Public health in digital cities

Knowledge &
Innovation
Community
EIT ICT Labs
Technical Report TR2011-003
Innovation Radar
Public health in digital cities
December 2011
Public health in digital cities:
Syndromic surveillance
1
Introduction
The sustained migration towards cities, for work and for improved living conditions, places increasingly high demands on urban infrastructures. Cities of the
future will require more efficient means of transportation, equal distribution of
resources such as water and electricity, and robust healthcare networks to accommodate larger populations. To aid the growing numbers of city inhabitants
in collectively asserting their will in issues that affect their life chances, the
cities of the future will also require inclusive means of political participation,
and tools for increasing the transparency of decision making processes [1].
Digital cities of the future is a European Institute of Innovation & Technology (EIT) thematic action line that investigates the future of urban environments, with a focus on the information technologies employed in these settings.
This report focuses on the intersection of the digital city, and an informationintensive form of public health surveillance called syndromic surveillance, where
data gathered for non-diagnostic purposes are used to identify signals relevant
to public health.
Urban surveillance systems can be found in almost any location linked to
consumption: supermarkets, gas stations, shopping centres, cinemas, restaurants, etc. The most common form is a security camera that monitors the
premises. Camera-based surveillance is also present in traffic, to enforce speed
limits, detect red light violations, and control toll roads. The motivation behind
the development of these systems, or of any surveillance system in general, often
focuses on permutations of three issues: efficiency, safety, and convenience [2,
p.96]. In contemporary societies, systems of surveillance take the form of information systems; collecting, storing, and distributing information on populations.
The digital city thematic action line proposes to create new, robust, and in
many cases, invisible information systems that are part of the urban infrastructure. This invisibility is imagined to be a result of the seamless integration of
new information systems with the existing methods of travel, recreation, and
consumption within the city. Currently, technologies such as electronic travel
cards, credit cards, and personal identification cards are used by city dwellers
in their daily activities. Attempts to integrate these systems to combine the
benefits, and increase convenience at the same time, occur often. Examples
include credit cards with photos of their owners that can be used for identification [3], and travel cards with built-in credit [4]. Proposals for new syndromic
surveillance systems can also be situated within the same trend, aiming to take
advantage of increased data collection for public health purposes. The digital
1
city, envisioned as a central hub for data collection activities, fulfils the functional requirements for the employment of syndromic surveillance systems.
1.1
Defining syndromic surveillance
Syndromic surveillance is a recent addition to the set of methods collectively
described as disease surveillance. Disease surveillance is an epidemiological
practice where the activities of a population are monitored for predefined signs.
These signs are then interpreted in an attempt to prevent or minimise the spread
of the disease. Disease surveillance is performed for both communicable diseases (influenza, chlamydia, salmonella, etc.) and non-communicable diseases
(asthma, cancer, diabetes, etc.). The surveillance of the former is called infectious disease surveillance, and it most often involves analysing case reports or
lab reports filed after doctors’ visits. In this analysis, the lab-verified results are
often used as highly accurate indicators of the disease, but the delay between
the onset of symptoms and the verification of the diagnosis through the reports
may be several days to weeks depending on the disease, the diagnosis and the
local infrastructure available to the practitioners. To reduce the delay, data
originally collected for other purposes have been proposed as additional indicators to aid the understanding of infectious diseases. This approach is called
syndromic surveillance, and its practitioners collect and analyse data from different data sources including pre-diagnostic case reports, number of hospital
visits, over-the-counter drug sales and Web search queries, among many other
sources.
One of the most comprehensive and influential definitions of syndromic
surveillance was given by the United States Centers for Disease Control and
Prevention (CDC):
Syndromic surveillance for early outbreak detection is 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 (e.g., by reportable disease surveillance and telephone consultation). The distinguishing characteristic of syndromic
surveillance is the use of indicator data types. [5, p.2]
Practitioners have criticised the usage of syndromic surveillance as imprecise and misleading, because many of the systems described by the term do
not actually monitor syndromes, the association of signs and symptoms often
observed together, but other non-health related data sources such as over-thecounter medication or ambulance dispatches [6, 7]. Despite its shortcomings,
the term remains the most widely recognised among the alternatives. Other
terms that describe similar or equivalent activities include: early warning systems, prodrome surveillance, pre-diagnostic surveillance, outbreak detection systems, information system-based sentinel surveillance, biosurveillance systems,
health indicator surveillance, nontraditional surveillance, and symptom-based
surveillance. Some developers of syndromic surveillance systems have argued
for broader terms, such as biosurveillance, to describe their work, in an attempt
2
to unify outbreak detection and outbreak characterisation, claiming that epidemiologists may consider outbreak characterisation to be separate from public
health surveillance [8, p.3].
3
2
Background
Syndromic surveillance is a recent development in the long history of disease
surveillance. While a full review of developments in disease surveillance is outside the scope of this report, a brief look at current disease surveillance systems
from the perspective of information technologies may be helpful [9].
The unifying property of complex disease surveillance systems as information
and communication systems is that they are designed to function without human intervention, performing statistical analyses at regular intervals to discover
aberrant signals that match the parameters set by their operators. Recent advances in information and communication technology (ICT) have made the development and operation of such systems technically feasible, and many systems
have been proposed to interpret multiple data sources, including those containing non-health related information, for disease surveillance. The introduction
of these systems to the public health infrastructure has been accompanied by
significant criticism regarding the diverting of resources from public health programs to the development of the systems [10, 11], the challenges of investigating
the alerts raised by the systems [6], and the claims of rapid detection [12, 13].
The motivation behind developing complex ICT systems for disease surveillance can be partially explained by the observation that epidemiologists tasked
with monitoring communicable diseases are expected to maintain an awareness
of multiple databases during their daily work. To provide the experts with a
rapid overview of all available data, and to equip them with additional information to make decisions, development of ICT systems are proposed. Efficiently
interpreting the combined output of these systems, however, remains a technical challenge. In many cases, the populations represented in the data sources
monitored by the systems differ significantly, preventing the application of traditional statistical methods to analyse the collected data. In theory, syndromic
surveillance complements traditional disease surveillance in order to increase the
sensitivity and specificity of outbreak detection and public health surveillance
efforts. However, the syntactic and semantic diversity of the syndromic data
sources complicates such efforts.
2.1
Constituents of disease surveillance systems
Conceptually, disease surveillance systems may be partitioned into collection,
analysis, and notification. The collection component contains lists of available
data sources, collection strategies for data sources, instructions for formatting
the collected data, and storage solutions. The analysis component stores a
wide variety of computational methods used to extract significant signals from
the collected data. The final component, notification, contains the procedures
for communicating analysis results to interested parties. The results may be
presented in many forms: numerical output from statistical analysis, incident
plots displaying exceeded thresholds, maps coloured to indicate different levels
of observed activity, or simply as messages advising the experts to check a data
source for further information.
