Epilogue The Future of Biosurveillance


An author's decision whether to speculate about the future is an example of a "high-stakes" decision taken under uncertainty. There is potential cost to the author, who risks being a source of amusement to future generations. The potential benefit is to the reader, who may be better equipped to allocate her organization's resources, prioritize research, develop a curriculum, or manage her career. So with these caveats (and the prospects of author humiliation) in mind, let us proceed.


Population growth, globalization, urbanization and bioterror-ism have shaped the recent history of biosurveillance. These megatrends and others will also shape its future. The following is a discussion of some of the more important trends. As in the past, advances in science and technology will influence whether the burden of infectious disease in the world increases or decreases, as will the world's ability to incorporate them into biosurveillance practice.

2.1. Population Growth

The U.N. predicts that the world's population will grow by 40% over the next 45 years, from 6.5 billion to 9.1 billion (United Nations Population Division, 2004). This growth will be accompanied by ecological changes such as deforestation that facilitate the emergence of new diseases. Global warming may cause additional ecological changes and natural disasters, which are often accompanied by disease outbreaks. Although improved biosurveillance is not the primary remedy for these problems, absent direct solutions, better biosurveillance will be required if the world is not willing to accept an increase in levels of disease.

2.2. Population Density

The U.N. projects that the 40% growth in the world's population over the next 45 years will occur entirely in urban areas. In 1975, there were only five urban areas in the world with more than 10 million inhabitants, known as "megacities." (For an annual series of distinguished lectures on megacities that began in 1997, see http://www.megacities.nl/.) By 2005, there were 25 (Brinkhoff, 2005). The five largest megacities are the greater Tokyo area with a population of 34.1 million people, Mexico City with 22.7 million, Seoul with 22.3 million, New York-Newark with 21.9 million, and Sao Paulo with 20.2 million. A total of 437 urban areas have populations greater than or equal to one million (Brinkhoff, 2005). See Figure 1.

The significance of this trend toward megacities with increased population density is that the rate at which outbreaks of contagious diseases grow is a function of the size and density of a population. The number of individuals who are exposed to accidental or intentional contaminations of food, water, or air is also a function of the size and density of a population. Better biosurveillance will be required if cities wish to grow without accepting an increase in disease.

2.3. Globalization of Travel, Trade, and the Food Supply

The volume of travel and trade among cities and countries is increasing steadily. The globalization megatrend will increase the number of outbreaks caused by imported disease. It will also amplify the economic impact of such outbreaks. For example, the economic impact of SARS on East Asia was approximately US$18 billion—equivalent to 0.6% of the gross domestic product of the region (Fan, 2003). This impact was disproportionate to its effect on health alone, which comprised approximately 8000 people infected worldwide with 800 deaths. The economic impact of a pandemic of avian influenza could be as high as $US800 billion (World Bank East Asia and Pacific Region, 2005).

At present, the world reacts to diseases such as SARS by curtailing travel and trade. If globalization is to continue at its current pace, the international community must develop biosurveillance systems that not only contribute to the containment of outbreaks before they become pandemics, but also provide real-time, accurate risk information to countries, businesses, and individuals so that their reactions to such events match the actual risk.

2.4. Bioterrorism

The threat of bioterrorism will continue to stimulate research and development of biosurveillance systems. Designers of biosurveillance systems assume that intentional attacks will attempt to avoid detection by biosurveillance systems, thereby creating a blue-team red-team situation in which each side tries to understand and exploit the weakness of the other.

Interestingly, data compiled in the Monterey WMD Terrorism Database—the largest unclassified catalog of worldwide incidents involving the attempted acquisition, possession, threat and use of unconventional weapons by non-state actors—suggests that the frequency of events has been low, that is, there has been a low level of use of biological agents by terrorists and in discovered plots (Figure 2). Note that these figures are for publicly known events and do

Stetten» tu 41

Ctr pzn


UWOH (114) London (12.0) CarolU5:ti kanüut i 11.41

Karachi 114-tj CafcuUa 1,15.&l

Beijing (11:9l

Karachi 114-tj CafcuUa 1,15.&l

figure 2. Actual bioterrorism uses (solid line, squares), plots (dashed line, diamonds), and hoaxes (dotted line, triangles), 1960-2005.The left axis (0-20) is for actual uses of biological agents and plots involving their use, the right axis (0-250) is for hoaxes. There were three additional uses of biological agents in 1933, 1947, and 1952 (not shown). The ten uses in 2001 include multiple anthrax letters. (Data courtesy of Monterey Weapons of Mass Destruction Terrorism Database.)

figure 2. Actual bioterrorism uses (solid line, squares), plots (dashed line, diamonds), and hoaxes (dotted line, triangles), 1960-2005.The left axis (0-20) is for actual uses of biological agents and plots involving their use, the right axis (0-250) is for hoaxes. There were three additional uses of biological agents in 1933, 1947, and 1952 (not shown). The ten uses in 2001 include multiple anthrax letters. (Data courtesy of Monterey Weapons of Mass Destruction Terrorism Database.)

not include events whose existence may be secret. (Public information about unreported plots is important to researchers and to designers of biosurveillance systems because a plot represents a potential outbreak scenario that biosurveillance systems ideally must be capable of detecting.) Readers interested in the details of specific events can request access by writing to [email protected], but note that a .gov or .mil email account or special permission is required for access. There is a published analysis of events in the database through 1998 (Tucker, 2003). It is fortunate that such events have not occurred with frequency as biosurveillance methods are not adequate for these threats at present.

2.5. Information Technology

Advances in information technology may partly offset the negative effects of the above trends by accelerating and improving virtually every step of the biosurveillance process. Computers and networks have increased in speed and storage capacity in an exponential manner for the past 40 years. The approximate rate—a doubling every 12 to 24 months—was predicted by Gordon Moore in 1965 and is referred to as Moore's Law. (Moore's paper is available at ftp://download.intel.com/ museum/research/arc_collect/history _docs/pdf/icfuture.pdf'.)

