Public Health Surveillance and Infectious Disease Detection

Abstract

Emerging infectious diseases, such as HIV/AIDS, SARS, and pandemic influenza, and the anthrax attacks of 2001, have demonstrated that we remain vulnerable to health threats caused by infectious diseases. The importance of strengthening global public health surveillance to provide early warning has been the primary recommendation of expert groups for at least the past 2 decades. However, despite improvements in the past decade, public health surveillance capabilities remain limited and fragmented, with uneven global coverage. Recent initiatives provide hope of addressing this issue, and new technological and conceptual advances could, for the first time, place capability for global surveillance within reach. Such advances include the revised International Health Regulations (IHR 2005) and the use of new data sources and methods to improve global coverage, sensitivity, and timeliness, which show promise for providing capabilities to extend and complement the existing infrastructure. One example is syndromic surveillance, using nontraditional and often automated data sources. Over the past 20 years, other initiatives, including ProMED-mail, GPHIN, and HealthMap, have demonstrated new mechanisms for acquiring surveillance data. In 2009 the U.S. Agency for International Development (USAID) began the Emerging Pandemic Threats (EPT) program, which includes the PREDICT project, to build global capacity for surveillance of novel infections that have pandemic potential (originating in wildlife and at the animal-human interface) and to develop a framework for risk assessment. Improved understanding of factors driving infectious disease emergence and new technological capabilities in modeling, diagnostics and pathogen identification, and communications, such as using the increasing global coverage of cellphones for public health surveillance, can further enhance global surveillance.

Infectious diseases remain major causes of disease and death worldwide. In addition, recent years have witnessed the appearance of “emerging infections”—that is, infectious diseases that are novel or are rapidly increasing in incidence or geographic range.^1,2 These include such well-known examples as human immunodeficiency virus/AIDS (HIV/AIDS); severe acute respiratory syndrome (SARS), which originated from Asia in 2003; hemolytic uremic syndrome caused by certain strains of the bacterium Escherichia coli; avian influenza H5N1; H1N1-2009 pandemic influenza; and numerous others. In addition, bioterrorism has increasingly become a concern in light of the anthrax mailings in 2001.^3,4 Our vulnerability to such events emphasizes the need for effective surveillance systems to provide early warning. The importance of strengthening public health surveillance has been the primary recommendation of all expert studies over the past 2 decades.^5–8 In the words of a 1992 Institute of Medicine report, Emerging Infections, “The key to recognizing new or emerging infectious diseases, and to tracking the prevalence of more established ones, is surveillance.”^5(p113)

This review discusses our progress toward that goal and considers perspectives for the future. In general, there have been improvements in the past decade, but public health surveillance capabilities remain limited and fragmented, with uneven global coverage.

Several recent initiatives are beginning to address this issue, and new technological and conceptual advances could facilitate reporting and diagnostic capability worldwide. Traditional public health surveillance remains the mainstay, but new nontraditional tools and technologies could complement, or even eventually supplant, traditional public health surveillance. Whatever methods are used, sustained capacity building is essential to achieve this goal.

History and Characteristics of Public Health Surveillance

While the reporting and tabulation of infectious diseases has a long history, Alexander Langmuir at the then–Center for Disease Control (CDC) is generally credited with developing our current concepts of disease surveillance in the 1960s (he was also the originator of CDC's Epidemic Intelligence Service [EIS], which remains active to this day).^9,10 Langmuir suggested that surveillance could be applied to diseases as a public health tool, replacing an earlier notion of surveillance as watching individuals for signs of illness.⁹ He defined surveillance as “the continued watchfulness over the distribution and trends of incidence through the systematic collection, consolidation and evaluation of morbidity and mortality reports and other relevant data.” ^9(p182) The definition currently used by the U.S. Centers for Disease Control and Prevention and widely accepted is similar and elaborates on this concept: “Public health surveillance is the ongoing, systematic collection, analysis, interpretation, and dissemination of data regarding a health-related event for use in public health action to reduce morbidity and mortality and to improve health.”^11(p2) Continuing the interest in surveillance during that time, in 1968 the World Health Assembly (the governing body of the World Health Organization) held technical discussions on “national and global surveillance of communicable diseases,” although the discussions did not lead to implementation plans at that time.^10,12

In general, a surveillance system traditionally requires 3 key elements: (1) a clinical facility or other suitable locale to identify and report affected or exposed individuals, (2) epidemiologic capacity to identify additional cases and determine the source and mode of transmission (and possible interventions), and (3) laboratory capacity to identify the disease agent. In relatively few places are all these elements readily found together, and at the global level the task of developing, linking, and sustaining these capacities becomes extremely challenging.

Originally, and still today, surveillance was disease specific and generally based on identification of the responsible pathogen in the laboratory or on clinical case definitions, usually with subsequent laboratory confirmation of at least some of the cases.^13,14 In this traditional schema, most surveillance systems are passive, requiring someone (usually the proverbial “astute clinician”) to notice a possible disease of interest and to act appropriately by reporting to appropriate authorities and by providing access to the patients and suitable specimens. Surveillance is labor intensive and relatively expensive, but often only limited resources are available. Active surveillance systems, in which interested agencies (such as health departments) make intensive outreach efforts, are particularly resource and labor intensive and therefore less common. Some active systems are short term, when a particular disease condition is of immediate concern, such as during an outbreak.

