1. Smith, Kathryn McKelvey BSN, MPH, RN, CHPN

Article Content

Key Points


* The tracking of human disease has been implemented since ancient times in order to contain and control its spread in societies, with 19th- and 20th-century scientific discoveries creating the field of epidemiology.


* With the advent of Internet technology in the 21st century, there has been an explosion of capability, not only to track cases but also to create data algorithms to predict trends or ascertain etiologies using epidemic intelligence and biosurveillance system ontologies.


* In light of these new technologies, implications and recommendations for nursing practice are discussed.


Epidemic intelligence (EI) refers to the framework of information that can be used to predict, inform, and verify trends of disease risk, outbreak, and spread. Historically, this gathering of information has been done via local health departments and providers by case reporting, thereby suggesting a trend that can be interrupted for the health good of the population. Because of limits of time and locality, these reports were often of epidemics already in progress. Modern modalities of information sharing therefore lay the basis for EI to become preemptive in disease control.1


In the last 25 years, this historical method has been augmented by the ability to monitor disease outbreak by using interagency databases, the Internet, worldwide news sources, social media, and, of late, biosurveillance ontology software programs. Several generations of such software have provided data collection unprecedented in geographical scope and timeframe. In turn, these data can be analyzed by expanding monitoring systems not only for earlier reporting but also for predictive algorithm creation. Such containment, before tragedy strikes, is the focus of EI and the impetus of current development of biosurveillance modalities.1,2 This article explores the history of epidemiology, current mechanisms in use for disease tracking, and recommendations and implications for the future.



One historical account described how disease control has been practiced since biblical times, with the containment of lepers in colonies to decrease spread. In the Dark and Middle Ages in Europe, plagues were managed by isolating patients and burning the infectious bodies of the deceased. It is acknowledged that Florence Nightingale's sentinel work with improving treatment of the sick during the Crimean War of the mid-1800s was based largely on her well-kept records of the incidence of sickness and associated conditions. By the late 1800s, immigrants had to prove fitness of mental and physical health at Ellis Island to gain access to the United States, thereby limiting the introduction of endemic disease.3


Disease control in the 19th and 20th centuries depended on case reporting by a care provider or laboratory. A drawback to this model was that cases were largely passively identified; that is, they were not always sought out but reported if the patient presented. Still in use today, such control programs may seek to identify and treat cases or have eradication as a goal. An example of the former might be treatment of gonorrhea, and an example of the latter would be smallpox or, more recently, the Ebola virus disease (EVD). Eradication of smallpox was conducted in the mid-18th century by the then-new procedure of the immunization of a "wide circle of contacts" around each positive patient, thus eliminating spread of the disease.4(p1)


Surveillance quality indicators were developed in the late 1980s by the Pan American Health Organization, in the context of assessing the effectiveness of worldwide polio eradication programs. One such indicator was developed to ensure that zero cases really meant zero cases, and their strategy improved specificity and sensitivity in monitoring at this historical point. Since 1994, the Centers for Disease Control and Prevention (CDC) in the US has developed and monitored a variety of surveillance indicators for vaccine-preventable diseases. In order to ensure that zero case reporting is accurate, the CDC continues to monitor case-finding efforts and looks at external standards such as laboratory reports of associated diseases.4


With the attacks in New York and Washington on September 11, 2001, came the realization that terrorism-type attacks could happen on US soil. Biosurveillance needed to expand to include detection of possible biological warfare via intentional infections. Barras and Greub5 stated that bioterrorism dates back some 24 centuries; in addition, in the last 600 years, governments have participated in "the intentional spread of epidemic diseases such as tularaemia, plague, malaria, smallpox, yellow fever, and leprosy."5(p497) Modern scientific advances in manipulation of biologics now allow widespread access to dangerous organisms. Anthrax was used as a bioterrorism tool in World Wars I and II and in September of 2001 when it was shipped through the US mail service.5 Mandates for systems to identify early bioterrorism reports now drive the development of paradigms that could be used to distinguish natural disease outbreaks from those intentionally designed.6



Researchers compared the attributes of three of the best known biosurveillance systems in the world and discussed their optimal use. BioCaster (a Twitter feed) is a system that scans the news sources such as Google, Reuters, the World Health Organization (WHO) site, the CDC site, and the European Media Monitor Alerts. This system scans for multiple languages but prioritizes reports from Asia/Pacific Rim in national languages. A mapping capability allows filtering of results.7 Another application, EpiSPIDER, available as a Twitter feed and an application from sourceforge (, collects from sources such as Google, WHO, and ProMED, but also the social medium Twitter. It contains a mapping/filter feature, timelines and word clouds as well, but scans only for media in English. Finally, HealthMap ( scans in Arabic, Chinese, English, and western European languages from similar news and social media sources. Users can interact with the reports, giving feedback and human intelligence collaboration.7


