Anda di halaman 1dari 26

Data Warehousing for Bioterrorism



Shipra Singal 08BM8055

Piyush Kumar 08BM8048

Ankit Maheshwari 08BM8083

Vinod Gupta School of Management

IIT Kharagpur

This paper introduces the reader to the term Bioterrorism and its
threat on civilian populations. Further, it delves into the need, usage
and application of data warehousing in Bioterrorist attacks and its
surveillance. It also describes the features and technical challenges in
developing an effective bioterrorism surveillance system.
A demonstration of a Bioterrorism surveillance system in the State of
Florida further showcases these ideas.

Survey of Literature

Donald J. Berndt et al [1] in their paper have discussed several technical

challenges in the development of an effective bioterrorism
surveillance system. They have used health care data warehousing
research as the basis to try to resolve these challenges. The difference
between the health care data warehousing and bioterrorism data
warehousing as described in this paper is that surveillance systems for
bioterrorism require more timely data and real-time data warehousing
approaches. They have introduced the concept of flash data
warehousing to compare real-time healthcare data with historical
patterns of key surveillance indicators.

Carol C. Diamond et al in their paper on Health IT systems
emphasize that Technology should enable researchers, practitioners,
and public health officials to share data across networks, while
protecting patients’ privacy. Health information technology (IT) has
great potential to transform health care and inform population health
goals in clinical research, quality measurement, and public safety. But
to fully realize the benefits of health IT for population health, focus
must be on new models that maximize efficiency, encourage rapid
learning. In this paper the authors explore the advantages of a
networked model for analysing population health information and
provide several examples.

Lori Uscher-Pines et al [3] in their research on public health surveillance
systems studied 8 US states. Their objective was to describe current
syndromic surveillance system response protocols and develop a
framework for health departments to use as a guide in initial design
and/or enhancement of response protocols. The research was carried
out by conducting in depth interviews with health department staff.
The conclusion from the research was that health departments had
not prioritized the development and refinement of response protocols
due to reasons like lack of guidance, limited resources for
development of response protocols, and few examples of syndromic
surveillance detecting previously unknown events of public health
significance. The authors have proposed a framework which can guide
health departments in creating protocols that will be standardized,
tested, and relevant given their goals with such systems.

Pascal Cre´pey et al in their paper develop a method of detecting
correlations between epidemic patterns in different regions that are
due to human movement and introduce a null model in which the
travel-induced correlations are cancelled. This method is then applied
to cases of seasonal influenza outbreaks in the United States and
France. This paper basically tells how to interpret data statistically.

Madjid et al in their paper elaborate that natural outbreak of
emerging infections or release of biologic agents during a bioterrorism
attack could have a considerable impact on the cardiovascular systems
of those exposed to the agents. The authors discuss issues

surrounding basic, clinical, and population science research and
training needs with regards to emerging infectious diseases and
biological threats. They describe the need for surveillance systems
that might increase our ability to quickly identify disease outbreaks
and track their course.

Buehler et al in their paper describe that the detection of a
bioterrorism-related epidemic depends on population characteristics,
availability and use of health services, the nature of an attack,
epidemiologic features of individual diseases, surveillance methods,
and the capacity of health departments to respond to alerts. The
authors emphasize that understanding their effect on epidemic
detection should help define the usefulness of syndromic surveillance
and identify approaches to increasing the likelihood that clinicians
recognize and report an epidemic.

The article by Arnold F. Kaufmann et al [7] covers the economic impact

of a bioterrorist attack. It estimates the cost assumes a suburban city
of population 100,000 when attacked by Bacillus anthracis, Brucella
melitensis, and Francisella tularensis. The paper takes into account
the cost of hospitalization based on various parameters like days a
patient need to be in the hospital etc. It also evaluates the economic
preparedness to fight bioterrorism.

The article by Brian C Lein [8] has drawn the attention towards the risk
assessment of bioterrorism. It discusses the detection and prevention
of the population from bioterrorism. It discusses evaluation of threat
and risk assessment. It also recommends consolidating all biological
defense funding, research, and coordination.