In all countries, most surveillance systems in use today are still disease-specific systems requiring the identification of specific pathogens or groups of pathogens. In the United States, examples include surveillance for methicillin-resistant Staphylococcus aureus (MRSA); a number of foodborne infections; waterborne infections, in a surveillance system managed jointly by CDC and the Environmental Protection Agency (EPA); HIV; influenza; poliovirus; and many others.^13,14 Internationally, surveillance includes influenza, polio, HIV, and a number of others. At each level, priority will often appropriately be given to diseases of particular concern in that locality, country, or region, or to infections that are targeted by special programs, such as smallpox during the smallpox eradication program or polio today. When appropriately implemented, such systems can be very effective for surveillance of the pathogens of interest, but they require appropriate laboratory capacity and capabilities for specimen collection and transport. At each level, whether local, municipal, state, national, or international, priorities and effectiveness of surveillance for a particular disease will vary with the interests and capacity of those carrying out the surveillance and with the perceived importance of the disease. The problem of comparing and tabulating surveillance data is further complicated by differences in the habits of data sharing among reporting jurisdictions and a lack of standardization in data formats. Finally, these systems are generally not designed to identify diseases or pathogens outside the scope of the system, which can also result in the loss of critical information.

In 2004, Bravata and colleagues conducted an extensive literature review and identified some 115 surveillance systems that could be used for early detection of bioterrorism-related diseases; many of them were systems for public health surveillance. They noted:

Existing surveillance systems for bioterrorism-related diseases vary widely with respect to the methods used to collect the surveillance data, surveillance characteristics of the data collected, and analytic methods used to determine when a potential outbreak has occurred.^15(p910)

Many of the existing surveillance systems have been well described and catalogued in several recent reviews, which are recommended for more detailed information.^14,16–18 Unfortunately, like the assessment by Bravata and colleagues, the picture that emerges is of a fragmented system with a large number and diversity of surveillance systems, many dedicated to special purposes, often utilizing divergent data formats, and very uneven in coverage.^16–18 This fragmentation is often an impediment to timely information exchange.

The field also needs consistent terminology and a more clearly articulated overarching strategy and goals. Some useful definitions have been offered by Hitchcock and colleagues.¹⁶ There have also been recent efforts to develop national strategies for surveillance. In the U.S., in 2007 the White House issued Homeland Security Presidential Directive 21 (HSPD-21),¹⁹ which detailed a National Strategy for Public Health and Medical Preparedness, building on principles set forth in an earlier government document, Biodefense for the 21st Century (April 2004).²⁰ The strategy in HSPD-21 called for the development of a nationwide approach to biosurveillance in order to enhance the U.S.'s ability to detect and respond to biological threats. The Department of Health and Human Services (HHS) was given responsibility for biosurveillance for human health and designated CDC for this purpose.²¹ HSPD-21 defines biosurveillance as

… the process of active data-gathering with appropriate analysis and interpretation of biosphere data that might relate to disease activity and threats to human or animal health—whether infectious, toxic, metabolic, or otherwise, and regardless of intentional or natural origin—in order to achieve early warning of health threats, early detection of health events, and overall situational awareness of disease activity.¹⁹

The report of the National Biosurveillance Advisory Subcommittee reviews the state of biosurveillance in the U.S. and makes recommendations for improvements.²²

Revisions to the International Health Regulations

Another very important step toward an overall global strategy to strengthen surveillance and control diseases at their source is embodied in the revised World Health Organization (WHO) International Health Regulations (IHRs).²³ Originally based on its historical development from 19th and early 20th century concerns, the IHRs formerly required the international reporting of only 4 diseases: cholera, plague, yellow fever, and, until it was eradicated, smallpox. It had been recognized for some time that the regulations were outdated and of limited usefulness in today's highly globalized world.

Consequently, the WHO and World Health Assembly undertook a revision of the IHRs, in an effort led by David Heymann, as a mechanism to strengthen surveillance and response. This process was completed in 2005, and the new regulations, called IHR (2005), entered into force in 2007. According to WHO, the purpose and scope of the IHR (2005) are

to prevent, protect against, control and provide a public health response to the international spread of disease in ways that are commensurate with and restricted to public health risks, and which avoid unnecessary interference with international traffic and trade.²³

The new IHR concept is a major paradigm shift, and the document introduces several important innovations. It replaces the old list of 3 diseases with a broader syndrome-oriented approach (based on case definitions for health events) that would encourage surveillance for both known and previously unknown infectious diseases. In an important innovation, national core capacity requirements for surveillance and response are enumerated. The concept of a “public health emergency of international concern” (PHEIC), requiring reporting within 24 hours, is introduced and defined. Perhaps most important of all, for the first time a decision instrument has been developed to specify criteria for reporting and response. Although there is a need, of course, for further development of decision criteria and triggers for response, these innovations are a major advance. Implementing the new IHR (2005) will require each nation to have a real-time event monitoring system and strengthened surveillance capabilities.²⁴ There will be significant challenges to implementation of this important initiative, even in the U.S.²⁵ Additionally, each country must fund the program from its own resources, and many will require financial help and incentives.

Event-Based Surveillance

Most conventional surveillance systems are unable to identify emerging infectious diseases, which by definition are unexpected and may be caused by unusual or previously unknown pathogens.² Timeliness is also often an issue, especially for rapidly moving outbreaks such as SARS or pandemic influenza. In recent years, approaches have been developed to circumvent these limitations of traditional surveillance. Outbreaks or other occurrences of concern are often referred to as “health events.” Considerable interest in recent years has focused on the detection of health events, rather than specific diseases, as a pathway to more generic and timely surveillance. Some authors now make a distinction between “event-based” and traditional (or “indicator-based”) surveillance.^26,27 The revised IHR document also emphasizes event-based reporting.²³

Event-based surveillance can be viewed as a way of improving sensitivity and timeliness, with the goal of near real-time detection. In the past decade or 2, several very important developments have helped to move this goal closer to reality. These include “syndromic surveillance” (emphasizing the use of nontraditional and automated data sources), the evolution of “digital surveillance” using the internet and other computer-based systems, and the development of new enabling technologies in communications and diagnostics. These categories are not mutually exclusive, and there is considerable overlap.

