A WEB-BASED SYSTEM TO REDUCE THE NOSOCOMIAL
INFECTION IMPACT IN HEALTCARE UNITS
Hugo Rigor, Jos´e Machado, Ant´onio Abelha, Jos´eNeves
Departamento de Inform´atica, Universidade do Minho, Campus de Gualtar, Braga, Portugal
Carlos Alberto
Centro Hospitalar do Porto, Porto, Portugal
Keywords:
Web Engineering, System Integration, Databases.
Abstract:
The nosocomial infection has a critical impact on the mortality and morbidity of the patients in healthcare units,
especially in intensive care units, and has been studied in order to be mitigated. The registration of information
about this phenomenon in databases is, more and more, a reality, turning viable the representation of this
information through mathematical formalisms that, conjugated with the application of Artificial Intelligence
techniques, will allow the discovery of knowledge related with the critical factors, processes and infectious
agents. The ultimate goal has been to construct a web-based computational tool to automate the registration
process to support the clinical body work, monitoring the performance and identifying the procedures that can
be implemented in order to reduce the impact of the infections.
1 INTRODUCTION
Medicine has been for some years a very attractive
domain for Computer Science (CS) researchers, in
general. There is a great potential for information
automation, and a lot remains to be done. Med-
ical Informatics (MI) is indeed an issue of study in
which Medicine and Computing overlap. Another
reason for this increasing interest was costs. Today’s
strained health-care economics makes it necessary for
expensive resources to be efficiently managed. CS
researchers have long used Medicine to elaborate on
their own work. The field is probably one of the most
knowledgeintensive ones, loadedwith human reason-
ing, with most of the procedure relying exclusively
on the clinical experts. This makes health-care a per-
fect target for CS, since conventional systems are nat-
urally bounded by their lack of rich knowledge rep-
resentation and proof schemes. On the other hand,
MI Systems have to be addressed in terms of a wide
variety of heterogeneous distributed systems speak-
ing different languages, integrating medical equip-
ment and customized by several companies, which
in turn were developed by people aiming at different
goals. This lead us to consider the solution to a par-
ticular problem, to be part of an integration process of
different sources of information, using different pro-
tocols, in terms of an Agency for the Integration, Dif-
fusion and Archive (AIDA) (Figure 1) of medical in-
formation, and the Electronic Medical Record (EMR)
software (in Portuguese referred as PCE - Processo
Cl´ınico Electr´onico), bringing to the healthcare arena
new methodologies for problem solving, computa-
tional models, technologies and tools, using ambient
intelligence (Removed for blind review). The homo-
geneity of clinical, medical and administrative sys-
tems is not possible due to financial and technical re-
strictions, as well as functional needs. The solution
is to integrate, diffuse and archive this information
under a dynamic framework, in order to share this
knowledge with every information system that needs
it. Indeed, to build systems for real health care en-
vironments, the infrastructure must meet a range of
basic requirements with respect to security, reliabil-
ity and scaling. With access granted to Clinical and
Historical Databases, agent technology may provide
answers to those who give assistance to patients with
a maximum of quality and medical evidence (Maes,
1995).
264
Rigor H., Machado J., Abelha A., Neves J. and Alberto C. (2008).
A WEB-BASED SYSTEM TO REDUCE THE NOSOCOMIAL INFECTION IMPACT IN HEALTCARE UNITS.
In Proceedings of the Four th Inter national Conference on Web Information Systems and Technologies, pages 264-268
DOI: 10.5220/0001522602640268
Copyright
c
SciTePress
Figure 1: AIDA.
2 AGENT ORIENTED
PROGRAMMING
The traditional programming languages do not sup-
port the description of certain types of behaviour
which usually involves computational agents (Weiss,
1999). In genesis, systems that incorporate those
functionalities have a multi-layer architecture, evolve
from esoteric software sub-systems, network proto-
cols, and the like. An agent must be able to manage
its knowledge, beliefs, desires, intentions, goals and
values ((Analide et al., 2006)). It may be able also to
plan, receive information or instructions, or react to
environment stimulus. It may communicate with oth-
ers agents, share knowledge and beliefs, and respond
to other agents upon request. It may cooperate for
diagnosing errors or information faults in its knowl-
edge bases, sharing resources, avoiding undesirable
interferences or joining efforts in order to revisit the
knowledge bases of its own and of its peers, in or-
der to reach common goals (Machado et al. 2006).
Agents exchange messages which are well-formed
formulae of the communication language, performing
acts or communicative actions ((Bradshaw, 1997)).