4
2.1.1
Collection
The set of accessible data sources is the most important factor in determining the
capabilities of a disease surveillance system. Once again, following the growth
of syndromic surveillance, a wide variety of sources have been proposed for
monitoring. They may be divided into three groups based on when they become
visible to the system relative to the patients’ status: pre-clinical, clinical prediagnostic, and diagnostic [14]. An alternative method of categorising the data
is to group according the type of patient behaviour that produces the data:
information seeking after onset of symptoms; care seeking where the patient
attempts to contact the healthcare provider or decides to purchase medication;
and post-contact, when the patient becomes visible in traditional public health
surveillance systems. Data sources most often used for syndromic surveillance,
ordered by availability, from earliest to latest, are as follows [13, 15]:
•
•
•
•
•
•
•
•
•
•
•
•
over-the-counter drug sales
triage nurse line calls
work and school absenteeism
prescription drug sales
emergency hotline calls
emergency department visit chief complaints
laboratory test orders
ambulatory visit records
veterinary health records
hospital admissions and discharges
laboratory test results
case reports
In a survey of operational syndromic surveillance systems [16], it is reported
that the 52 respondents monitor the following data sources: emergency department visits (84%), outpatient clinic visits (49%), over-the-counter medication
sales (44%), calls to poison control centres (37%), and school absenteeism (35%).
Another review by [17] examines 56 systems and presents a comparable distribution of data source usage.
The availability of data sources depends on the local context of the project:
jurisdiction of the organisation responsible for the system, diagnoses to be monitored, existing laws regulating data access, and technical concerns such as ensuring sustained connectivity to the data sources. Recent research suggests that
additional sources such as Web search queries [18, 19], and Twitter posts [20]
can also contain indicators for disease surveillance.
The timeliness of a data source is often inversely proportional to its reliability [14]. Sources with immediate availability such as Twitter posts or search
queries often contain large amounts of false signals, and usually lack geographic
specificity. In contrast, laboratory test results provide definitive diagnostic information, but they are not available early. An example between the two extremes is chief complaint records from emergency departments. These records
are available on the same day as the visit, contain specific signs and symptoms
as well as geographic information, but initially lack diagnoses [21].
5
2.1.2
Analysis
In traditional disease surveillance systems, the data forwarded by the collection component is associated with a diagnosis directly, and analysis begins. In
syndromic surveillance systems, the data may contain signals for multiple diagnoses. Therefore, every data stream is assigned a syndrome category before it
can be investigated for statistically significant signals. Syndrome categories are
lists of signs and symptoms that indicate specific diseases; examples include respiratory, gastrointestinal, influenza-like, and rash. The assignment proceeds in
two steps: first, the information relevant to the categorisation is extracted from
the collected data, and second, the extracted information is used to associate
the data with the syndrome category. The extraction procedure is trivial for
pre-formatted data sources such as over-the-counter drug sales (the data are already categorised by drug type), but may require complex methods for free-text
data sources such as emergency department chief complaints. The extracted
information is then associated with one or more syndrome categories, either by
using a static mapping of data sources to diseases, or by an automated decisionmaking mechanism. An example of the former is CDC’s syndrome categories
in BioSense, and of the latter, BioStorm’s ontology-driven assignments. Both
systems are described in more detail in the next section. In existing syndromic
surveillance systems, Bayesian, rule-based, and ontology-based classifiers have
been used to assign syndrome categories [17, p.53].
After the categories are assigned, statistical analysis is used to detect significant signals. These signals may be short-term changes such as sharp increases
or decreases in the number of cases, indicating emerging outbreaks or effects
interventions; or long-term shifts, indicating the appearance of the disease in
previously unaffected age groups or geographical regions. The literature on statistical analysis of disease surveillance data is vast, and interested readers are
recommended to refer to previously published reviews [22, 23, 24, 25, 26] for a
more thorough analysis of existing methods.
Most of the algorithms used in disease surveillance are adapted from other
fields such as industrial process control or econometrics [27], but some have
been developed specifically for disease surveillance. Time series methods, meanregression methods, auto-regressive integrated moving average (ARIMA) models, hidden Markov models (HMMs), Bayesian HMMs, and scan statistics [28]
are among the most commonly used algorithm classes. The spatial and spacetime statistics, specifically, have gained popularity among practitioners as methods for the detection of disease clusters [29].
The detection algorithm is chosen based on the needs of the users, and
the available data sources. To aid the decision, the performance of aberrancydetection algorithms are often expressed in terms of sensitivity, specificity, and
timeliness [30]. Sensitivity (true positive rate) is the probability that an alarm
is raised given that an outbreak occurs. Specificity (true negative rate) is the
probability that no alarm is raised given that no outbreak occurs. Timeliness is
the difference in time between the event and the raised alarm. Additionally, to
be able to understand and describe the detection algorithms better, researchers
have proposed a classification scheme algorithms that considers the types of
information and the amount of information processed by the algorithms: number of accessible data sources, number of covariates in each source, and the
availability of spatial information [31].
6
2.1.3
Notification
The results of the analyses are visualised and communicated to the users by
the notification component. The simplest method of communication is by providing the statistical output from the analysis method directly to the user, as
a table or in plain text. Although the output contains all the essential information, understanding these reports is often time-consuming, and they quickly
become overwhelming if many data sources, diagnoses, or geographic regions
are involved in the analysis.
The most common way of summarising the results is by displaying their
values at different time points, using line charts or bar graphs. The variable may
be case reports, ambulance dispatches, drug sales, or any other indicator used in
surveillance. If previously computed historical baselines exist, they may also be
plotted on the same graph, to put the current results in a larger context. Scatter
plots and pie charts may also be used to summarise non-temporal components
of the analysis.
Figure 1: Sample output graph from an outbreak detection system in use at the
Swedish Institute of Communicable Disease Control.
When a spatial analysis method is used, the same variable may be displayed
using a map where colours, shades, or patterns illustrate the differences between
geographical regions [32]. Results of clustering methods, such as spatial scan
statistics, may also be visualised on maps, often using geometric shapes or grids
drawn on the map in addition to the regional borders [33]. Visualising the results
7
of hybrid spatio-temporal analysis methods may be achieved by animating the
map, or presenting snapshots from the same map at different time points sideby-side.
Alternatively, Geographic Information Systems (GIS) may be used to visualise spatial or spatio-temporal analysis results if the data source contains
detailed geographical information. These systems are commonly used in disease
surveillance [34], and epidemiologists are more likely to be familiar with GIS
software given the long tradition of their usage [35]. In some cases, GIS include
their own analysis tools [36], but these may be bypassed by importing the analysis results directly to the visualisation component. A simpler system, Google
Earth [37], also provides similar functionality for spatial visualisation.
The result reports and visualisations are communicated to the users periodically through email, SMS, automated phone calls, Web sites, or a dedicated
display unit placed at the institution tasked with disease surveillance.