Syndromic Surveillance

There are many definitions of syndromic surveillance, but most highlight the use of “nondiagnostic” data—that is, information on possible health events before, or without, definitive laboratory identification of the pathogen. Unfortunately, there is confusion about the terminology, as the term had already been used to refer to surveillance based on clinical presentation. Clinical case definitions have long been used in surveillance, particularly for newly recognized diseases before laboratory tests have been developed, and they are used in the revised IHR. This approach has been used successfully in the smallpox and polio eradication programs and proposed for surveillance of emerging infections.¹⁸ Syndromic surveillance, as the term is currently used, by contrast includes a wide variety of other and nontraditional data sources.

There has been an increasing interest in this type of syndromic surveillance, especially since 2001, when an important conference (supported by the Alfred P. Sloan Foundation) led to the formation of the International Society for Disease Surveillance (http://www.syndromic.org/) to advance research and development in the field.

Although there is some general agreement about the data sources and methods that fall under the rubric of syndromic surveillance, definitions are often unclear. As Henning noted, “Specific definitions for syndromic surveillance are lacking, and the name itself is imprecise.”^28(p8) One widely cited definition comes from CDC's document for evaluating public health surveillance systems for early detection of outbreaks; it defines syndromic surveillance as

an investigational approach where health department staff, assisted by automated data acquisition and generation of statistical signals, monitor disease indicators continually (real-time) or at least daily (near real-time) to detect outbreaks of diseases earlier and more completely than might otherwise be possible with traditional public health methods.^29(p2)

Although this often involves automated data collection, which offers obvious advantages, it is not a requirement.

Many municipalities and agencies have piloted syndromic surveillance systems, with many different data sources, including hospital emergency department data, sales of prescription or over-the-counter pharmaceuticals, employee absenteeism, hospital admissions, medical billing or laboratory records, and many others, limited only by ingenuity and data availability.^30–32 The New York City Department of Health & Mental Hygiene, for example, routinely uses emergency department chief complaint data (collected overnight and analyzed early each morning), but it has also collected data from many other sources, including pharmaceutical sales and ambulance dispatch data.³³ In the Washington, DC, metropolitan area, the ESSENCE II system (Electronic Surveillance System for the Early Notification of Community-based Epidemics), which was originally developed for the Department of Defense Global Emerging Infections System (GEIS), combines local military and civilian information and collects such clinical data as emergency department chief complaints, private practice billing codes grouped into syndromes, and veterinary syndromes, as well as employee absenteeism, nurse hotline calls, prescription medications, and over-the-counter medication sales.³⁴

Syndromic surveillance understandably has a number of skeptics, who note, among other valid criticisms, that it has not yet provided advance warning of an outbreak.³⁵ The current situation may be best summarized by Balter and colleagues, based on their experience with emergency department syndromic surveillance for gastrointestinal infections in New York City:

Syndromic surveillance signals occur frequently, are difficult to investigate satisfactorily, and should be viewed as a supplement to, rather than a replacement for, well-maintained traditional surveillance systems that rely on strong ties between clinicians and public health authorities.^36(p175)

While syndromic surveillance shows great promise and may provide valuable information that would be missed by conventional systems, there is still a need for evaluation of these systems to understand which data are best for which situations, how best to interpret these data, and how these sources can be combined to provide a more accurate or complete picture and context. Syndromic surveillance continues to evolve, and its utility as an early warning system should increase with time and experience. It is likely to be especially useful for early warning of unexpected or emerging infections and in those unfortunately all too common situations where the ties between clinicians and public health authorities are not sufficiently strong.

Digital Surveillance

Another major development is the advent of what has been termed “digital surveillance” or “digital disease detection.”³⁷ Although there is no precise definition of the term, it broadly includes the use of the internet and computer technologies for collecting and processing health information, including outbreak reports and surveillance data. These data may come from news reports on the internet, electronic submissions of surveillance data or reports from workers in the field or laboratory, or many other avenues. As such, there is some overlap with syndromic surveillance, although digital surveillance implies data collection through the use of electronic media and may include news reports or other sources that are not part of what would generally be considered syndromic surveillance.

In an attempt to respond to what many saw as the fragmentation of disease surveillance systems and the lack of global capacity, ProMED (the Program for Monitoring Emerging Diseases) was begun in 1993 by a group of scientists, under the auspices of the Federation of American Scientists, as an international follow-up to earlier meetings, especially a 1989 National Institutes of Health (NIH) meeting on emerging viruses and the 1992 Institute of Medicine report.^1,5 At meetings in Geneva and elsewhere, the ProMED Steering Committee recommended developing a system of regional centers to identify and respond to unusual disease outbreaks.³⁸ This could be seen as elaborating on the system D. A. Henderson originally proposed at the 1989 NIH meeting.⁶

The original ProMED concept was for a surveillance network that could provide early warning of both emerging (previously unknown or unanticipated) infections as well as those more familiar. The strategy developed was vigilance for unusual clinical presentations of special concern based on specific case definitions (such as encephalitis or acute respiratory distress with fever in adults); a set of minimum microbiology capabilities at each site, to identify common diseases; and a system to refer unidentifiable samples to successively more sophisticated reference laboratories, through the network, for possible identification.³⁸ The plan also included epidemiologic capacity, which could be provided rapidly through the network if needed.³⁸