Agents in a healthcare facility configure applications
or utilities that collect information about the assets
in the organization ((Alves et al., 2005)). Once that
information has been collected it can be posted di-
rectly to other entities (e.g. a physician), a server,
saved to a file, emailed to someone to be handled
at a later date or sent using HL7 (Health Level 7)
(http://www.hl7.org). Indeed, in a hospital, the col-
lection of vast amounts of medical data will not only
support the requirements of archiving, but also pro-
vide a platform for the application of data mining and
knowledge discovery techniques to determine possi-
ble medical trends and the real data to support ed-
ucational training. Knowledge discovery techniques
can be applied to identify pathologies and disease
trends. The data can also be used for educational
and training purposes; unique cases can be identi-
fied and used to advise practitioners. This may lead
to one’s goal of Ambient Intelligence at Health Care
Units at our doorstep (Costa et al., 2007) (Abelha
et al., 2007). HL7 plays an essential role in ex-
tending the interoperability for the development of
health information exchange, in the standardizationof
XML medical document structures and in the specifi-
cation of robust vocabulary definitions for use in clin-
ical messages and documents (e.g. SNOMED CT)
(http://www.ihtsdo.org/) enabling functional specifi-
cations for the EMR.
3 THE ELECTRONIC MEDICAL
RECORD
The Electronic Medical Record (EMR) is a core ap-
plication which covers horizontally the health care
unit and makes possible a transverse analysis of
medical records along the services, units or treated
pathologies, bringing to the healthcare arena new
methodologies for problem solving, computational
models, technologies and tools. One aims to develop
a comprehensive, structured approach to EMR devel-
opment and analysis. Indeed, this paper will thrash
out the inner features of intelligent agents to be used
in the EMR, in the context of the Telemedical Infor-
mation Society, a step in the direction of Intelligent
Health Care Units ((Hendler, 1996)). The process
to collect data comes from Problem Oriented Med-
ical Record (POMR) method ((Weed, 1969)). This is
a format for clinical recording consisting of a prob-
lem list, a database including the patient history with
physical examination and clinical findings, diagnos-
tic, therapeutic and educational plans ((Machado and
Alves, 2005) and a daily SOAP (Subjective, Objec-
tive, Assessment and Plan) progress note ((of Medi-
cine, 1991)). The problem list serves as an index for
the reader, each problem being followed through until
resolution. This system widely influences note keep-
ing by recognizing the five different phases of the de-
cision making process, i.e. data collection, problem
specification, devising a management plan, review-
ing the situation and revising the plan if necessary.
One’s goal is to replace hard documents by electronic
ones, increasing data processing and reducing time
and costs. The patient assistance will be more effec-
tive, faster and the quality of service will be improved
((Machado et al., 2006)). The system uses freeware
tools or software database packages which licenses
belong to the Portuguese Health Ministry (e.g. Or-
acle software). Messages are sent by agents using
XML or HL7. According to the ontology, messages
are processed, integrated and archived in large data-
bases. The ontology is defined by the administrators
and can be managed using web tools. The ”intelli-
A WEB-BASED SYSTEM TO REDUCE THE NOSOCOMIAL INFECTION IMPACT IN HEALTCARE UNITS
265
gence” of the system as a whole arises from the inter-
actions among all the system’s components. The in-
terfaces are based on Web-related front-ends, query-
ing or managing the data warehouse. Such an ap-
proach can provide decision support. A context de-
pendent formalism has been used to specify the AIDA
system incorporatingfacilities such as abstraction, en-
capsulation and hierarchy, in order to define the sys-
tem components or agents; the socialization process,
at the agent level and the multi-agent level, following
other possible way of aggregation and cooperation;
the coordination procedure at the agent level; and the
global system behaviour. The data can also be used
for educational and training purposes because maybe
one of the unique cases can be identified and used in
expert system like applications to advise practitioners
((Analide et al., 2006)). An screen view is shown in
Figure 2.
Figure 2: Procedure Registration in the EMR.
4 A WEB-BASED SYSTEM FOR
THE NOSOCOMIAL
INFECTION REGISTRATION
Despite the advancements in the area of the health
care, every year around one hundred thousand pa-
tients contract an infection at the hospital environ-
ment and more than ten thousand die in consequence
of these infections. Beyond the morbidity and con-
siderable mortality, the hospital infection provokes a
considerable increase of costs with trials of diagnosis,
of therapy and with the increaseof the time of hospital
admission.
The hospital or nosocomial infection is an infec-
tion that is notpresent at the momentof patient admis-
sion to the hospital. It can be provoked by internal or
external agents to the patient. Some of the factors that
prepare a patient to the development of a nosocomial
infection are directly related with the age, the state of
his immunity, the use of antibiotherapy, the time of
hospitalization, the techniques of diagnosis, among
others. The dynamics and the hospital environment
provide, on the other hand, diverse potential foci of
infection as the patients, the visits, the personnels, the
equipment, the installations, the environment, the in-
vasive devices, etc. All these factors are potentiated
when the patients are at Intensive Care Units, where
the problem of the nosocomial infectionhas the great-
est impact inside the Hospital environment.