2.2
State-of-the-art
Considering the extensive history of public health literature, the development
of complex information systems for disease surveillance, mainly in the form of
syndromic surveillance, is a recent addition. The first systems that proposed to
monitor non-health related data sources for indicators of public health appeared
in late 1990s [38], and the development of larger systems began after 2001, most
of them in the United States [39]. The last ten years have seen the development of a surprisingly large number of systems with diverse functionality. Four
directions taken by developers of syndromic surveillance can be identified from
existing literature on the topic. These systems are described further in the
following subsection.
• Integrating data collected from many institutions tasked with public health
response to provide an overview of events concerning public health at
the national level. BioSense, one of the largest syndromic systems ever
deployed, accomplishes this by combining the data from diverse health
facilities in 26 states in the U.S. [40].
• Understanding signals in many public health data sources in relation to
each other, during the collection process, at the institute tasked with
collection. RODS achieves this task by providing a self-contained system
that can be deployed independently at multiple public health facilities [41].
• Structuring the collected data and the methods of analysis in order to
ease the difficulty of adding new sources or methods to existing systems.
BioStorm provides ontologies that can classify analysis methods based on
the goal of the analysis. It matches available data sources with suitable
analysis methods [42].
• Increasing the visibility of results of disease surveillance analyses. The
Web-based HealthMap is accessible over the Internet without any additional authentication [43].
Syndromic surveillance literature published in English in the last decade is dominated by systems developed in and intended to be deployed in the United States,
with a few exceptions [44, 45]. More systems, developed in European states have
8
been documented in the broader field of disease surveillance and outbreak detection [46]. Additionally, the European Commission has sponsored several large
projects on Europe-wide systems for awareness and monitoring of pandemics,
but the scope has been very wide and the software output has been modest; see,
e.g., the INFTRANS (Transmission modelling and risk assessment for released
or newly emergent infectious disease agents) project on the sixth framework
programme, 2002–2006 [47].
The most recent development in European syndromic surveillance at the
time of writing is an ongoing project titled Triple-S. It is co-financed by the
European Commission through the executive agency for health and consumers,
and aims to create “an inventory of existing and proposed syndromic surveillance systems in Europe, and provide an in-depth understanding of selected
systems” [48]. The three-year project, initiated in September 2010, involves
twenty-four organisations from thirteen countries.
Personal communication with the project revealed that an investigation
about the existing systems had already been performed, and additional information about the systems would be provided by them at a later, unspecified
date.
2.3
Implementations of syndromic surveillance
Four implementations are presented in this section to illustrate different properties of existing syndromic surveillance systems. For additional information on
other systems, the reader is recommended to refer to [17], which includes an
overview of 50 syndromic surveillance systems and examines eight in further
detail. An earlier review of 115 disease surveillance systems, including nine
syndromic surveillance systems, by [49] is also informative.
2.3.1
BioSense
BioSense is a CDC (Centers for Disease Control and Prevention) initiative that
aims to “support enhanced early detection, quantification, and localisation of
possible biologic terrorism attacks and other events of public health concern on
a national level” [40, p.1]. The software component of the initiative is called the
BioSense application. The development of the application started in 2003, and
the first version was released in 2004.
Initially BioSense included three national data sources: United States Department of Defence military treatment facilities, United States Department of
Veterans Affairs treatment facilities, and test orders from Laboratory Corporation of America. In a later technical report, the BioSense data sources were
reported to also include state/regional surveillance systems, private hospitals
and hospital systems, and outpatient pharmacies [50]. As of May 2008, 454
hospitals from 26 US states were sending data to BioSense.
The BioSense application classifies incoming data into eleven syndrome categories: botulism-like, fever, gastrointestinal, hemorrhagic illness, localised cutaneous lesion, lymphadenitis, neurologic, rash, respiratory, severe illness and
death, and specific infection. The daily statistical analysis is performed using
CUSUM [51], SMART [52], and W2 (a modified version of the C2 method [51]
for anomaly detection). The data reporting component displays the results of
9
the analyses as spreadsheets of observed case counts, time series graphs, patient
maps, or detailed case reports.
In 2010, CDC started redesigning the BioSense program. The redesign aims
to expand the scope of BioSense beyond early detection to contribute information for “public health situational awareness, routine public health practise,
[and] improved health outcomes and public health” [53]. Earlier presentations
about the future of the project have noted additional goals about improving the
usability of biosurveillance tools and “reducing excessive features which miss
the needs of the users” [54, p.19]. Open sourcing of the system is also included
as a possibility for the redesigned BioSense project [55].
The initial motivation for the development and operation of the BioSense application was expressed primarily in terms of preventing biologic terrorism [40].
As part of the redesign, the motivation for developing the system is broadened
considerably:
The goal of the redesign effort is to be able to provide nationwide and
regional situational awareness for all-hazard health-related threats
(beyond bioterrorism) and to support national, state, and local responses to those threats. [53]
The BioSense program has also contributed to the International Society for
Disease Surveillance report on developing syndromic surveillance standards and
guidelines for meaningful use [56]. The current BioSense application is one of
the largest syndromic surveillance systems in existence, and the scope of its
next iteration is likely to be influential in defining what is viable in the field of
disease surveillance systems.
2.3.2
RODS
The development of the Real-Time Outbreak and Disease Surveillance system
(RODS) began in 1999 at the University of Pittsburgh for the purpose of detecting the large-scale release of anthrax [41]. The sixth iteration of the software
is currently reported to be under development and the source code for several
versions licensed under GNU GPL or Affero GPL are available from the RODS
Open Source Project Web site [57].
The first implementation of RODS in Pittsburgh, Pennsylvania collected
chief complaints data from eight hospitals, classified them into syndrome categories, and analysed the data for anomalies [58]. The system was then expanded
to collect additional data types and deployed in multiple states. It was also used
as a user-interface to the American National Retail Data Monitor [59], which
collects over-the-counter medication sales. The most recent publicly available
version of RODS supports user-defined syndrome categories. Implementations
of the recursive least-squared (RLS) algorithm [60] and an initial implementation of the wavelet-detection algorithm [61] are also included. The results of the
analyses can be displayed as time series graphs, or work with a GIS to create
maps of the spatial distribution.
From a data collection perspective, RODS is the decentralised counterpart of
BioSense. Unlike BioSense, which collects data from a large number of sources
centrally within a single implementation, RODS is designed to be installed at facilities on the sub-national level to collect and analyse the available data locally.
In 2009, more than 300 healthcare facilities in 15 states in the U.S., more than
10
200 in Taiwan, and an unspecified number in Canada were being monitored by
independent RODS implementations [57].
At the time of writing, no updates to the RODS open source project have
been committed to the code repository for the last two years, and the latest
available RODS publications date back to 2008. It is unclear if RODS 6 will be
released in the future, but the availability of the source code for many earlier versions makes RODS an important resource for developers of disease surveillance
systems.
2.3.3
BioStorm
The BioStorm system (Biological spatio-temporal outbreak reasoning module)
has been developed at the Stanford Center for Biomedical Informatics Research
in collaboration with McGill University. The goal of the project is to “develop
fundamental knowledge about the performance of aberrancy detection algorithms used in public health surveillance” [62]. The source code for the system
is available at [63].