It soon became apparent that the 60 or so steering committee members from around the world had no consistent means of communicating with each other. As a result, in 1994, ProMED connected all its steering committee members by e-mail. Hard as it may seem to believe today, in 1994 the internet was only beginning to develop, e-mail was still not widely used, and the World Wide Web we take for granted today was virtually nonexistent, with little publicly available information or news sources. Internet coverage in many parts of the world was so limited that satellite uplinks were required (provided through another nonprofit organization, SatelLife, in Boston) to connect some of the members. The e-mail system, originally envisioned as a direct scientist-to-scientist network, rapidly grew into a prototype outbreak reporting and discussion list, and the decision was made almost immediately to make it publicly available to all at no charge. In October 1999, ProMED-mail became a program of the International Society for Infectious Diseases and has grown along with the internet, currently reaching more than 40,000 subscribers in at least 185 countries. It is now also available in several languages (Portuguese, Spanish, Russian, and French, to support regional efforts in South America, Africa, Russia and the former Soviet Union, and southeast Asia).³⁹ Reports are available in real time by e-mail subscription at no charge, or on the web, where all the reports are also archived (www.promedmail.org). All reports are edited (moderated) by a group of scientists, including public health and infectious disease experts, who vet the reports for scientific plausibility and provide commentary. The context and comments from the moderators (who are subject matter experts) are a unique feature that many subscribers have noted as especially useful, emphasizing the value of contextualizing raw information. A system like ProMED-mail, in addition to outbreak reporting, provides the ability to aggregate and correlate reports, allowing people to recognize that what they are observing locally may be happening in other places as well. Probably the best known outbreak first reported on ProMED-mail was the early report of SARS in China (February 10, 2003).³⁹

ProMED-mail received an unusual compliment from Steven Johnson in his book The Ghost Map, about cholera in Victorian London:

The popular ProMED-mail e-mail list offers a daily update on all the known disease outbreaks flaring up around the world, which surely makes it the most terrifying news source known to man.^40(p219)

With the rapid growth of the internet and the web in the past decade, it is now possible to have reports contributed from much of the world. At the same time, although strides have been made toward the original goal of peri-urban centers with clinical and diagnostic capacity for surveillance, there is still no fully functional global network of regional centers of the sort envisioned by D. A. Henderson or by the original 1996 ProMED proposal.^6,38

Since ProMED-mail was started as an experimental system over a decade ago, it has helped to demonstrate the power of networks and the feasibility of designing widely distributed low-cost reporting systems (“distributed surveillance”), and it has encouraged the development of other systems using additional technologies. These concepts help to begin building the heavily networked surveillance systems that will be needed to deal with public health threats in an increasingly globalized and unpredictable world.

The original use of the internet (through e-mail and later the World Wide Web) in surveillance was to provide disease reports sent in to ProMED-mail by subscribers. The next major innovation harnessed the availability of news sources on the web, a result of the explosion in the web's information content. Since the late 1990s, news sources from around the world, in both English and other languages, have become readily available on the web. Several start-up companies (such as Factiva, started in 1996) were formed as “aggregators,” to collect and index this new flood of information. In 1998, the Canadian government, in an initiative led by Rudi Nowak and Ron St. John at what is now the Public Health Agency of Canada, thought to utilize this strategy for collecting news reports on health events. The result was the Global Public Health Intelligence Network (GPHIN).^41,42 Major sources for GPHIN include Factiva and the Arab language Al Bawaba news service. In November 2004, the second phase of GPHIN, GPHIN II, was launched by the government of Canada and the Washington-based Nuclear Threat Initiative (NTI). Expanding on the original GPHIN, which offered reports in English only, GPHIN II has capacity in 7 languages: Arabic, English, French, Russian, Simplified and Traditional Chinese, and Spanish.⁴¹ It can provide documents in the language of the user's choice, as well as translate articles from English to the other languages and vice versa, and there are plans to expand it further. GPHIN is available to governments, the WHO, and to others by subscription. The automated search algorithms use keywords relating to health events of public health importance to identify the reports of interest. From this subset of aggregated news stories, GPHIN analysts read the reports and identify those that appear of greatest relevance and post them, often after some additional editing, translation, or polishing. A new version, GPHIN III, was in development at the time of this writing.

More broadly, this has now become the dominant strategy for most digital surveillance, and it is used extensively (in some cases exclusively) by most web-based surveillance systems, including ProMED-mail, GPHIN, and HealthMap. ProMED-mail continues to receive information from its traditional sources, such as first-hand reports from the field by clinicians and public health workers, but news reports (selected and edited by the moderators) have increasingly become a dominant information source.

HealthMap, started in 2006 by John Brownstein and colleagues at Children's Hospital Boston and Harvard Medical School, added an additional innovation, the use of geographic information. Each outbreak report is not only listed but pinpointed on a map, so that more precise locations (to the extent that they can be deduced from the original reports) and clusters could be visualized. HealthMap has partnered with ProMED-mail for several years. In addition to ProMED-mail, a major source, HealthMap receives information from several news aggregators such as Google News, Moreover (a commercial service provided by VeriSign), GeoSentinel (a program for sentinel surveillance of individual travelers, operated by the International Society of Travel Medicine and CDC), and others, as well as several official sources.³⁷ HealthMap has also been pioneering other innovations, such as the mobile app “Outbreaks Near Me” or “Flu Near You” (in partnership with the American Public Health Association; https://flunearyou.org/) for Apple iPhone or iPad.

Over the past decade, the WHO developed its own network of networks, originally called GOARN (the Global Outbreak and Response Network), which includes a number of sources both governmental and informal, including GPHIN and ProMED-mail, making a sort of all-source system. An additional salutary effect of the new IHR has been enhancement of the verification and risk management process, through the designation and involvement of national IHR Focal Points.