The research paper by Howard Kirk Mardis discusses the threat of
bioterrorism and how to make the system more dynamic and efficient
to cater to threats of Bioterrorism. It covers the need to adopt better
information management and human resources systems to fight
bioterrorism. It mentions the trends in bioterrorism and how the
bioterrorist group is different from the rest of the groups. It focuses on
information management and Information structure and how they
help in countering terrorism. The paper mentions the need of special
information managers and information broker to manage information
related to bioterrorism. The paper mentions about analytics and the
human resources also.

The article by Dana A Shea and Sarah A Lister discusses the
biowatch program. It gives insight as what the program is and which
all US cities are covered under it. It discusses the technical issues
about the issues regarding sensors which can detect bioterrorism. It
also stated bioterrorism analytical issues. The paper also covers the
future of biowatch program and also the concerns regarding the policy
which is in place. It mentions the distributed detection networks to be
used under the biowatch program.

The paper by Farzad Mostashari and Jessica Hartman covers the
early warning for bioterrorist attack and bio surveillance. The paper
highlights new uses of analytical techniques. It highlights the

challenges of evaluation of early warning of bioterroism. It also
discusses what all is required to make surveillance to reach its full
potential data standardization, data flow, data security etc.

The paper by Amanda Hodges and Rick Sapp covers the threat of bio
terrorism and agroterrorism to the state of floirda. The paper
examines vulnerability of various industries to bio attacks with special
focus on plant and animal industries. The paper further examines the
national diagnostic network.

The paper by Donald J. Berndt, Sunil Bhat, John W. Fisher, Alan R

Hevner and James Studnicki discusses use of data analytics for
bioterrorism surveillance. The paper also explains the use of catch
data warehouse. Further the paper examines the surveillance system
at the state of florida using data analytics.

The Paper by Fu – Chiang Tsui, Jeremy U Espino, Virginia M. Dato, Per

H. Gesteland, Judith Hutman, and Michael M. Wagner defines the
design and implementation of the Real time outbreak and disease
surveillance (RODS) system, a computer based public health
surveillance system for early detection of disease outbreaks.

The Paper by Parsha Mirhaji examines the current state of the

conceptualization, design, analysis, and implementation of PHS
systems from a translational informatics perspective. The paper
applies concepts from cognitive science and knowledge engineering to
suggest directions for improvement and further research.

The book by Ray R. Arthur, James W. Leduc, James M. Hughes looks

into the various infectious diseases which spread rapidly. It examines
various surveillance and response networks. It also examines the
critical role of the laboratory in surveillance of bio terrorism.

Section 1: Introduction

Bioterrorism and the Threat

Bioterrorism is terrorism by intentional release or dissemination

of biological agents (bacteria, viruses, or toxins); these may be in a
naturally-occurring or in a human-modified form.

According to the US-based Centres for Disease Control and


A bioterrorism attack is the deliberate release of viruses, bacteria, or

other germs (agents) used to cause illness or death in people, animals,
or plants. These agents are typically found in nature, but it is possible
that they could be changed to increase their ability to cause disease,
make them resistant to current medicines, or to increase their ability
to be spread into the environment. Biological agents can be spread
through the air, through water, or in food. Terrorists may use
biological agents because they can be extremely difficult to detect and
do not cause illness for several hours to several days. Some
bioterrorism agents, like the smallpox virus, can be spread from
person to person and some, like anthrax, cannot.

Bioterrorism Agents: A biological agent (or pathogen) is a

disease-causing organism or toxin produced from a biological
source. These can include a bacterium, virus, fungus, or
biological toxin.

Bioterrorism agents can be separated into three categories, depending

on how easily they can be spread and the severity of illness or death
they cause. Category A agents are considered the highest risk and
Category C agents are those that are considered emerging threats for

Category A: These high-priority agents include organisms or toxins

that pose the highest risk to the public and national security because:

 They can be easily spread or transmitted from person to


 They result in high death rates and have the potential for
major public health impact

 They might cause public panic and social disruption

 They require special action for public health preparedness.