The aim of the BioStorm project is to create a scalable system that integrates
multiple data sources, includes support for many problem solvers, and provides
flexible configuration options [42]. The defining feature of the project is the
central use of ontologies. A data-source ontology is used to describe data sources.
The descriptions are then used to map to suitable analysis methods available in
the system’s library of problem solvers [64]. Intermediate components such as
a data-broker, a mapping interpreter, and a controller are used to connect the
data sources to the analysis methods. The use of ontologies is intended to ease
the process of adding new data sources and new analysis methods to an existing
BioStorm implementation. No existing syndrome categories or visualisation
components are provided, but any category or visualiser can be added to the
system according to the needs of its users.
The BioStorm project differs from the majority of disease surveillance systems primarily due to its highly complex mechanism for classifying data sources
and problem solvers. The developers reflect on the high overhead of this approach, but state that the overhead is acceptable for systems that connect to
many data sources and require diverse analysis methods [65]. In contrast, most
of the existing syndromic surveillance systems do not suffer from this overhead,
but require additional programming to accommodate new data sources or methods.
The complexity of the BioStorm project creates a significant obstacle for
implementation in a public health facility for day-to-day monitoring. However,
the feature set provided by the project is ideal for systematically comparing
and evaluating the performance of different analysis methods on different data
sources. The possibility of categorising not only the syndromes, but also the
data sources and the analysis methods [66] promises to simplify experiment
design for evaluating detection algorithms.
The BioStorm source code has not been updated since 2010, and the most
recent publication related to the project dates back to 2009, but the source code
continues to be available from the Stanford Center for Biomedical Informatics
Research [63].
11
2.3.4
HealthMap
HealthMap is a freely accessible Web site that integrates data from electronic
sources, and visualises the aggregated information onto the world map, classified
by infectious disease agent, geography, and time. The project aims to deliver
real-time information for emerging infectious diseases. It has been online since
2006, and its current data sources include Google News, the ProMED mailing
list, World Health Organisation announcements and Eurosurveillance publications, among others [67]. HealthMap uses automated text processing to classify
incoming alerts and to create or update points of interest on the world map
based on the classification results [43]. The time-frame of alerts, the number
of alerts, and the number of sources providing information are reflected by the
colour of the markers for the points of interest on the world map. HealthMap
also includes an interface for users to report missing outbreaks.
HealthMap’s reliability, much like any other system, depends on the reliability of its data sources. Since it accesses less reliable sources compared to the
systems discussed previously, different weights are assigned to different sources
based on their credibility when creating reports to offset the influence of less
reliable reports [68].
HealthMap is unique among the systems discussed so far because its analysis
results are available to all World Wide Web users instead of a small group of
experts. The results can be made available without major privacy concerns
because all of the incoming data are also publicly available on the Web. Another
notable system that employs a similar approach is EpiSpider [69].
A recent HealthMap feature, Outbreaks Near Me [70], provides the users
with mobile tools to report and view outbreaks. Accessing the system without
a standard browser requires a smart-phone which limits its availability, but
it is argued that such limitations will eventually be overcome with cheaper
devices [71]. The development of HealthMap-like systems signifies the presence
of a different perspective in public health surveillance, where a larger group
of users are able to influence the surveillance process and access the results of
statistical analyses.
12
3
Employing syndromic surveillance in the digital city
The analysis of the possibilities of employment for syndromic surveillance in
this section begins with a theory of surveillance, which examines the drivers of
surveillance in societies. These drivers are used to reveal future directions that
may be taken by urban applications of syndromic surveillance. The analysis
then proceeds to look at current problems faced by, or posed by, systematic
applications of surveillance, and attempts to extrapolate them into the future.
3.1
A theory of surveillance
“(S)urveillance is a focused attention to personal life details with a
view to managing or influencing those whose lives are monitored. It
may involve care, and more often, control. It is, in other words, the
power of classification, of social sorting.” [72, p.152]
The demand for surveillance appears increasingly often as societies reconfigure
themselves to emphasise the privacy of the individual. This seemingly paradoxical phenomenon is due to the necessity of recognising participants in social
exchanges; the disappearance of local networks of identification (the village, the
extended family, the household) leads to an increase in the demand for “tokens
of trustworthiness” [72, p.66] about the individual. Surveillance is uniquely positioned to generate these tokens, in different forms depending on the contents
of the social exchange: credit cards in financial transactions, access cards when
negotiating physical access, or identification documents when crossing national
borders. A significant driver for the employment of surveillance technologies
is fear: “Fear of attack, of intrusion, of violence, prompts efforts to ward off
danger, to insulate against risk” [72, p.58]. Risk management, in diverse forms,
aims to alleviate this fear with promises of awareness provided by surveillance.
Communicable disease control is one example in which populations are protected
from undesirable biological agents.
3.2
Trends
Current trends that drive surveillance can be understood through five concepts:
rationalisation, technology, sorting, knowledgeability, and urgency [2, p.26], all
of which are present in syndromic surveillance:
• An attempt to understand the behaviour of infectious disease, a seemingly
chaotic phenomenon,
• through the development of new systems that include previously unused
data sources,
• to be able to identify, contact, and respond to the infected individuals,
• (without requiring the awareness of the subjects of surveillance),
• more rapidly than methods used for traditional disease surveillance.
13
These five concepts provide directions in which syndromic surveillance are
most likely to develop in. Advances in syndromic surveillance technologies are
taken for granted within this report, and the four remaining concepts are used to
trace what shape those technologies may take. From the rationalisation perspective, it would be possible to observe a movement towards methods that provide
better detection algorithms. These better detection algorithms are likely to require more computational power to tackle the ever-expanding data sources, and
take into account more variables to make their predictions. The sorting perspective drives the search for more data sources that can be used in syndromic
surveillance. Any large-scale information system intersecting with urban populations can be seen as an immediate candidate for syndromic surveillance, and
therefore of public health interest in urban environments, while remaining within
the limits of established data protection laws and regulations. Knowledgeability,
referring to the awareness among the subjects of surveillance about the existence
and the functions of surveillance systems, would continue to influence how individuals react to newly developed systems, leading to explicit acceptance or
rejection if the systems are revealed from the beginning, or a gradual awareness
and eventual reaction if the systems attempt to conceal their presence. The final concept, urgency, would drive developers of syndromic surveillance systems
to seek data sources that contain indicators for increasingly rapid detection.
These extremely early indicators may be sought in genetic information, purchase habits, or any other source that records an individual before her contact
with the infectious agent.
3.3
Current issues
Computerised systems for public health surveillance have clear advantages compared to their earlier, paper-based versions. When the required information
infrastructure for the operation of the system is in place, these systems reduce
the number of humans required to detect and communicate health risks, and
they can even automate daily monitoring and detection tasks. Additionally, the
systems function without any additional input from individuals whose data are
used. No one has to opt in to be included in an outbreak investigation, and
individuals are very rarely identified during such investigations.