Researchers found that approximately 10% of reports were redundant and that some lag in reporting time was related to the time zone of the activity. Mapping was compared to see if the systems found similar geographic distributions. There were differences; for example, HealthMap had more activity in Argentina than EpiSPIDER, and BioCaster omitted activity in Bangladesh. Each system has a niche, which, when used in concert, will maximize EI. Unique incident reporting and weighting are therefore features that will be necessary to incorporate when writing accurate detection algorithms.7



Worldwide control and isolation of EI are often contingent on an individual state's compliance. Cultures vary as to their approach to control and their willingness to share information. With international travel, diseases emerging half a world away can now very quickly spread globally. Three uses of technology will be examined with attention to preexisting systems in place, utilization of surveillance, and outcomes obtained.


Tuberculosis and China

The severe acute respiratory syndrome (SARS) outbreak in the early 2000s was a pivotal point in modern disease reporting.1 China's deficiencies in public health reporting came to light as the disease became pandemic with worldwide travel. In response, the Chinese created the national Infection Disease Reporting System, which has been used successfully to control tuberculosis (TB).8 One million active cases of TB are managed each year in China. Prior to the inauguration of the TB Information Management System (TBIMS) in 2005, all tracking and follow-up were done on paper, and loss to follow-up at 2 months after diagnosis was approximately 27%. As TBIMS overtook the paper system, discontinued in 2009, loss to follow-up dropped to 3%. Another improvement seen with use of the tracking system was an increase in the under-24-hour reporting timeframe: 25% in 2005 to 100% in 2011.8


While the development of this surveillance system has been touted as an international example by the WHO, some deficiencies remain. Only respiratory TB is reportable; cases of nonpulmonary TB may not be included in the database. Second, China has approximately 230 million migrants who may not be getting healthcare, as their access is based on province of residence. Finally, not all hospitals interface with the reporting system; only those associated with their CDC have access.8


Ebola and Africa: When Information Is Not Enough

The EVD outbreak in West Africa first was identified as epidemic in March 2014. The National Biosurveillance Integration Center disseminated information to the US Department of Homeland Security, US Department of Health and Human Services, and international agencies. Despite availability of information, the WHO did not declare a state of Public Health Emergency of International Concern until September 2014. What delayed action for 6 months to stop this epidemic, which resulted in a death rate of 40%, or 11 300, of the more than 28 600 cases?9,10


The 2014 EVD epidemic illustrates the need for communication and cooperation, as well as public health infrastructure. Knowledge of outbreak and case reporting is not enough. West African countries were without substantial public health policies, training, and care delivery methods and, when coupled with historic mistrust of Western aid, a containable disease ran rampant. Future outbreaks can be prevented by incorporating lessons learned in all governmental levels in preparation and prevention.9


Vaccine Adverse Event Monitoring

Use of the electronic health record (EHR) is evolving as a data source for monitoring adverse events from new vaccines. In one study, near-real-time vaccination surveillance systems (NRTVSSs) were used to detect seasonal flu vaccine safety problems in very close to real time. The EHR is useful for biosurveillance in modernized Western countries where extensive use of these records is in place. Use of data in near real time is particularly appropriate for seasonal vaccine surveillance as compared with other vaccines. Testing of new seasonal vaccines must be done quickly, as the illness cycle can be completed before the vaccine testing ends. Because of these limitations, improved NRTVSS technology would increase safety and confidence for vaccine consumers.11



German researchers undertook a systematic review comparing event and incident reporting. Incident reporting refers to traditional identification of cases and helps to support known seasonal or geographic data, such as rabies cases per summer in a certain county. Authorities can track cases and make adjustments in intervention activities, such as offering free animal vaccination programs. This traditional system uses the advantages of technology in reporting and is measurable and reliable. Disadvantages include time lag from outbreak to report, resulting in spread of disease while possible cases are overlooked.12


