage or dialogue with public health officials and online information sources: to provide feedback through chat sessions and blogs; to access additional information such as online tools for finding nearest locations to get vaccinated; to monitor indicators of influenza-like illness using internet search terms such as Google Flu Trends; or to conduct online surveys to estimate vaccine coverage among pregnant women. Perhaps the most important information tool from the pandemic was the ability for disparate data to converge in online communities where the public could access free web-based tools for analysis, interpretation, and visualization. Sites such as the Global Initiative on Sharing All Influenza Data (GISAID) endorsed by WHO CCs, or the Influenza Research Database (IRD) and GenBank made available by the US National Institutes of Health [63], brought hundreds of thousands of influenza genome sequences, protein structures, epidemiologic information, and other data together for use by researchers, modelers, vaccine manufacturers, diagnostic test developers, and public health officials. These platforms provided structure and standardization for available information to converge and facilitate influenza prevention and control. Without these tools, the numerous and distributed pockets of available influenza data would be unconnected and unorganized.

Looking forward, connected and converging information may further assist with influenza diagnosis, treatment, prevention, outbreak detection, and pandemic response. As electronic health records are more widely used, surveillance may be greatly facilitated for influenza-like illness, laboratory-confirmed influenza, and influenza-associated hospitalizations and deaths, and may have the potential to be timelier, more representative, and less resource-intensive for public health agencies. Applications on smart phones could provide patients and clinicians with easy to use algorithms for diagnosing influenza, locating sites for antiviral treatment or vaccination, and for connecting to sites for additional information or participating in surveys. After treatment or vaccination, patient follow-up could occur through email, cell phone text messages, Facebook, and other means to remind patients to complete treatment courses and to collect data useful for monitoring adverse events and vaccine effectiveness. Point-of-care and in-home mobile diagnostic devices already connect to the internet for monitoring glucose and hemoglobin A1C results for diabetic patients. Influenza diagnostic test manufacturers are exploring this same approach for clinician offices and pharmacies to diagnose influenza rapidly, upload the results to electronic health records for clinician and patient access online, and send de-identified laboratory results to cloud-based public health agency reporting sites for influenza surveillance. Cell phone coverage is exploding at a rapid pace in resource-limited settings, allowing many of these tools to be accessible in those locations as well. Although there is great promise in the age of information, public health authorities, academic researchers, and others will need to coordinate efforts to assure that information is shared, rapidly available, readily accessible, and well-structured to facilitate prevention and control efforts.

Conclusions

The world of influenza is complex and interconnected. Human influenza viruses carry within them the history of multiple avian, swine, and human gene origins, reflecting a continuous and opportunistic ability of the virus to reinvent itself and reinfect populations. Pandemics, sometimes with tremendous mortality, can arise from changes to influenza viruses circulating back and forth among people and their animals in isolated farming communities and markets around the globe. The opportunities for exchanging viruses between species and for reassortment of their genes have increased as the populations of humans, swine, and birds have increased, all in close proximity to one another in growing urban communities, and living only one flight (and one incubation period) apart from each other.

Influenza is a global challenge, and the prevention and control of influenza requires a commensurate global response. International networks of laboratories and public health agencies have been established, and have also adapted and reinvented themselves by incorporating new technologies, new regulations, and new information sharing tools and platforms. With an increasingly crowded and connected world, influenza viruses have the opportunity to spread and adapt rapidly. To counter that, new molecular diagnostic and genetic sequencing capabilities are available to public health officials, researchers, and clinicians for rapidly determining appropriate treatment, control measures, and prevention strategies. The greatest remaining challenge is maintaining the resources, innovation, and global coordination for responding to influenza viruses.

Acknowledgments

Any views expressed in the Work by contributors employed by the United States government at the time of writing do not necessarily represent the views of the United States government, and the contributor's contribution to the Work is not meant to serve as an official endorsement of any statement to the extent that such statement may conflict with any official position of the United States government.

References

1. World Health Organization. Programmes and projects. Immunization, vaccines and biologicals: influenza. 2008. Available from: http://www.who.int/immunization/topics/influenza/en/index.html (accessed 13 August 2012).

2. Molinari NA, Ortega-Sanchez IR, Messonnier ML, Thompson WW, Wortley PM, Weintraub E, et al. The annual impact of seasonal influenza in the US: measuring disease burden and costs. Vaccine. 2007;25(27):5086–96.

3. McLeod A. The economics of avian influenza. In: Swayne DE, editor. Avian influenza, 1st ed. Ames, IA: Blackwell Publishing; 2008. p. 537.

4. Kilbourne ED. Influenza pandemics of the 20th century. Emerg Infect Dis. 2006 Jan;12(1):9–14.

5. Taubenberger JK, Morens DM. 1918 Influenza: the mother of all pandemics. Emerg Infect Dis. 2006;12(1):15–22.

6. Dawood FS, Iuliano AD, Reed C, Meltzer MI, Shay DK, Cheng PY, et al. Estimated global mortality associated with the first 12 months of 2009 pandemic influenza A H1N1 virus circulation: a modelling study. Lancet Infect Dis. 2012;12(9):687–95.