However, these systems require considerable infrastructure as prerequisites
before they can even be implemented. For example, areas without reliable communication networks rarely benefit from the implementation of such systems.
Also following directly from the previous advantage, it is not possible to opt
out of many public health surveillance systems, and as a result the majority
of current implementations are undemocratic. Individuals, whose data inform
public health, have no say in how their data are used, or whether it should be
used at all.
Contemporary syndromic surveillance systems have a tendency to generate
more false alarms than traditional disease surveillance methods. Practitioners of
public health tasked with monitoring such alarms have stated that without the
resources to investigate detected events, the detection capability is useless [6].
In future systems, this may lead to several new research directions: Syndromic
surveillance algorithms may aim to have higher specificity and lower sensitivity,
or they may require more data sources to confirm the detected event before an
alert is raised. To accomplish this cross-checking, more data sources that are
14
representative of the same population would have to be monitored.
Quantifying the representativity of a data source, and judging how similar
two separate data sources are in terms of representativity is yet another investigation that is likely to take place in institutions that employ syndromic
surveillance methods. Surveillance of data sources that include multiple activities linked to the same identity are likely candidates for syndromic surveillance.
Since these systems link diverse data sources, they require a complex network
of users and machines to support their daily functions; the maintenance of
syndromic surveillance systems are non-trivial. They connect to software that
may change over time, and as the number of sources increases, compatibility
issues also increase. While there are standards for storing and communicating
health information digitally, the field of syndromic surveillance today is far
from standardisation. If a single project were to gain enough momentum, such
a standard could be established. Factors that affect the adoption of one project
are fairly similar to other software standards, with one important difference:
Public health is performed at the national level, each nation state exercises its
public health duties in different ways, and those ways are rigidly encoded in the
laws of the state. For this reason, the standardisation of syndromic surveillance,
much like any other information system dealing with health at the national level,
would require consensus at the international level.
3.3.1
Commerce and public health
Syndromic surveillance, as well as other forms of surveillance for public health
purposes, are performed with the explicit aim of raising the health of populations. The information gathered is used to establish norms of disease, in order
to be able to detect when those norms are violated, and to judge if the costs
of intervention serve the aim of raising the health of the population. This calculation of costs is inherently incompatible with the notion of profit. Since the
aim is to raise the health of the whole population indiscriminately, any potential gains, be it knowledge, or accumulation of resources, are committed to the
same aim. Creating other venues where capital can be generated for those who
operate these mechanisms is bound to encounter an internal inconsistency.
This inconsistency may take either the form of withholding, where resources
generated from the surveillance process are not completely committed to the
improvement of that same process, or it may take, in the worst case, the form
of active deception, where the primary aim of public health is subverted to
benefit not the whole but only certain parts of the population without acknowledging the change in recipients of the benefits explicitly. For this reason, a
commercial entity, predicated by the motive of creating additional value for its
shareholders, very rarely functions in the public health domain without critical
conflicts. A deeper theoretical investigation of the causes of this incompatibility is described in a report from the United Nations Research Institute for
Social Development [73], which uses cross-country comparable data to illustrate,
among many other comparisons, a negative relationship between life expectancy
and private health care expenditure.
Regarding commercialisation, public health surveillance does not differ from
health care significantly. In a practice where subjects rarely volunteer to be
included, problems of redistribution arise immediately when commercial interests appear. Attempts to convince individuals to volunteer, to offer them other
15
advantages in exchange for their data, have succeeded in domains of social communication through information systems. Google’s automatic mining of emails
to provide personalised ads, or Facebook’s initiative to insert individuals to ads
personalised for the viewing of that individual’s friends are existing examples of
such technology. These measures have been contested [74, 75, 76], and continue
to be contested, but their presence has not led to a rejection of the services as
a whole by their users. In public health, however, attempts to convince individuals to volunteer must be held to much closer scrutiny. Existing research
on ethics, such as Gary Marx’s ethics of surveillance [77], may provide an idea
of what forms this scrutiny should take. The issue is ensuring that the governed population receives equal benefit from public health measures, including
health surveillance, and that they are provided with opportunities to exercise
such benefits. Emphasising individual choice leads naturally to the possibility
that the wrong choice will be made by some, and that they will be responsible
for the consequences of those wrong choices. The ability to make choices, especially of making the right choice given a particular circumstance, is strongly
influenced by prior privilege; those equipped with more privilege are more likely
to make the better choice. The logic of choice is incompatible with the imperative of public health, because the outcomes linked to individual choice are
strongly linked to prior privilege, and the only part prior privilege can play in
the discourse of public health is that of an obstacle to be overcome.
Commercial attempts at entering the domain of public health surveillance
vary based on how the expertise of the commercial organisation is positioned.
Using the collection, analysis, and notification framework for syndromic surveillance, these entry points can be made slightly more visible. At the collection
stage, a company can propose to handle data collection and storage tasks, by
proposing similarities to data warehousing that is very common in companies
that work with information technologies. The arguments that such companies
would propose are likely to mention the difficulty of storing large databases in
hardware, managing backup methods, or distributing resources to optimise the
speed of retrieval. In the analysis stage, the role of the company can be stated
as a consultancy, bringing expertise (statistical, computational, etc.) to analysis tasks. For example, companies with data mining technologies may present
proposals where their methods are used, or companies that manage large groups
of experts may provide the actual worker who would perform and interpret the
statistical analyses chosen by the operators of the surveillance systems. Finally,
at the notification stage, companies may provide novel visualisation methods
that are inaccessible to health agencies, due to laws regulations surrounding intellectual property for example, or they may claim to have more suitable technological solutions for the reliable communication of messages such as secure
cellular networks.
3.4
The future of syndromic surveillance
It is possible to provide a rudimentary method for inventing new applications
of syndromic surveillance: Proposals for future information systems or newly
implemented systems are searched for specifications of the data they will collect
or generate. Then, the behaviour of infectious diseases of interest are crossreferenced one by one to with these specifications of data sources, to see if there
are any intersections. These intersections may occur as transmission mechanics,
16
symptoms of the diseases, treatment methods, etc. If any intersection exists, an
investigation can be made regarding the applicability of syndromic surveillance
to the data source. This investigation would have to take into account what
part of the population the data source represents, and how reliable (in terms of
sensitivity, specificity, and timeliness) the signals derived from the data source
would be. Some example scenarios are as follows:
• Records of energy usage in households (the smart home scenario): Does
sickness change the energy consumption behaviour of the inhabitants in
ways that are visible to the smart home data sources?
• Transportation: Electrical cars are plugged to a grid for charging. Given
a car user that drives to work every day, would it be possible to assume
the presence of sickness if the car remains plugged in to the grid during
work hours?
• Supermarket customer loyalty cards: Can personal purchase records be
analysed real-time to detect the presence of communicable diseases from
the consumption of remedies?