1) Anthrax: Anthrax is a non-contagious disease. An anthrax

vaccine does exist but requires many injections for stable use.
When discovered early anthrax can be cured by
administering antibiotics. Anthrax was used in a series of
attacks on the offices of several United States Senators in late

2001. The anthrax was in a powder form and it was delivered
by the mail.
2) Smallpox: Smallpox is a highly contagious virus. It is
transmitted easily through the atmosphere and has a
high mortality rate (20-40%). Smallpox was eradicated in the
world in the 1970s. However, some virus samples are still
available in Russian and American laboratories. As a biological
weapon smallpox is dangerous because of the highly
contagious nature of both the infected and their pox. Also, the
infrequency with which vaccines are administered among the
general population since the eradication of the disease would
leave most people unprotected in the event of an outbreak.
Smallpox occurs only in humans, and has no external hosts or

Category B: These agents are the second highest priority because:

 They are moderately easy to spread

 They result in moderate illness rates and low death rates

 They require specific enhancements of CDC's laboratory

capacity and enhanced disease monitoring.

Example: typhus, brucellosis, cholera

Category C: These third highest priority agents include emerging

pathogens that could be engineered for mass spread in the future

 They are easily available

 They are easily produced and spread

 They have potential for high morbidity and mortality rates and
major health impact

Example: multiple drug-resistant tuberculosis strains

Bioterrorism has become an attractive weapon because biological

agents are relatively easy and inexpensive to obtain or produce, can
be easily disseminated, and can cause widespread fear and panic
beyond the actual physical damage they can cause.

, even though biological warfare itself is as old as the human race.

Over the centuries, there have been many attempts to initiate the
spread of infectious diseases. Recently, the phenomenon of
bioterrorism has gained momentum and bioweapons have become a
threat to civilian populations.

Need for Data Warehousing

Responsibility for bioterrorism acts is rarely claimed by terrorist

groups, as is usually the case with other types of attacks. A covert
attack may thus take a long time to be detected. As the first visible
indication that a bioweapon had been used would be a great number

of patients with similar symptoms and/or presence of an unusual

The effects of the attack will be visible on a number of levels like:

 Physical - actual diseases;

 Psychological - fear, mass panic;
 Economic - travel restrictions, business shut-down;
 Environmental – visible on humans, animals, Plants

To detect warning signs for Bioterrorism attacks, the medical

community should look out for unusual diseases not typically seen in
the area and the general public should constantly be vigilant for

Because of the difficulty in detection, Data warehousing can be used

to facilitate rapid detection of a future bioterrorist attack. Data
warehousing Surveillance systems that target the early manifestations
of bioterrorism-related disease are required. The surveillance efforts
and preparedness towards biowar would be fool-proof if the DW
technique is used and methods of analyzing the situation and data
collection are error-free and prompt.

The goal of these surveillance systems will be to enable earlier

detection of epidemics and a more timely public health response,
hours or days before disease clusters are recognized clinically, or
before specific diagnoses are made and reported to public health
authorities. Establishing a diagnosis is critical to the public health
response to a bioterrorism-related epidemic, since the diagnosis will
guide the use of vaccinations, medications, and other interventions.

Current response to Bioterrorism

To strengthen the area of biodefence, the US senate has passed the

‘Bioterrorism Act of 2002’. According to this law, there is an essential
element of national preparedness against bioterrorism and the focus
is on safety of drugs, food and water from biological agents and toxins.
However, in India such law on bioterrorism is still lacking. In India,
efforts have been put forward by the Department of Defence, through
its arm called the Defence Research and Development Organisation
(DRDO) to develop counter measures against bioterrorism. A few
laboratories of the DRDO are active in this area of research. To
prepare for nuclear, biological and chemical warfare these
laboratories have developed protective systems and equipment for

In 1999, the University of Pittsburgh's Centre

for Biomedical Informatics deployed the first automated bioterrorism
detection system, called RODS (Real-Time Outbreak Disease
Surveillance). RODS is designed to draw collect data from many data
sources and use them to perform signal detection, that is, to detect a
possible bioterrorism event at the earliest possible moment.