Event-based systems use less conventional media, such as news or social outlets and biosurveillance software. They found that many universities, nongovernmental organizations, and governments used this ontology-based software and included systems such as BioCaster, EpiSPIDER, and ProMED e-mail as discussed previously. Each system also had a different focus such as early detection or improved communication and ranged from free to subscribed access. Advantages seen included faster identification of new or mutant strains of disease. Real-time case identification could also be improved by social media reporting or collecting information from less typical sources.12 Event triggers were highly sensitive, but the difficulty in reliability would be ascertaining which signals were significant. They discovered that while event-based sources were able to identify trends earlier, very few of these sources were incorporated into incident-based surveillance systems. They recommend continued research using event-based data to enhance traditional methods of disease and risk identification.12


A worldwide Early Alerting and Reporting (EAR) project, which includes monitoring for biological, chemical, and radionuclear activity, was shown to have success in an article by a group of international researchers. Started in 2008, this was the first and largest study of its kind involving event surveillance data from multiple systems and with cooperation of international organizations. Over the course of the study, the analysts who received the data were able to incorporate possible bioterrorism components into their usual disease reporting protocols without missing an outbreak or creating a false warning. The EAR created the beginning of a mechanism to monitor both natural and nefarious outbreaks because the latter are relatively rare and could be overlooked.6



The data-information-knowledge-wisdom (DIKW) nursing information model can be used in understanding the nursing implications in disease surveillance science.13 The union of computer science, information science, and nursing science makes the DIKW model ideal for working with biosurveillance. One of the criticisms of the DIKW model is its linearity. Some say the theory ideally operates more as a pyramid than a line. Critics have held that movement between all levels needs to be possible. In discussion of EI, some of these criticisms could be fleshed out in the translation of data to information. For example, increased disease rates would need to be validated through knowledge before they could be included in the information phase, and revisions may need to be made between the DIKW levels as new information is collected through surveillance sources.14


The DIKW model needs clarity in scope of use in order to work to its potential as applied to EI. The e-patient is a participant in tracking his health through informatics, taking crowd sourcing to new levels as one contemplates individual health records being used for a larger, collaborative EI database. People and populations cannot always be reduced to data points, so dedication to the entire model is essential. By the same token, there are data that are not yet associated with people or populations, and these data must have a way of being perceived. Data are abundant and are then categorized into information. Knowledge can be empirical or intuitive, and use of EI requires wisdom and ethics in application.14



Informatics nurses as well as public health nurses and other primary care providers can affect and be affected by information generated from EI and biosurveillance sources. They are in the unique position of using both traditional methods of disease reporting and having access to these newer forms through online state reporting systems. In addition, nurses can educate the public in the appropriate uses of crowd-sourcing data via social media such as Twitter and Google platforms and with regard to Internet security and safeguarding personal health records against unauthorized uses. For example, sources that use advertisements on their Web site could have a bias with regard to integrity in reporting data or news.


Nurses can also educate the public in interpretation of news stories so as to help quell panic without factual basis. Much misinformation accompanies early stories about new diseases, as studies will not have been completed in the beginning of an outbreak. The current Zika outbreak is an example, with early reports of fetal microcephaly not yet proven to be caused by the virus.15 Nurses should base their understanding on research and best practices and model these responses for the public.12


Health promotion and disease prevention are seen as cost-efficient, but incorporating large-scale biosurveillance technology is not without costs. Launching monitoring systems and training people in the data-mining, interpretation, and dissemination processes will be expensive and cumbersome. Parsing information across countries and cultures will also carry financial requirements. Who pays for these systems and who has access to the data are ongoing policy issues to be examined now and in the future.12



An interesting future application is the use of digital drivers outside the traditional fields of medicine and disease surveillance. A group of researchers discussed such a framework as optimizing historically reported public health information with what they term as identification of "antecedent conditions."16(p1285) Drivers can analyze aspects of meteorological, zoological, and social (eg, war) variables, which ultimately could suggest an impending outbreak in disease and include such factors as rainfall, land use, temperature, wildlife migration or obliteration, and lack of public health infrastructure. Because use of digital drivers is in the early phase of development, provision to control false positives and negatives must be made to both mount responses to real outbreaks and prevent panic from rumors.16


For the future, data captured by driver and ontology surveillance would need to be immediately available and able to be depoliticized. Ideally, such knowledge would be used for resource allocation, but sadly, using epidemiological information as political or economic leverage is a possibility. A useful interface should be secure and accepted across cultures and governments and would generate a variety of possible outcomes from myriad scenarios.