As stated previously, data collected by a wide variety of information systems in urban environments can be used in syndromic surveillance with the aim
of extracting health-related information. Current examples include ambulance
dispatch records and over-the-counter medicine sales that are analysed statistically to detect unexpected changes. Following the same trend, other data
sources may be used in the near future for disease surveillance tasks. One particularly complicated scenario links cellphone logs to disease spread. Assuming
full availability of cellphone handover logs between base stations, it would be
possible to trace potential contacts of infected individuals in order to warn them
about the effects of exposure, and recommend them to take precautionary measures to avoid spreading the disease further. The privacy implications of this
scenario depend primarily on the roles of the actors who would define the practical effects of the precautionary measures to be taken, and how those measures
should be enforced. Given a wealth of real-time or near-real-time data sources,
many detection methods previously thought impossible become possible. However, these implementations should be accompanied with a discussion of their
implications: how do these methods shape society, who enjoys the benefits, who
suffers the costs, and finally, which possible futures become impossible, or highly
unlikely, once such systems are in place?
3.4.1
Exclusion
Syndromic surveillance systems must strive to avoid excluding parts of the population. Two factors contribute to this requirement: First, the imperative of
public health places strong emphasis on equal inclusion, and to deviate from it
would make syndromic surveillance less compatible with the practice of public
health. Second, communicable diseases spread indiscriminately. Not taking into
account parts of the population would only lead to partitions where communicable disease can spread unchecked among available hosts, which would not only
make syndromic surveillance less efficient, but also increase the human suffering
caused by communicable diseases that syndromic surveillance aims to alleviate.
17
Where is exclusion observed in the methods and systems of syndromic surveillance? It appears, most importantly, in the data sources. If parts of the population are not represented in the data sources, syndromic surveillance will not
say anything about their suffering caused by communicable diseases. Exclusion
may also appear during the analysis stage where the intersection of multiple data
sources are investigated. If those sources do not represent identical segments of
the population, which is the usual case, individuals outside of that intersection
would not benefit from the surveillance even though their data appears in some
of the data sources.
Urban environments do not have rigid borders; they grow, recede, and shift
as the cities evolve. While this is, from a humanitarian perspective, a property
highly worth preserving, it introduces complications for a public health practice that aims to be non-exclusive. New inhabitants of the city do not become
visible to public health surveillance as easily as already established inhabitants,
and consequently, delivery of interventions or preventive measures may not encompass the recently settled. Under the assumption that the future influx of
populations to cities will tend to increase, new systems of surveillance, and particularly of syndromic surveillance, must be designed to take into account the
necessity of including those that have recently arrived in the city.
3.5
A candidate for future case studies
The London 2012 Olympics is one example where urban syndromic surveillance
systems of the future will become observable. Relatively short in duration, but
very large in terms of attendance, the Olympics creates a unique opportunity
where extreme migration challenges existing infrastructure, when an urban event
becomes highly complex due to the sudden increase in the population. Proposals for temporary syndromic surveillance systems, constructed with the aim of
monitoring the population during the event, would be very valuable for observing the trends discussed in this report in action. The discussions surrounding
the hosting of the Olympics in a highly populated capital are informative, because they illustrate what the organising institutions consider to be inadequate
for the influx of visitors in the near future. These discussions can be extended
into the further future, under the assumption that problems caused by increased
migration are likely to take similar forms. After the event, evaluations of the
proposed remedies can serve as additional indicators of the future issues.
3.6
Conclusion
No information system stands independent of the population it informs or monitors. It is of utmost importance that the potential social effects of surveillance
systems be put under close scrutiny before any resources are committed to realising them, but the complete picture, with regards to the redistribution of
privilege due to the operation of the system, will never be possible to predict
in its entirety. In fact, the act of prediction modifies the outcome, especially
if it leads to media attention, political scrutiny, or any form of campaigning
for, or against, the proposal. The best outcome that planners and developers
of syndromic surveillance can hope to achieve, while staying committed to the
imperative of public health, is to minimise the tendencies of their systems to
selectively privilege only certain parts of the population.
18
The main author of this report is Baki Cakici, Ph.L., KTH. The editing of
the report was done by Magnus Boman, professor, KTH.
19
References
[1] EIT ICT Labs. Digital cities of the future. http://eit.ictlabs.eu/
ict-labs/thematic-action-lines/digital-cities-of-the-future,
2011. Accessed 2011-10-06.
[2] David Lyon. Surveillance Studies: An Overview. Polity, first edition, August 2007. ISBN 074563592X.
[3] Bank of America. Bank of America - Photo Security, 2011. URL
http://www.bankofamerica.com/creditcards/index.cfm?template=
cc_features_photo_security. Accessed 2011-11-24.
[4] Transport for London. Pay as you go - Transport for London, 2011. URL
http://www.tfl.gov.uk/tickets/19798.aspx. Accessed 2011-11-24.
[5] J. W Buehler, R. S Hopkins, J. M Overhage, D. M Sosin, and V. Tong.
Framework for evaluating public health surveillance systems for early detection of outbreaks. MMWR Morbidity and Mortality Weekly Report, 53:
1–11, 2004.
[6] F. Mostashari. Syndromic surveillance: a local perspective. Journal of
Urban Health: Bulletin of the New York Academy of Medicine, 80:i1–i7,
2003. ISSN 1099-3460.
[7] Kelly J Henning. What is syndromic surveillance? MMWR. Morbidity and
Mortality Weekly Report, 53 Suppl:5–11, September 2004.
[8] Michael M. Wagner, Andrew W. Moore, and Ron M. Aryel. Handbook of Biosurveillance. Academic Press, first edition, June 2006. ISBN
0123693780.
[9] Baki Cakici. Disease surveillance systems. Licentiate dissertation, Royal
Institute of Technology (KTH), 2011.
[10] V.W. Sidel, R.M. Gould, and H.W. Cohen. Bioterrorism preparedness:
cooptation of public health? Medicine & Global Survival, 7(2):82–89, 2002.
[11] K. C Dowling and R. I Lipton. Bioterrorism preparedness expenditures
may compromise public health. American Journal of Public Health, 95
(10), 2005.
[12] A. Reingold. If syndromic surveillance is the answer, what is the question?
Biosecurity and Bioterrorism: Biodefense Strategy, Practice, and Science,
1(2):77–81, 2003.
[13] Magdalena Berger, Rita Shiau, and June M Weintraub. Review of syndromic surveillance: implications for waterborne disease detection. Journal
of Epidemiology and Community Health, 60(6):543–550, June 2006.
[14] David L. Buckeridge, Justin Graham, Martin J. O’Connor, Michael K.
Choy, Samson W. Tu, and Mark A. Musen. Knowledge-based bioterrorism
surveillance. Proceedings of the AMIA Symposium, pages 76–80, 2002.