The existence of scientific studies and guidelines,
such as norms to control the infection and consensus
for the diagnosis and handling of the infection, con-
tribute for a deeper understandingand diffusion of the
phenomena associated with the infections, in terms
of its prevention, diagnosis and handling. However,
despite such advancements, an efficient diagnosis on
time, as well as the decision about the handling or
therapeutic, is in itself based on the empirical knowl-
edge of each professional of health. Frequently, the
inherent information concerned with the process is
not analyzed, and the results are not registered in sys-
tematized and adequate formats to a future process-
ing. The monitorization assumes, therefore, an im-
portant role in the course of this trial, once it would
permit to collect helpful and valid information for the
creation of models to support the diagnosis and the
medical decision process.
In the area of the Clinical Information Systems,
we must highlight the EMR and the Laboratory In-
formation System, as part of the platform AIDA that
is today a reality within the Hospital Geral de Santo
Ant´onio (HGSA), that has acted in collaboration with
the University of the Minho. It makes possible to col-
lect in real time data from laboratory trials, patient
and clinical information (e.g., medical images).
The making of data or knowledge bases with lab-
oratory and clinical information, enabled with AIDA,
complemented with information about the diagnosis
and the trial decisions, allow for the Knowledge Dis-
covery and Date Mining techniques in use, leading to
the induction of formal models more adjusted to the
patient profiles, and to the admission context.
5 THE IMPLEMENTATION
5.1 Preamble
The implementation plan of the registration system
for infection control in HGSAs Intensive Care Unit
(ICU), emerged from the necessity of such unit of
monitorizing and accessing quality of health care pro-
vided. In order to accomplish the proposed task, it
was fundamental to acquire specific knowledge re-
lated with epidemiologic vigilance. Since there was
WEBIST 2008 - International Conference on Web Information Systems and Technologies
266
no institutionalized pattern for structuring the registry
form, operational parameters selection was performed
in three separate ways:
Discussion sessions with the ICU clinical team;
Bibliographic research;
Operational ”catch up” with similar initiatives in
other clinical facilities.
Since HGSA is also engaged in taking
part in Helics III project (see http:helics.univ-
lyon1.fraboutHELICSIII.htm), it was of major
importance to maintain the compatibility between the
selected terminologies. The unique symbiosis pro-
vided by the conceptual approach behind the EMR at
HGSA, granted enormous easiness in implementing
the referred system. It utilizes XML as an integration
media, built on top of a relational database. This
way, realization itself was not the hardest task. It
uses a simple xml skeleton which contains specific
formatting parameters to be interpreted by an external
application or viewer. Different combinations of
GUI Widgets can be used like combo and check
boxes, labels, text fields, etc. This provided unique
flexibility to the structure, while still maintaining
a familiar user interface for the clinician. Such
approach has also allowed restricting possible choice
values, enabling data uniformity and, consequently,
the accurate storage methodology in order to offer
useful or meaningful information for comparative
and statistical studies.
Due to the extent of the covered clinical parame-
ters, it was decided to divide the main structure into
two separate systems:
The General Record;
The MRB (Multiresistant bacteria) Individual
Record.
5.2 General Record for the NI Control
As the system was planed in order to follow patient’s
ICU internment, record field line up and arrangement
were chosen following clinician’s own registry order.
The first part of the registry contains information re-
lated with patient’s internment like origin, possibility
of infection’s presence already at admission, previ-
ous hospitalization, antibiotic therapy, etc. All data
already registered and available through EMR is au-
tomatically retrieved as a mean to avoid information
redundancy.
As described above, there was constantly the con-
cern to reduce record subjectivity to a minimum, re-
straining, when feasible, the clinical registry to a de-
fined set of possible values.
The second part of the form covers clinical ar-
guments that are registered during the first days of
internment like the SAPS II ((Gall et al., 1993))
and Glasgow indexes, imunodepression subsistence,
trauma, etc. Lastly, the rest of the structure accounts
for controlling the use of invasive devices (intubation,
central venous catheter, urinary catheter) and inci-
dence of infection. In addition to the usual recording
of first device application, the structure can store se-
quent similar proceduresfor future data mininganaly-
sis. It also previews simple colonization recording. In
terms of infection, the main sites have been covered
(urinary infections, surgical site infections, nosoco-
mial pneumonia and nosocomial bacteraemia). All
intermediary registries for each infection were para-
meterized under Helics III specifications so that maxi-
mum compatibility could be provided,foreseeingdata
sharing with European common database initiative
(Helics III) (Figure 3).