Data also need to be reliable and able to be interpreted and trusted by the public. One nurse fell victim to misinformation about EVD when a family member traveled to Africa in the fall of 2014, but not to a country with reported cases of EVD. Work colleagues became alarmed at the news of African foreign travel and demanded quarantine for many weeks from the workplace. After conferring with a CDC state epidemiology field officer who ensured no risk, the nurse relayed the information to the employer. However, the entity in question chose to adhere to these scientifically unfounded requirements. Such hysteria from healthcare workers is doubly damaging, as the public expects sound and evidence-based decisions and policies put forth by knowledgeable professionals.



This article has sought to examine the next phase of technology development in the field of EI. From ancient times, man has sought to contain disease to keep the population safe. Although not new, bioterrorism has wider potential for global destruction with rapid global travel and international delivery systems. New and deadly strains of disease continue to mutate and cause widespread epidemics.


Biosurveillance by ontology software is now being used widely by universities, governments, and world health organizations in order to establish trends that can indicate advanced signs of disease outbreaks. Early warning will allow prompt intervention and decreases in affected cases. Having effective infrastructure in place to accommodate potential epidemics will work in tandem with early identification methods. Deadly epidemics such as EVD can be a thing of the past with both biosurveillance and local health systems' preparation and coordination. Nurses working in public health and informatics will be key participants in using technology for biosurveillance and in educating the public about new modalities such as understanding EI news reports and providing epidemiological data via social media outlets.




1. Collier N. Uncovering text mining: a survey of current work on Web-based epidemic intelligence. Glob Public Health. 2012;7(7): 731-749. [Context Link]


2. Hartley DM, Nelson NP, Arthur RR, et al. An overview of Internet biosurveillance. Clin Microbiol Infect. 2013;19(11): 1006-1013. [Context Link]


3. Lombardo JS, Ross D. Disease surveillance, a public health priority. In: Lombardo JS, Buckeridge DL, eds. Disease Surveillance: A Public Health Informatics Approach. Hoboken, NJ: John Wiley & Sons; 2007: 1-6. [Context Link]


4. Roush SW. Manual for the Surveillance of Vaccine-Preventable Diseases. Centers for Disease Control and Prevention. Published April 1, 2014. Accessed May 4, 2016. [Context Link]


5. Barras V, Greub G. History of biological warfare and bioterrorism. Clin Microbiol Infect. 2014;20(6): 497-502. [Context Link]


6. Riccardo F, Shigematsu M, Chow C, et al. Interfacing a biosurveillance portal and an international network of institutional analysts to detect biological threats. Biosecur Bioterror. 2014;12(6): 325-336. [Context Link]


7. Lyon A, Nunn M, Grossel G, Burgman M. Comparison of Web-based biosecurity intelligence systems: BioCaster, EpiSPIDER and HealthMap. Transboundary Emerg Dis. 2012;59: 223-232. [Context Link]


8. Huang F, Cheng S, Du X, et al. Electronic recording and reporting system for tuberculosis in China: experience and opportunities. J Am Med Inform Assoc. 2014;21(5): 938-941. [Context Link]


9. Hickey A, Barthel SC, Blessington T, et al. An analysis of the international and domestic health hazards posed by the 2014 west African Ebola virus disease outbreak. Online J Public Health Inform. 2015;7(1): e79. [Context Link]


10. Centers for Disease Control and Prevention. Ebola (Ebola virus disease). Published April 14, 2016. Accessed May 4, 2016. [Context Link]


11. Leite A, Andrews NJ, Thomas SL. Near real-time vaccine safety surveillance using electronic health records-a systematic review of the application of statistical methods. Pharmacoepidemiol Drug Saf. 2016;25: 225-237. [Context Link]


12. Velasco E, Agheneza T, Denecke K, Kirchner G, Eckmanns T. Social media and Internet-based data in global systems for public health surveillance: a systematic review. Milbank Quarterly. 2014;92(1): 7-33. [Context Link]


13. Graves JR, Corcoran S. The study of nursing informatics. Image J Nurs Sch. 1989;21(4): 227-231. [Context Link]


14. Ronquillo C, Currie LM, Rodney P. The evolution of data-information-knowledge-wisdom in nursing informatics. Adv Nurs Sci. 2016;39(1): e1-e18. [Context Link]


15. Meaney-Delman D, Hills SL, Williams C, et al. Zika virus infection among U.S. pregnant travelers-August 2015-February 2016. Centers for Disease Control and Prevention. Morb Mortal Wkly Rep. 2016;65: 211-214. [Context Link]


16. Olson SH, Benedum CM, Mekaru SR, et al. Drivers of emerging infectious diseases as a framework for digital detection. Emerg Infect Dis. 2015;21(8): 1285-1292. [Context Link]