20
[15] Steven Babin, Steven Magruder, Shilpa Hakre, Jacqueline Coberly, and
Joseph S. Lombardo. Understanding the data: Health indicators in disease
surveillance. In Joseph S. Lombardo and David L. Buckeridge, editors,
Disease Surveillance: A Public Health Informatics Approach, chapter 2,
pages 43–90. Wiley-Interscience, first edition, apr 2007.
[16] J. W Buehler, A. Sonricker, M. Paladini, P. Soper, and F. Mostashari.
Syndromic surveillance practice in the united states: findings from a survey
of state, territorial, and selected local health departments. Advances in
Disease Surveillance, 6(3):1–20, 2008.
[17] Hsinchun Chen, Daniel Zeng, and Ping Yan. Infectious Disease Informatics:
Syndromic Surveillance for Public Health and Bio-Defense. Springer, first
edition, December 2009. ISBN 1441912770.
[18] Anette Hulth, Gustaf Rydevik, and Annika Linde. Web queries as a source
for syndromic surveillance. PloS One, 4(2), 2009. ISSN 1932-6203.
[19] Jeremy Ginsberg, Matthew H. Mohebbi, Rajan S. Patel, Lynnette Brammer, Mark S. Smolinski, and Larry Brilliant. Detecting influenza epidemics
using search engine query data. Nature, 457(7232):1012–1014, February
2009.
[20] Vasileios Lampos, Tijl Bie, and Nello Cristianini. Flu detector - tracking
epidemics on twitter. In José Luis Balcázar, Francesco Bonchi, Aristides
Gionis, and Michèle Sebag, editors, Machine Learning and Knowledge Discovery in Databases, volume 6323, pages 599–602. Springer Berlin Heidelberg, Berlin, Heidelberg, 2010.
[21] D. Travers, C. Barnett, A. Ising, and A. Waller. Timeliness of emergency
department diagnoses for syndromic surveillance. In AMIA Annual Symposium Proceedings, volume 2006, page 769, 2006.
[22] Ron Brookmeyer and Donna F. Stroup. Monitoring the Health of Populations: Statistical Principles and Methods for Public Health Surveillance.
Oxford University Press, USA, 1 edition, October 2003. ISBN 0195146492.
[23] Christian Sonesson and David Bock. A review and discussion of prospective
statistical surveillance in public health. Journal of the Royal Statistical
Society: Series A (Statistics in Society), 166(1):5–21, February 2003.
[24] Andrew B. Lawson and Ken Kleinman. Spatial and Syndromic Surveillance
for Public Health. Wiley, May 2005. ISBN 0470092483.
[25] W.K. Wong and A.W Moore. Classical time-series methods for biosurveillance. In Michael M. Wagner, Andrew W. Moore, and Ron M. Aryel, editors, Handbook of biosurveillance, chapter 14. Academic Press, jun 2006.
[26] H. Burkom. Alerting algorithms for biosurveillance. In Joseph S. Lombardo
and David L. Buckeridge, editors, Disease Surveillance: A Public Health
Informatics Approach, chapter 4. Wiley-Interscience, first edition, apr 2007.
21
[27] David L Buckeridge, Anna Okhmatovskaia, Samson Tu, Martin O’Connor,
Csongor Nyulas, and Mark A Musen. Understanding detection performance
in public health surveillance: Modeling aberrancy-detection algorithms.
Journal of the American Medical Informatics Association, 15(6):760–769,
November 2008.
[28] Anita Pelecanos, Peter Ryan, and Michelle Gatton. Outbreak detection
algorithms for seasonal disease data: a case study using ross river virus
disease. BMC Medical Informatics and Decision Making, 10(1):74, 2010.
[29] Martin Kulldorff, Farzad Mostashari, Luiz Duczmal, W. Katherine Yih,
Ken Kleinman, and Richard Platt. Multivariate scan statistics for disease
surveillance. Statistics in Medicine, 26(8):1824–1833, 2007.
[30] K. P. Kleinman and A. M. Abrams. Assessing surveillance using sensitivity,
specificity and timeliness. Statistical methods in medical research, 15(5):
445, 2006.
[31] David L. Buckeridge, Howard Burkom, Murray Campbell, William R.
Hogan, and Andrew W. Moore. Algorithms for rapid outbreak detection: a
research synthesis. Journal of Biomedical Informatics, 38(2):99–113, April
2005.
[32] R. G Cromley and E. K Cromley. Choropleth map legend design for visualizing community health disparities. International Journal of Health
Geographics, 8(1), 2009.
[33] Francis P. Boscoe, Colleen McLaughlin, Maria J. Schymura, and Christine L. Kielb. Visualization of the spatial scan statistic using nested circles.
Health & Place, 9(3):273–277, September 2003.
[34] Candace I. J. Nykiforuk and Laura M. Flaman. Geographic information
systems (GIS) for health promotion and public health: A review. Health
Promotion Practice, 12(1):63–73, January 2011.
[35] K. C. Clarke, S. L. McLafferty, and B. J. Tempalski. On epidemiology
and geographic information systems: a review and discussion of future
directions. Emerging Infectious Diseases, 2(2):85–92, 1996.
[36] Kyusuk Chung, Duck-Hye Yang, and Ralph Bell. Health and GIS: toward
spatial statistical analyses. Journal of Medical Systems, 28(4):349–360,
2004.
[37] Google. Google earth. http://earth.google.com, 2011. Accessed 201104-22.
[38] R. Heffernan et al. New york city syndromic surveillance systems. MMWR.
Morbidity and mortality weekly report, 53:25–27, 2004.
[39] D. M. Sosin and J. DeThomasis. Evaluation challenges for syndromic
surveillance – Making incremental progress. MMWR. Morbidity and mortality weekly report, 53:125, 2004.
22
[40] C. A. Bradley, H. Rolka, D. Walker, and J. Loonsk. BioSense: implementation of a national early event detection and situational awareness system.
MMWR. Morbidity and Mortality Weekly Report, 54:11, 2005.
[41] 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: JAMIA, 10(5):399–408, 2003. ISSN 10675027.
[42] Martin J. O’Connor, David L. Buckeridge, Michael Choy, Monica Crubézy,
Zachary Pincus, and Mark A. Musen. BioSTORM: a system for automated
surveillance of diverse data sources. AMIA Annu Symp Proc, 2003:1071–
1071, 2003.
[43] Clark C Freifeld, Kenneth D Mandl, Ben Y Reis, and John S Brownstein.
HealthMap: global infectious disease monitoring through automated classification and visualization of internet media reports. Journal of the American Medical Informatics Association, 15(2):150–157, March 2008.
[44] L Josseran, J Nicolau, N Caillère, P Astagneau, and G Brücker. Syndromic
surveillance based on emergency department activity and crude mortality:
two examples. Eurosurveillance, 11(12):225–229, 2006.
[45] C. van den Wijngaard, W. van Pelt, N. Nagelkerke, M. Kretzschmar, and
M. Koopmans. Evaluation of syndromic surveillance in the netherlands: its
added value and recommendations for implementation. Eurosurveillance,
16(9), 2011.