Cellphones and “Participatory Epidemiology”

The advent of the internet and digital data sources has transformed our ability to collect and report potential health events. But still greater changes are in store. When ProMED-mail started in 1994, it required a satellite connection to reach most of the locations in Africa and Asia. Now, many of these same places have internet access, often broadband. Even more strikingly, many once remote locations in developing countries have mobile phone coverage, which continues to expand explosively. The International Telecommunication Union estimates that, while there were 2 billion internet users (more than half in developing countries), there were 5.3 billion mobile cellphone subscriptions in 2010, almost 75% in developing countries, with about 90% of the world now covered.⁴³

Consequently, surveillance data or disease reports can now be sent from almost anywhere in the world, including field locations or centers that were once totally inaccessible, and the astute clinician no longer need be isolated.^44,45 A number of pioneering efforts, such as Voxiva, had previously used cellphones to send health alerts.⁴⁶ So many people now have mobile phones that reporting can be done by almost anyone on the spot. This has led to an interest in developing “participatory epidemiology.”⁴⁵ Community-based reporting can be done by individuals who are not necessarily medical personnel but are trained to recognize events of interest, and there has been recent interest in self-reporting of disease symptoms either by cellphone or on the web, or through social media.^26,45 The rise of social networking systems has the potential to play an increasingly important role in the future. The challenge will be in using these tools effectively. As with digital disease detection, this revolution in communications technology promises to break down many of the barriers to reporting, but at the risk of increasing the noise level and volume of raw data and the difficulty of verifying reports or establishing representativeness (if needed), much of which is best done by humans interpreting the data. However, there are unparalleled opportunities for detecting health events, whether a sick bushmeat hunter or a wildlife die-off seen by a game warden.

Pathogen Discovery

Pathogen identification, or diagnostic capacity, is often the rate-limiting step in infectious disease detection. Although many bacteria and some other pathogens can be identified by traditional methods such as morphology and growth characteristics, this requires suitably equipped laboratories with trained personnel and can take days or weeks. Viruses can be especially challenging. Until recently, the ability to identify new or uncultivatible pathogens was therefore severely limited. The use of new molecular methods, such as genomic, or “deep,” sequencing, nucleic acid arrays, and various adaptations of the polymerase chain reaction (PCR) for conserved sequences shared by a viral or microbial group, now allow potential identification even of unknown pathogens based on commonalities with known related organisms.^47,48 This open-ended identification of pathogens has been termed “pathogen discovery.” It was made possible by the impressive evolution of both molecular diagnostic methods and biomedical informatics (in part enabled by enormous advances in computing power), providing the ability to separate potential microbial signal from background noise.⁴⁷

Emerging Infectious Diseases and the USAID Predict Project

Despite improvements in recent years, most existing surveillance systems are still unable to identify emerging infectious diseases.^18,49 The very uneven distribution of surveillance capacity in the world is also a major cause for concern.

Most emerging infections are zoonotic introductions from other vertebrate species, including wildlife.^50–53 For example, HIV (or its ancestors) very likely entered the human population from chimpanzees or other nonhuman primates, probably through hunting and butchering of infected animals for meat.^52,54 The ancestor of the virus that causes SARS appears to be a natural infection of certain bats and is thought to have infected humans through mixing of species in live animal markets and food handling practices.^55,56 Despite this, and the longstanding recommendations of expert groups such as the Institute of Medicine, there has not been a global program for surveillance of emerging infections before they reach the human population.¹⁸

In 2009 the U.S. Agency for International Development (USAID) rose to this challenge by starting the Emerging Pandemic Threats (EPT) program, which includes PREDICT, a project to build global capacity for surveillance and risk assessment of novel zoonotic infections that have pandemic potential.⁵⁷ EPT grew out of USAID's earlier programs in avian influenza, which demonstrated the importance of the One Health approach.⁵¹ PREDICT uses the One Health approach to target and integrate surveillance across species, in partnership with governments and agencies such as CDC, FAO, WHO, and OIE. In addition to PREDICT, the EPT program includes RESPOND, a project led by Development Alternatives, Inc. (DAI), to develop training for outbreak investigation and response that merge animal- and human-health approaches, in order to build capacity for disease detection and control; IDENTIFY, which brings together the WHO, the U.N. Food and Agriculture Organization (FAO), and the World Organization for Animal Health (OIE) to develop laboratory networks and to implement strategies for strengthening diagnostic laboratory capacity globally; and PREVENT, led by FHI360, which focuses on behavior change for risk reduction.⁵⁸

PREDICT, the surveillance component of EPT, is intended to build global capacity for infectious disease surveillance, especially for emerging and zoonotic infections. Many of the risk factors, or drivers, of emergence increase pathogen transfer across the interface between humans and other animals (or between animal species). Human activities that can facilitate this process often involve changes in land use or population patterns. These include, among others, farming, hunting, live animal markets, and urbanization.^2,50–52 PREDICT focuses on the animal-human interface to better understand the pathogen background in other species coming into contact with humans and the risk factors for emergence of new zoonoses. Current activities are ongoing in about 20 developing countries. Capacity building, to enable countries to enhance their own surveillance and diagnostic capabilities, is at the heart of the project. Conducted in partnership with national and local governments and in-country scientists and other local personnel, activities include field observation and sample collection, reporting, and both broad viral testing/pathogen discovery and conventional laboratory microbiology as feasible. As of late 2011, the project has already identified approximately 100 viruses (with bats, rodents, and non-human primates the most intensively studied), spanning a number of viral families. A digital data system is being used for storing and correlating the data obtained from these diverse sources.

An innovative and promising approach developed in the past few years is the concept of mapping “hotspots,” or areas that have historically been associated with the emergence of new infections.⁵⁹ There are plans to use data from PREDICT to refine hotspot mapping and modeling strategies and to test hypotheses about zoonotic transmission. Data and technical expertise will be shared appropriately with the participating countries for their own public health planning, and subsequently made publicly available through HealthMap.org on the web.