7. The World Bank. Issue briefs. Animal and pandemic influenza. 2012. Available from: http://go.worldbank.org/E3NZHTSNE0 (accessed 13 August 2012).

8. United Nations Population Fund. State of world population 2011: people and possibilities in a world of 7 billion. 2011. Available from: http://www.unfpa.org/webdav/site/global/shared/documents/publications/2011/EN-SWOP2011-FINAL.pdf (accessed 22 April 2012).

9. United Nations, Department of Economic and Social Affairs. Population division, population estimates and projections section. World Population Prospects, the 2010 Revision. 2011. Available from: http://esa.un.org/unpd/wpp/Documentation/pdf/WPP2010_Highlights.pdf (accessed 22 April 2012).

10. United Nations, Department of Economic and Social Affairs. Population division, population estimates and projections section. World Urbanization Prospects, the 2011 Revision. 2012. Available from: http://esa.un.org/unpd/wup/pdf/WUP2011_Highlights.pdf (accessed 22 April 2012).

11. United Nations Human Settlements Programme. The challenge of slums: global report on human settlements. 2003. Available from: http://www.unhabitat.org/pmss/listItemDetails.aspx?publicationID=1156 (accessed 22 April 2012).

12. United Nations Human Settlements Programme. State of the world's cities, 2006/7. 2006. Available from: http://www.unhabitat.org/pmss/listItemDetails.aspx?publicationID=2101 (accessed 22 April 2012).

13. Central Intelligence Agency, United States of America. World fact book, country comparison, area. Avail­able from: https://www.cia.gov/library/publications/the-world-factbook/rankorder/2147rank.html (accessed 22 April 2012).

14. Banglapedia: National Encyclopedia of Bangladesh. Dhaka division. Available from: http://www.banglapedia.org/HT/D_0157.HTM

15. International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B). Kamalapur 2005–2007 census results. 2008. Available from: https://centre.icddrb.org/images/SP130.pdf (accessed 22 April 2012).

16. Homaira N, Luby SP, Petri WA, Vainionpaa R, Rahman M, Hossain K, et al. Incidence of respiratory virus-associated pneumonia in urban poor young children of Dhaka, Bangladesh, 2009–2011. PLoS ONE. 2012;7(2):e32056.

17. Brooks WA, Goswami D, Rahman M, Nahar K, Fry AM, Balish A, et al. Influenza is a major contributor to childhood pneumonia in a tropical developing country. Pediatr Infect Dis J. 2010;29(3):216–21.

18. Jones KE, Patel NG, Levy MA, Storeygard A, Balk D, Gittleman JL, et al. Global trends in emerging infectious diseases. Nature. 2008;451(7181):990–3.

19. Chang WK. National influenza experience in Hong Kong, 1968. Bull World Health Organ. 1969;41(3):349–51.

20. Eggo RM, Cauchemez S, Ferguson NM. Spatial dynamics of the 1918 influenza pandemic in England, Wales and the United States. J R Soc Interface. 2011;8(55):233–43.

21. Relman DA, Choffnes ER, Mack A. Infectious Disease Movement in a Borderless World. Institute of Medicine. Washington, DC: The National Academies Press; 2010.

22. International Air Transport Association (IATA). Passenger forecast 2007–2011. . Available from: http://www.iata.org/pressroom/pr/pages/2007-24-10-01.aspx (accessed 24 April 2012).

23. UNWTO (United Nations World Tourism Organi­zation). 2008a. Tourism highlights, 2012 edition. Available from: http://mkt.unwto.org/en/publication/unwto-tourism-highlights-2012-edition (accessed 20 April 2012).

24. Russell CA, Jones TC, Barr IG, Cox NJ, Garten RJ, Gregory V, et al. The global circulation of seasonal influenza A (H3N2) viruses. Science. 2008;320(5874):340–6.

25. Miller JM, Tam TW, Maloney S, Fukuda K, Cox N, Hockin J, et al. Cruise ships: high-risk passengers and the global spread of new influenza viruses. Clin Infect Dis. 2000;31(2):433–8.

26. Byerly CR. The U.S. military and the influenza pandemic of 1918–1919. Public Health Rep. 2010;125(Suppl. 3):82–91.

27. McLeod M, Baker M, Wilson N, Kelly H, Kiedrzynski T, Kool J. Protective effect of maritime quarantine in south pacific jurisdictions, 1918–19 influenza pandemic. Emerg Infect Dis. 2008;14(3):468–70.

28. Crosby AW. Samoa and Alaska. America's Forgotten Pandemic: The Influenza of 1918, 2nd ed. Cambridge, UK: Cambridge University Press; 2003. p. 227.

29. Sattenspiel L, Herring DA. Simulating the effect of quarantine on the spread of the 1918 flu in central Canada. Bull Math Biol. 2003;65:1–26.

30. CDC. Update: novel influenza A (H1N1)