In Europe, disease surveillance is beginning to be organized on the
continent-wide scale needed to track a biological emergency. The
system not only monitors infected persons, but attempts to discern
the origin of the outbreak. Researchers are experimenting with
devices to detect the existence of a threat:

 Tiny electronic chips that would contain living nerve cells to

warn of the presence of bacterial toxins (identification of
broad range toxins)
 Fiber-optic tubes lined with antibodies to light-emitting
molecules (identification of specific pathogens, such as
anthrax, botulinum, ricin)

Thus, Bioterrorism poses obvious challenges for civilian authorities.

Such a situation calls for close local co-operation between civilian
authorities from different sectors (public health, law enforcement,
prosecution and customs), national leadership to hold all sectors
together, and international co-operation within the areas concerned.
These could include intelligence services, medicine, the fight against
terrorism and implementation of international conventions.

The health sector should be adequately prepared, that is, not only
stocked with necessary supplies for treatment (vaccines, antibiotics)
but also aware of existing dangers so as to be able to detect a covert
biological attack.

Section 2: Bioterrorism Surveillance System

Developing accurate and reasonable specific surveillance systems

capable of identifying bioterrorism accurately and swiftly is one of the
most active areas of development and research relating to

A surveillance system which can be used for early detection should

include the following

Data Sources: Web sites of relevant government and nongovernment


Data Abstraction: Reports that described or evaluated systems for

collecting, analyzing, or presenting surveillance data for bioterrorism-
related illnesses or syndromes.

Data Extraction: Method of collection, analysis, and presentation of

surveillance data; and outcomes of evaluations of the system

Data Synthesis: The data gathered is reviewed. The systems collecting

syndromic surveillance data and detection system data were designed.
Syndromic surveillance systems have been deployed for both event-
based and continuous bioterrorism surveillance.

Data Sources: The data is gathered from databases of peer-reviewed
articles, government reports, and web sites of relevant government
and commercial entities. Social media sites data is also incorporated.
Government agencies most likely to fund, develop, or use bioterrorism
systems should be identified.. We searched the Web sites of these
government agencies and other academic and commercial sites.
Finally, we identified additional articles from the bibliographies of
included articles and from conference proceedings. Data can also be
gathered from programs like MATRIX (Multi-state Anti-terrorism
Information Exchange system) in USA. . This program helps to facilitate
collaborative information sharing and factual data analysis.

Data Abstraction: System should reviewed titles, abstracts, and full-

length articles to identify potentially relevant articles. An example of
diagnostic decision support systems can be - If we recorded whether
bioterrorism-related illnesses were included in the system's
knowledge base, how the system enabled updates of the probability
of bioterrorism-related illness as epidemic progresses.

Data Extraction: Extraction is the operation of extracting data from a

source system for further use in a data warehouse environment. This
is the first step of the ETL process. After the extraction, this data can
be transformed and loaded into the data warehouse. The source
systems for a data warehouse are typically transaction processing
applications and in this case would be the data source system which
gives the input of the terrorists.

Data Synthesis:
The data gathered from the various sources should be analyzed
through various tools and techniques and appropriate patterns should
be identified. Meaningful conclusions should be derived from the data
gathered for further action.

Surveillance Systems should be installed for monitoring various

scenarios. The systems should be developed for identification of the
scenarios discussed below.

Surveillance Systems for Collecting Environmental Detection Data:

Detection Systems should be deployed which transmit data collected
from environment. These systems differ in the type and location of
sample collected (for example, aerosol samples continuously taken
from locations in fixed sites, such as airports or public buildings If a
peak increase in any chemical is detected, the instrument
automatically collects an sample and alerts the control center.