Future Needs

In the past, surveillance has been highly dependent on diagnostic facilities and health infrastructure. Reporting was often spotty and delayed. In recent years, there have been major improvements both in diagnostic technology and in syndromic and digital surveillance capabilities, greatly increasing the timeliness of reporting.⁴⁹ These systems have less specificity than the traditional systems but can be used to target additional efforts and greatly complement conventional capabilities.

In the meantime, diagnostic capabilities have also markedly improved. Molecular tests, including increasingly widespread use of PCR, array-based (“chip”) technologies, and even genomic sequencing directly from specimens without the need to isolate the pathogen, allow the identification of pathogens from previously unstudied sources, such as many wildlife species, and a better understanding of pathogen background in other species and the environment.^47,48 A number of these technologies are adaptable to resource limited settings.

Improved surveillance for early warning of infectious disease threats, whether in a human population somewhere in the world or waiting in the wings as a potential zoonotic introduction, is therefore both essential and at least conceptually possible. The ability to identify pathogens, at times in mere hours or days, in virtually any species we can test is an impressive improvement in capability, but it also emphasizes the nascent state of our risk assessment ability. Pathogen discovery provides many insights but will also bring many hitherto unknown pathogens to light. It is likely that most of them are relatively well adapted to their hosts and probably would not cause human disease. If we wish to identify potential threats to public health, how to determine which of the many pathogens we will discover in other species are of concern? In other words, how do we predict rather than discover and watch? These are among the issues that surveillance programs, such as PREDICT, will be wrestling with. As capabilities increase, the perennial need to distinguish signal from noise will become increasingly critical.

This requires interpreting the surveillance data, and we are in the early stages of the learning curve. Our shortcomings in predicting avian influenza H5N1 (or, for that matter, influenza generally) exemplify our current state of knowledge. This infection clearly is devastating to poultry farmers' livelihoods and to those people who are unfortunate enough to acquire the disease, usually through contact with infected poultry. As of late January 2012, there have been some 583 human cases, with 344 deaths, in 15 countries.⁶⁰ At one time, there was considerable concern that the virus might evolve to spread efficiently person-to-person and become the next pandemic influenza, but fortunately this has not happened (and may never occur in nature, although recent reports, widely covered in the press, of adapting H5N1 to mammalian transmission in the laboratory have caused concern in the scientific and biosecurity communities). Whether it might happen some day, and if so whether the virus will remain highly virulent in humans, is unknown and, at this point, essentially unpredictable. This is but one of many examples demonstrating our general inability to make accurate predictions about infections from other species. Clearly, surveillance is essential to determine whether there are changes in the behavior of the pathogen. Once an infection is in humans, the surveillance strategies discussed above clearly apply, and epidemiology can be used to determine whether person-to-person transmission is occurring.

Other clues may also be available to help assess risk of a newly discovered pathogen, such as broad host range, or acquiring the ability to infect new host species. The concept of risk mapping, as in the “hotspots” maps discussed earlier, is a promising approach that continues to be refined.⁵⁹

Effective global public health surveillance can also lead to better understanding of the “drivers” of emerging infections. Many disease problems are in areas undergoing major changes in land use or population patterns, including urbanization.² The relative importance of such factors as which family a pathogen belongs to, origins from certain natural hosts or the hosts' phylogenetic relatedness to humans (eg, nonhuman primates), and frequency or intensity of exposure remain open questions.⁵³

An increasing amount of data also means an increasing need for interpretation. Effective surveillance therefore requires human capital. The power of the human mind to make sense of the data is too often taken for granted and overlooked. Whatever the technical advances, there is no substitute for the alert and informed mind, whether the “astute clinician” or an epidemiologist who notices a significant correlation. While true even today, with the more rapidly accumulating information we can expect in the future, it will become even more essential to distinguish signal from noise and be able to place potentially significant events in context. The workforce is therefore of crucial importance. Well-trained professionals, and the need to provide sustainable career paths, are critical now and will be critical in the future.

In addition, surveillance is not enough. Although early warning is the necessary first step, it is not sufficient. For early warning to be useful, it must be backed up by a robust response system. This requires a strong national public health infrastructure in each country, both to foster and coordinate the local and national systems and to provide response capacity. Continued development of triggers and decision criteria, to ensure an appropriate response that is proportionate to the threat, also is essential.

Public health surveillance remains a work in progress. It is clear that the capacity for public health surveillance has markedly improved in the past 2 decades, but many needs remain.^18,61 For example, Feng and colleagues have noted that SARS and avian influenza H5N1 have greatly stimulated improved surveillance in China but that significant gaps remain in the ability to detect emerging infectious diseases.⁶¹ Concern is sometimes expressed whether surveillance systems for emerging or unusual infectious diseases, or for zoonotic infections, will be able to maintain capabilities while preparing for the next SARS outbreak or potential pandemic. But infectious disease outbreaks are occurring frequently all over the world, providing ample opportunities to maintain the capabilities of personnel and equipment. While some infectious diseases may become better controlled in the future, such as some of the vaccine-preventable diseases, it seems highly probable that infections will continue to emerge. The ecological conditions that drive emergence, allowing the introduction and potential dissemination of pathogens across the interface between humans and other animals, and increasing globalization are likely to continue resulting in emerging infections well into the foreseeable future.

The recent development of new and robust technologies for laboratory identification, and revolutionary advances in information technology and communications, enable more rapid reporting of potentially important health events, even from remote areas lacking communications or clinical infrastructure. For the first time, it may now actually be possible to develop global surveillance capabilities of a reasonably high order, even in resource-poor countries. The revised International Health Regulations also set a target for all members states to have the core capacities for surveillance by June 2012. While meeting this objective will probably require more time for most countries, it sets a good aspirational goal and specific plans toward which all countries can work.

With these new and powerful tools, there is great promise for public health surveillance. The greatest concern may well be how to sustain capacity and maintain momentum and political will over the long term.