[46] A. Hulth, N. Andrews, S. Ethelberg, J. Dreesman, D. Faensen, W. van
Pelt, and J. Schnitzler. Practical usage of computer-supported outbreak
detection in five european countries. Eurosurveillance, to appear, 15(36),
2010.
[47] European Commission. INFTRANS – infectious diseases: modelling for
control.
http://ec.europa.eu/research/fp6/ssp/inftrans_en.htm,
2008. Accessed 2011-04-27.
[48] The Triple-S project. Triple-S – The syndromic surveillance project. http:
//www.syndromicsurveillance.eu, 2011. Accessed 2011-10-06.
[49] Dena M. Bravata, Kathryn M. McDonald, Wendy M. Smith, Chara Rydzak, Herbert Szeto, David L. Buckeridge, Corinna Haberland, and Douglas K. Owens. Systematic review: Surveillance systems for early detection
of Bioterrorism-Related diseases. Annals of Internal Medicine, 140(11):
910–922, June 2004.
[50] CDC. Biosense technical overview of data collection, analysis, and reporting. Technical report, Centers for Disease Control and Prevention, June
2008.
[51] L Hutwagner, W Thompson, G M Seeman, and T Treadwell. The bioterrorism preparedness and response early aberration reporting system (EARS).
Journal of Urban Health: Bulletin of the New York Academy of Medicine,
80:i89–i96, 2003.
23
[52] Ken Kleinman, Ross Lazarus, and Richard Platt. A generalized linear
mixed models approach for detecting incident clusters of disease in small
areas, with an application to biological terrorism. American Journal of
Epidemiology, 159(3):217 –224, February 2004.
[53] CDC. CDC BioSense Home. http://www.cdc.gov/biosense, 2011. Accessed 2011-04-01.
[54] Taha A. Kass-Hout. BioSense: next generation. http://www.cdc.
gov/biosense/files/PHIN_PowerPoint_BioSense_Next_Generation_
FINAL_508.pdf, 2009. Accessed 2011-04-19.
[55] Taha A. Kass-Hout. BioSense: Going forward. http://www.cdc.gov/
biosense/files/BioSense_ISDS.ppt, December 2009. Accessed 2011-0419.
[56] ISDS. Final Recommendation: Core Processes and EHR Requirements
for Public Health Syndromic Surveillance. Technical report, International
Society for Disease Surveillance, January 2010. URL \url{http://www.
syndromic.org/projects/meaningful-use}.
[57] RODS.
The RODS Open Source Project.
sourceforge.net, 2009. Accessed 2011-04-04.
http://openrods.
[58] J. U Espino, M. Wagner, C. Szczepaniak, F-C. Tsui, H. Su, R. Olszewski,
Z. Liu, W. Chapman, X. Zeng, L. Ma, Z. Lu, and J. Dara. Removing a
barrier to computer-based outbreak and disease surveillance—The RODS
open source project. Syndromic Surveillance, page 32, 2004.
[59] Michael M. Wagner, J.Michael Robinson, Fu-Chiang Tsui, Jeremy U. Espino, and William R. Hogan. Design of a national retail data monitor for
public health surveillance. Journal of the American Medical Informatics
Association, 10(5):409–418, 2003.
[60] Monson H. Hayes. Statistical digital signal processing and modeling. John
Wiley & Sons, 1996. ISBN 0471594318.
[61] Jun Zhang, Fu -Chiang Tsui, Michael M Wagner, and William R Hogan.
Detection of outbreaks from time series data using wavelet transform.
AMIA Annual Symposium Proceedings, pages 748–752, 2003.
[62] BioSTORM. Welcome to BioSTORM. http://biostorm.stanford.edu/
doku.php, 2009. Accessed 2011-04-06.
[63] BioSTORM. BioSTORM SVN Repository. https://bmir-gforge.
stanford.edu/gf/project/biostorm/scmsvn, 2010. Accessed 2011-0406.
[64] Monica Crubézy, Martin O’Connor, David L. Buckeridge, Zachary Pincus,
and Mark A. Musen. Ontology-Centered syndromic surveillance for bioterrorism. IEEE Intelligent Systems, 20(5):26–35, 2005.
[65] D. L. Buckeridge, M. A. Musen, P. Switzer, and M. Crubézy. A modular
framework for automated Space-Time surveillance analysis of public health
data. In AMIA Annual Symposium, Washington, DC, pages 120–124, 2003.
24
[66] Zachary Pincus and Mark A Musen. Contextualizing heterogeneous data
for integration and inference. In AMIA 2003 Symposium Proceedings, pages
514–518, 2003.
[67] HealthMap. Global health, local information. http://healthmap.org/en,
2011. Accessed 2011-04-07.
[68] John S Brownstein, Clark C Freifeld, Ben Y Reis, and Kenneth D Mandl.
Surveillance sans frontières: Internet-based emerging infectious disease intelligence and the healthmap project. PLoS Med, 5(7):e151, 2008.
[69] Herman Tolentino, Raoul Kamadjeu, Paul Fontelo, Fang Liu, Michael Matters, Marjorie Pollack, and Larry Madoff. Scanning the Emerging Infectious Diseases Horizon-Visualizing ProMED Emails Using EpiSPIDER. Advances in Disease Surveillance, 2:169, 2007.
[70] HealthMap.
Outbreaks near me.
outbreaksnearme, 2011. Accessed 2011-04-08.
http://healthmap.org/
[71] Clark C. Freifeld, Rumi Chunara, Sumiko R. Mekaru, Emily H. Chan, Taha
Kass-Hout, Anahi Ayala Iacucci, and John S. Brownstein. Participatory
epidemiology: Use of mobile phones for community-based health reporting.
PLoS Med, 7(12), 12 2010.
[72] David Lyon. Surveillance Society: Monitoring Everyday Life. Open University Press, first edition, February 2001. ISBN 0335205461.
[73] Maureen Mackintosh. Health care commercialisation and the embedding
of inequality, RUIG/UNRISD health project synthesis paper. Geneva,
Switzerland: United Nations Research Institute for Social Development,
2003.
[74] Electronic Frontier Foundation. Lawsuit demands answers about socialnetworking surveillance, 2009.
URL https://www.eff.org/press/
archives/2009/11/30. Accessed 2011-11-30.
[75] Tor Project. Tor Project: Anonymity Online, 2011. URL https://www.
torproject.org/. Accessed 2011-11-30.
[76] Adblock Plus. Adblock Plus – for annoyance-free web surfing, 2011. URL
http://adblockplus.org/en/. Accessed 2011-11-30.
[77] Gary T. Marx. Ethics for the new surveillance. The Information Society,
14:171–185, August 1998. doi: 10.1080/019722498128809. URL http:
//www.tandfonline.com/doi/abs/10.1080/019722498128809.
25
Imprint:
EIT ICT Labs IVZW
22 Rue d'Arlon, 1050 Brussels
BELGIUM
ISBN 978-91-87253-04-1