Surveillance Systems Collecting Clinical Reports: Systems that collect

clinical information from networks of sentinel clinicians. These
systems will keep track of the supplies of some chemicals which can
be used by bioterrorist groups. Monitoring system which will track the
people who are purchasing these chemicals will also be a part of the
surveillance system.

Surveillance Systems Collecting Laboratory Data: Data from the lab

should be collected in order to access what all developments are going
on and in which direction. Potential people who are doing critical
developments on vaccines and biotechnology should be tracked as
these people can develop certain compounds which might be of use
for bioterrorists.

Surveillance Systems Collecting Illness Data: Systems which collect

the data related to illness relating to food items, chemicals or illness
which have been or have the potential to be widespread should be
kept an eye on. Systems should take into account data related to any
incidence which can be a test incidence for bioterrorist group.

Surveillance Systems Collecting Animal Disease Data: Systems which

collect animal disease data should be under surveillance as these data
can be a test data for bioterrorist testing hazardous chemicals.

Evaluation of Reports of Surveillance Systems: All the surveillance

systems need to be integrated. The reports generated by theses
surveillance systems need to be analyzed and proper meaning from
the data needs to be derived.

Section 3: Live case example: State of Florida

The state of Florida has its own surveillance system using data
warehousing to detect early signs of bio terrorist attacks. The
demonstration surveillance system in Florida uses the flash data
warehouse architecture.

The following section provides a brief description of the surveillance

system deployed in Florida

Bio Terrorism Threat Indicators

In any bio terrrorism surveillance system it is most important that the

threats are recognised at an early stage. This will help in controlling
the effect of the attack or the threat.The surveillance system in Florida
uses their expertise in CATCH health care data sets and biological
agents to key bio terrorism threat indicators.

Large Volume of data are collected from the following resources

 Air Quality Monitors

 Water Quality Monitors
 ER Signs and symptoms
 Hospital Admissions
 State Laboratories
 Pharmacy Data
 Practitioners office
Data collected from all the mentioned resources are analysed. They
are coupled with real time data from the hospitals for the
demonstration system. Potential threat indicators that can be defined
based on International Classification of Diseases (ICD) codes and
derived from hospital admission/discharge data.

Real-Time Data Feeds

The initial foci of the demonstration surveillance system will be on

hospital admission discharge data, clinical electronic laboratory
reporting, and existing efforts such as the Merlin and EpiCom
systems. The ongoing community assessment work sup ported by the
CATCH data warehouse already incorporates hospital discharge data
for a variety of health status indicators. In addition, some preliminary
experiments with real-time reporting from local hospitals have been
conducted with great success.

Therefore, the existing hospital admission/discharge data warehouse

components provide a sound foundation for experimentation.

The second area of data collection will focus on electronic laboratory

reporting (ELR).This area of research has received a good deal of
attention, with several successful efforts around the country. The
successful application of the HL7 messaging standards, laboratory test
standards such as Logical Observation Identifiers, Names, and Codes
(LOINC), and vocabularies like the Systematized Nomenclature of
Human and Veterinary Medicine (SNOMED) make this an appropriate
area for the demonstration of surveillance systems.
Pattern Recognition and Alarm Thresholds

One of the most challenging aspects of the demonstration system will

be developing pattern recognition algorithms and defining the nature
of “abnormal” patterns in the selected threat indicators. The early
detection of disease outbreaks is an emerging area of research that
demands “extreme timeliness of detection” for identifying and
responding to public health threats. There has been a surge of interest
in such early warning systems and corresponding attention to some
fundamental questions.
• “Which data are useful for early detection?”
• “What are the timeliness requirements for outbreaks caused by
different agents?”
• “How do we measure timeliness of a detection system for a specific
type of outbreak and especially for outbreaks such as large-scale
inhalation anthrax that have not occurred in areas monitored by the
new systems?”
Solutions to these complex pattern recognition and early detection
problems will only come from sustained research and development.
There is no silver bullet here.