Footnotes

Acknowledgments

The author gratefully acknowledges support from the Alfred P. Sloan Foundation, USAID Cooperative Agreement GHN-A-009-00010-00 (to University of California, Davis), and the Arts & Letters Foundation.

References

Morse

, Schluederberg

. Emerging viruses: the evolution of viruses and viral diseases. J Infect Dis, 1990; 162,1:1–7.

Morse

. Factors in the emergence of infectious diseases. Emerg Infect Dis, 1995; 1,1:7–15.

Jernigan

, Raghunathan

, Bell

et al. Investigation of bioterrorism-related anthrax, United States, 2001: epidemiologic findings. Emerg Infect Dis, 2002 Oct 8,10:1019–1028.

Hughes

, Gerberding

. Anthrax bioterrorism: lessons learned and future directions. Emerg Infect Dis, 2002 Oct 8,10:1013–1014.

Lederberg

, Shope

, Oaks

Jr . Emerging Infections. Microbial Threats to Health in the United States. Washington, DC: National Academies Press, 1992.

Henderson

. Surveillance systems and intergovernmental cooperation. Morse

. Emerging Viruses. New York and Oxford: Oxford University Press, 1993; 283–289.

World Health Organization. Emerging infectious diseases: memorandum from a WHO meeting. Bull World Health Organ, 1994; 72,6:845–850.

Smolinski

, Hamburg

, Lederberg

. Microbial Threats to Health: Emergence, Detection, and Response. Washington, DC: National Academies Press, 2003.

Langmuir

. The surveillance of communicable diseases of national importance. N Engl J Med, 1963 Jan 24 268:182–192.

10.

Declich

, Carter

. Public health surveillance: historical origins, methods and evaluation. Bull World Health Organ, 1994; 72,2:285–304.

11.

German

, Lee

, Horan

, Milstein

, Pertowski

, Waller

. Guidelines Working Group Centers for Disease Control and Prevention (CDC). Updated guidelines for evaluating public health surveillance systems: recommendations from the guidelines working group. MMWR Recomm Rep, 2001 Jul 27 50,RR-13:1–35.

12.

World Health Organization. Report of the technical discussions at the twenty-first World Health Assembly on national and global surveillance of communicable diseases. 1968 May 18A21/Technical Discussions/5.

13.

U.S. Centers for Disease Control and Prevention. Case definitions for infectious conditions under public health surveillance. MMWR Recomm Rep, 1997 May 2 46,RR-10:1–55.

14.

Lee

, Teutsch

, Thacker

, St. Louis

. Principles and Practice of Public Health Surveillance, 3rd. New York: Oxford University Press, 2010.

15.

Bravata

, McDonald

, Smith

et al. Systematic review: surveillance systems for early detection of bioterrorism-related diseases. Ann Intern Med, 2004 Jun 140,11:910–922.

16.

Hitchcock

, Chamberlain

, Van Wagoner

, Inglesby

, O'Toole

. Challenges to global surveillance and response to infectious disease outbreaks of international importance. Biosecur Bioterror, 2007; 5,3:206–227.

17.

Castillo-Salgado

. Trends and directions of global public health surveillance. Epidemiol Rev, 2010 Apr 32,1:93–109.

18.

Morse

. Global infectious disease surveillance and health intelligence. Health Aff, 2007; 26,4:1069–1077.

19.

The White House. Homeland Security Presidential Directive 21 (HSPD-21). Public Health and Medical Preparedness. October 18, 2007. http://www.fas.org/irp/offdocs/nspd/hspd-21.htm. 2012 January 19.

20.

The White House. HSPD-10. Biodefense for the 21st Century. April 28, 2004. http://www.fas.org/irp/offdocs/nspd/hspd-10.html. 2012 January 19.

21.

U.S. Centers for Disease Control and Prevention. Office of Surveillance, Epidemiology, and Laboratory Services, Public Health Surveillance Program Office. Biosurveillance Coordination (BC) July 25, 2011. http://www.cdc.gov/osels/ph_surveillance/bc.html. 2011 September 15.

22.

National Biosurveillance Advisory Subcommittee. Improving the Nation's Ability to Detect and Respond to 21st Century Urgent Health Threats: Second Report of the National Biosurveillance Advisory Subcommittee. http://www.cdc.gov/about/advisory/pdf/NBASFinalReport_April2011.pdf. 2012 January 19.

23.

World Health Organization. International Health Regulations, 2005 2nd. Geneva: World Health Organization, 2008.

24.

Baker

, Fidler

. Global public health surveillance under new International Health Regulations. Emerg Infect Dis, 2006; 12,7:1058–1065.

25.

Katz

, Kornblet

. Comparative analysis of national legislation in support of the Revised International Health Regulations: potential models for implementation in the United States. Am J Public Health, 2010; 100,12:2347–2353.

26.

Keller

, Blench

, Tolentino

et al. Use of unstructured event-based reports for global infectious disease surveillance. Emerg Infect Dis, 2009 May 15,5:689–695.

27.

World Health Organization. A Guide to Establishing Event-Based Surveillance. Manila: World Health Organization Western Pacific Regional Office, 2008.

28.

Henning

. What is syndromic surveillance? MMWR Morb Mortal Wkly Rep, 2004 Sept 24 53,Suppl:5–11.

29.

Buehler

, Hopkins

, Overhage

, Sosin

, Tong

. CDC Working Group. Framework for evaluating public health surveillance systems for early detection of outbreaks: recommendations from the CDC working group. MMWR Recomm Rep, 2004 May 7 53,RR-5:1–11.

30.

U.S. Centers for Disease Control and Prevention. Syndromic surveillance: reports from a national conference, 2003. MMWR Morb Mortal Wkly Rep, 2004 Sept 24 53,Suppl:1–264.