Figure 3 is Moore's hand-drawn chart. It shows that in 1962 the minimum manufacturing cost per transistor was obtained by fabricating "chips" (integrated circuits) that contained 10 transistors. Three years later, the minimum manufacturing cost per transistor had fallen more than 10-fold and was achieved by fabricating chips containing 100 transistors. By 2004, the Intel Itanium ® 2 processor contained 592 million transistors.

This hardware megatrend is making biosurveillance more cost effective. If Moore's Law holds for another decade—as experts believe it will—it may make the hardware needed for biosurveillance dirt (or at least silicon) cheap.

Although there is no equivalent law that describes trends in the cost of software development, we note that biosurveillance systems are developed by knitting together existing software components such as web browsers and database management systems. These components are developed for other purposes and are available either for free or under license at a minis-cule fraction of the actual development cost. There is little risk of future author remorse in predicting that the final cost of software development for biosurveillance systems will be small relative to the economic impact of SARS or a pandemic of avian influenza.

It is worth noting that advances in information technology are rarely, if ever, driven by biosurveillance, but rather occur for other reasons and must be harnessed by biosurveillance. For this reason, new developments in information technology (and other scientific fields) should be tracked for potential application to biosurveillance. Some developments to track include the following:

• The continuous improvement and proliferation of the

Internet, which will further increase its value as an

figure 3. Moore's graph. In 1965, Gordon Moore sketched out his prediction of the pace of silicon technology. Decades later, Moore's Law remains true, driven largely by Intel's unparalleled silicon expertise. Copyright © 2005 Intel Corporation.

electronic network that connects the myriad organizations and individuals that participate in biosurveillance and enable new types of biosurveillance systems. Trends toward the creation and adoption of paperless medical records, which will result in more and more data that are useful in biosurveillance being captured electronically. Regional health information organizations (RHIOs), which will integrate healthcare data regionally for the purpose of improving the quality of care. RHIOs will represent a single point of access to healthcare data needed for biosurveillance.

Ubiquitous personal devices connected to networks, which will support real-time physiological (e.g., skin temperature, heart rate, and coarse motion) and environmental (e.g., biological agent) monitoring. Consumer products such as cell phones and PDAs can now maintain a trace of all outdoor locations (through GPS and cell-tower communication logs) that their owner visits and when. Newer Bluetooth devices can even record a complete list of when, where, and how long they come in proximity of other devices. These capabilities suggest a possible future of massive parallel real-time contact-tracing. (Examples of these devices and methods can be found at http://www.bodymedia.com/main.jsp, http:// flat-earth.ece.cmu.edu/~eWatch, and http://www.cs.helsinki.fi/ group/context/.)

The development and deployment of new product-tracking technologies such as radio frequency identification (RFID). This technology will enable biosurveillance systems to trace the movement and history of items associated with disease outbreaks.

• Advances in bioinformatics and computer science that will improve algorithms and techniques used to process biological data. These techniques include advances in privacy-protected data mining, which allow organizations to pool data into aggregates needed for large-scale event detection without any participant (including the organization performing the data analysis) being able to infer the individual data contributions.

2.6. Artificial Intelligence

The encoding of human knowledge into computer-interpretable format will play an increasing role in biosurveillance over the next decades. As we discussed in this book, the professionals working in biosurveillance master large bodies of knowledge during their professional lives. The very best of these individuals apply this knowledge to the interpretation of seemingly sketchy clues to achieve startling leaps of insight about the nature and causes of outbreaks. The methods developed by the field of artificial intelligence—including diagnostic expert systems, Bayesian networks, speech recognition, natural language processing, and computational decision theory—are only beginning to encode this knowledge so that it can be used to automate and support the biosurveillance process. To date, only a tiny fraction of the knowledge of physicians, epidemiologists, microbiologists, sanitarians, and others has been represented in computer-interpretable format. This area of research and development has immense potential to increase the speed and efficacy of biosurveillance.

2.7. Diagnostic and Sensor Technology

We are in the midst of a diagnostic revolution in medicine in which rapid testing methods based on DNA and protein identification increasingly allow clinicians (and laypeople, such as in home pregnancy testing) to establish precise diagnoses early in the course of a condition. These technologies will enable the earlier detection of cases of disease at high levels of diagnostic precision. When coupled with electronic data transmission to central locations, they will make possible the detection of outbreaks that are too small or diffuse (spatially or temporally) to be detected at present. Laboratories already employ rapid genotyping of biological agents to determine with high confidence whether a set of individuals were infected by each other or by a common source. Many of these tests are microchip-based and therefore subject to Moore's Law. As the costs of these tests drops, it will become economical to use them for screening and routine monitoring of the microbial ecosystem for emergence of microbial variants that may represent a threat to human or animal populations.

2.8. Molecular Genetics

As the scientific community further elucidates the relationship between genotype and phenotype, there will be an ability to predict the emergence of new diseases and even their characteristics. With time, there will be predictive mathematical models that can answer critical questions such as the probability that the current H5N1 avian influenza will recombine with various human influenza strains and the expected transmissi-bility and virulence of the resulting strains.

Swine Influenza

Swine Influenza

SWINE INFLUENZA frightening you? CONCERNED about the health implications? Coughs and Sneezes Spread Diseases! Stop The Swine Flu from Spreading. Follow the advice to keep your family and friends safe from this virus and not become another victim. These simple cost free guidelines will help you to protect yourself from the swine flu.

Get My Free Ebook


  • rayyan
    What are some of the diseases that biosurveillance would be useful to track?
    6 years ago

Post a comment