Section 4: Conclusions

This paper identifies bioterrorism and its threats. It classifies the

agents of bioterrorism and establishes the fact that Bioterrorism has
become one of the major challenges of the 21st century. It identifies
the need for data warehousing to tackle bioterrorism. The paper
covers current response to bioterrorism. Factors incorporated in a
surveillance system used to prevent bioterrorism are identified and
systems needed for collecting various sources of data are identified. A
case of state of Florida is included as a live case study at the end of the


1. Donald J. Berndt, Alan R. Hevner, and James Studnicki,

Bioterrorism Surveillance with Real-Time Data Warehousing,
University of South Florida, Tampa, FL 33620
2. Carol C. Diamond, Farzad Mostashari, and Clay Shirky,
Collecting And Sharing Data For Population Health: A New
Paradigm, H e a l t h I T S y s t e m s, Health Aff
(Millwood) March 1, 2009 28:454-466
3. Lori Uscher-Pines, PhD, Corey L. Farrell, MPH, Steven M.
Babin, MD, PhD, Jacqueline Cattani, PhD, Charlotte A. Gaydos,
DrPH, Yu-Hsiang Hsieh, PhD, Michael D. Moskal, MBA, and
Richard E. Rothman, MD, PhD, Framework for the
Development of Response Protocols for Public Health
Syndromic Surveillance Systems: Case Studies of 8 US States,
Disaster Medicine and Public Health Preparedness, 2009, 3
4. Pascal Cre´pey and Marc Barthe´lemy, Detecting Robust
Patterns in the Spread of Epidemics: A Case Study of
Influenza in the United States and France, American Journal
of Epidemiology, Advance Access publication October 15,
2007,Vol. 166, No. 11
5. Mohammad Madjid, MD, Russell V. Luepker, MD, MS, FACC,
FAHA, Co-Chairs Kurt J. Greenlund, PHD, Kathryn A. Taubert,
PHD, FAHA, Michael J. Roy, MD, MPH, FACP, Rose Marie

Robertson, MD, FACC, FAHA, Cardiovascular Effects of
Emerging Infectious Diseases and Biological Terrorism
Threats, March 27, JACC Vol. 49, No. 12, 2007
6. James W. Buehler, Ruth L. Berkelman, David M. Hartley, and
Clarence J. Peters, Syndromic Surveillance and Bioterrorism-
related Epidemics, Emerging Infectious Diseases, Vol. 9, No.
10, October 2003
7. Arnold F. Kaufmann, Martin I. Meltzer, and George P. Schmid,
The Economic Impact of a Bioterrorist Attack: Are Prevention
and Post attack intervention Programs Justifiable, Centers for
Disease Control and Prevention, Atlanta, Georgia, USA
8. Brian C. Lein, A Bioterrorism Prevention Strategy for the 21st
9. Howard Kirk Mardis, Lt Col, Counter Bioterrorism US
intelligence challenges, USAF
10. The Biowatch by Dana A Shea and Sarah A Lister
11. Farzad Mostashari and Jessica Hartman, Syndromic
Surveillance: a Local Perspective
12. Amanda Hodges, Rick Sapp ,The Threat of Agroterrorism and
Bioterrorism in Florida- Prevention and a Coordinated
Response,Florida Department of Agriculture and Consumer
Services, October 2006
13. Donald J. Berndt, Sunil Bhat, John W. Fisher, Alan R Hevner
and James Studnicki ,Data Analytics for Bioterrorism
Surveillance, University of South Florida
14. Fu – Chiang Tsui, Jeremy U Espino, Virginia M. Dato, Per H.
Gesteland, Judith Hutman, and Michael M. Wagner , Technical
description of RODS: a real time public health surveillance

system , Journal of the American Medical Informatics
Association, Volume 10 issue 5 , October 2003
15. Parsa Mirhaji , Public Health Surveillance Meets Translational
Informatics : A Desiderata , Journal of the Association for
Laboratory Automation, Volume 14 Issue 3, June 2009
16. Ray R. Arthur, James W. Leduc, James M. Hughes,
Surveillance for Emerging Infectious Diseases and
Bioterrorism Threats, Tropical Infectious Diseases , 2006