31.

Mostashari

, Hartman

. Syndromic surveillance: a local perspective [editorial] J Urban Health, 2003; 80,suppl 1:i1–i7.

32.

Das

, Weiss

, Mostashari

et al. Enhanced drop-in syndromic surveillance in NYC following Sept 11, 2001. J Urban Health, 2003 Jun 80,2 Suppl 1:i76–i88.

33.

Heffernan

, Mostashari

, Das

et al. New York City syndromic surveillance systems. MMWR Morb Mortal Wkly Rep, 2004 Sept 24 53,Suppl:23–27.

34.

Lombardo

, Burkom

, Elbert

et al.

A systems overview of the Electronic Surveillance System for the Early Notification of Community-Based Epidemics (ESSENCE II)

J Urban Health, 2003; 80,Suppl 1:i32–i42.

35.

Reingold

. If syndromic surveillance is the answer, what is the question? Biosecur Bioterror, 2003; 1,2:77–81.

36.

Balter

, Weiss

, Hanson

, Reddy

, Das

, Heffernan

. Three years of emergency department gastrointestinal syndromic surveillance in New York City: what have we found? MMWR Morb Mortal Wkly Rep, 2005 Aug 26 54,Suppl:175–180.

37.

Brownstein

, Freifeld

, Madoff

. Digital disease detection—harnessing the Web for public health surveillance. N Engl J Med, 2009 May 21 360,21:2153–21552157.

38.

Morse

, Rosenberg

, Woodall

. ProMED global monitoring of emerging diseases: design for a demonstration program. Health Policy, 1996 Dec 38,3:135–153.

39.

Madoff

, Woodall

. The internet and the global monitoring of emerging diseases: lessons from the first 10 years of ProMED-mail. Arch Med Res, 2005 Nov-Dec 36,6:724–730.

40.

Johnson

. The Ghost Map. New York: Riverhead, 2006.

41.

Mykhalovskiy

, Weir

. The Global Public Health Intelligence Network and early warning outbreak detection: a Canadian contribution to global public health. Can J Public Health, 2006 Jan-Feb 97,1:42–44.

42.

Mawudeku

, Lemay

, Werker

, Andraghetti

, St. John

The Global Public Health Intelligence Network. M'ikanatha

, Lynfield

, Van Beneden

, de Valk

. Infectious Disease Surveillance. Malden, MA: Blackwell, 2007; 304–317.

43.

International Telecommunication Union. ICT Data and Statistics. 2010. http://www.itu.int/ITU-D/ict/statistics. 2011 September 15.

44.

Zamiska

. Fighting bird flu with cellphones; Indonesia goes wireless to overcome reporting delays on cases of deadly disease. Wall Street Journal, 2006 Aug 17 B4.

45.

Freifeld

, Chunara

, Mekaru

et al. Participatory epidemiology: use of mobile phones for community-based health reporting. PLoS Med, 2010 Dec 7 7,12:e1000376.

46.

Prahalad

. Voxiva: health alerts for all. Prahalad

. The Fortune at the Bottom of the Pyramid, Revised and Updated 5th Anniversary Edition: Eradicating Poverty Through Profits. Upper Saddle River, NJ: Pearson Prentice Hall, 2009; 345–356.

47.

Lipkin

. Microbe hunting. Microbiol Mol Biol Rev, 2010 Sep 74,3:363–377.

48.

Relman

. Microbial genomics and infectious diseases. N Engl J Med, 2011 Jul 28 365,4:347–357.

49.

Chan

, Brewer

, Madoff

et al. Global capacity for emerging infectious disease detection. Proc Natl Acad Sci U S A, 2010 Dec 14 107,50:21701–21706.

50.

Daszak

, Cunningham

, Hyatt

. Emerging infectious diseases of wildlife: threats to biodiversity and human health. Science, 2000; 287:443–449.

51.

Karesh

, Cook

. The human-animal link. Foreign Affairs, 2005 July-Aug 84:38–50.

52.

Pike

, Saylors

, Fair

et al. The origin and prevention of pandemics. Clin Infect Dis, 2010; 50,12:1636–1640.

53.

Woolhouse

MEJ

, Gowtage-Sequeria

. Host range and emerging and reemerging pathogens. Emerg Infect Dis, 2005 Dec 11,12:1842–1847.

54.

Keele

, Van Heuverswyn

, Li

et al. Chimpanzee reservoirs of pandemic and nonpandemic HIV-1. Science, 2006 Jul 28 313,5786:523–526.

55.

Guan

, Zheng

, He

et al. Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China. Science, 2003 Oct 10 302,5643:276–278.

56.

, Shi

, Yu

et al. Bats are natural reservoirs of SARS-like coronaviruses. Science, 2005 Oct 28 310,5748:676–679.

57.

University of California, Davis. PREDICT Website. http://www.vetmed.ucdavis.edu/ohi/predict/index.cfm. 2011 September 15.

58.

U.S. Agency for International Development. Emerging Pandemic Threats: program overview. Washington, DC: USAID, February 2010. http://www.usaid.gov/our_work/global_health/home/News/ai_docs/emerging_threats.pdf. 2011 September 15.

59.

Jones

, Patel

, Levy

et al. Global trends in emerging infectious diseases. Nature, 2008 Feb 21 451:990–994.

60.

World Health Organization. Cumulative number of confirmed human cases of avian influenza A/(H5N1) reported to WHO. http://www.who.int/influenza/human_animal_interface/H5N1_cumulative_table_archives/en/index.html. 2012 January 24.

61.

Feng

, Li

, Varma

. Gaps remain in China's ability to detect emerging infectious diseases despite advances since the onset of SARS and avian flu. Health Aff, 2011 Jan 30,1:127–135.