Figure 3: General Record for the NI control.
A WEB-BASED SYSTEM TO REDUCE THE NOSOCOMIAL INFECTION IMPACT IN HEALTCARE UNITS
267
5.3 The Registry for MRB Control
Antibiotic resistance stands for the ability of a mi-
croorganism to withstand the effects of an antibiotic.
This is recurrently accomplished via natural selec-
tion through random mutation and can be hastened
by a deficient politic for antibiotic usage. In order
to minimize this serious issue, an electronic record
form was created so that antibiotic resistance of sim-
ilar pathogens could be compared through time. This
way, specific data can be compared in order to rethink
antibiotic application actions. A scale of four possible
valueswas chosen for antibiotic resistance categoriza-
tion (Sensible, Intermediary, Resistent and unknown),
embedded into a XML format.
The structure also permits recording of some spe-
cific recognition methods like Screening or CMI (E-
test), used in VRSA (Vancomycin-resistant Staphylo-
coccus aureus) detection.
6 CONCLUSIONS
Future work will consider the knowledge discov-
ery about the nosocomial infections, on the basis of
guidelines, consensus, refereed articles, and infor-
mation registered on the EMR, crossing these data
with the complementarymeans of diagnosis - analytic
and, when applicable, imagiological; inducing patient
models for the early detection of nosocomial infec-
tions, with high acuteness and more adjusted to the
patients and to the environmentof the healthcare unit;
prompting models for the process of decision mak-
ing, and to a better selection of handlings. On the
other hand, we also intend to specify, to develop and
to implement a data processing computational system
starting from an online structured and systematized
recording of information about the nosocomial infec-
tion, enabling the recording of the diagnosis that were
effectively done, the register of the flat therapeutic
being prescribe, the interoperation between the lab-
oratory systems and the clinical information ones and
the implementation of prevention systems and, ac-
cordingly, of programs to improve and to measure the
quality of service.
ACKNOWLEDGEMENTS
We are thankful to the Hospital Geral de Santo
Ant´onio (HGSA), in Oporto, Portugal, for their help
to the analysis and development of the EMR system
referred to above, which is now being largely used in
their premises.
REFERENCES
Abelha, A., Analide, C., Machado, J., Neves, J., and No-
vais, P. (2007). Ambient intelligence and simulation
in health care virtual scenarios. In Camarinha-Matos,
L., Afsarmanesh, H., Novais, P., and Analide, C., ed-
itors, Establishing the Foundation of Collaborative
Network. Springer.
Alves, V., Machado, J., Abelha, A., and Neves, J. (2005).
Agent based decision support systems in medicine. In
WSEAS on Biology and Biomedicine, Issue 2, Volume
2.
Analide, C., Novais, P., Machado, J., and Neves, J. (2006).
Quality of knowledge in virtual entities. In Coakes, E.
and Clarke, S., editors, Encyclopedia of Communities
of Practice in Information and Knowledge Manage-
ment. Idea Group Reference.
Bradshaw, J. (1997). Software agents. Cambridge, MA:
MIT Press.
Costa, R., Neves, J., Novais, P., Machado, J., Lima, L., and
Alberto, C. (2007). Intelligent mixed reality for the
creation of ambient assisted living. In Neves, J., San-
tos, M., and Machado, J., editors, Progress in Artifi-
cial Intelligence, volume 4874. LNAI, Springer.
Gall, J., Lemeshow, S., and Saulnier, F. (1993). A New Sim-
plified Acute Physiology Score (SAPS II) Based on a
European/North American Multicenter Study. JAMA.
Hendler, J. (1996). Intelligent Agents: Where AI meets In-
formation Technology, Special Issue. IEEE Expert.
Machado, J., Abelha, A., Neves, J., and Santos, M. (2006).
Ambient intelligence in medicine. In Proceedings
of the IEEE-Biocas 2006, Biomedical Circuits and
Systems Conference, Healthcare Technology, Imperial
College, London, UK.
Machado, J. and Alves, V. (2005). Web-based simulation in
medicine. In Feliz-Teixeira, J. and Brito, A., editors,
MODELLING AND SIMULATION 2005.
Maes, P. (1995). Modeling Adaptive Autonomous Agents.
Artificial Intelligence Magazine.
of Medicine, I. (1991). The computer-based patient record:
an essential technology for health care.
Weed, L. (1969). Medical records, medical education, and
patient care. The problem-oriented record as a basic
tool. Case Western Reserve University, Cleveland,
OH.
Weiss, G. (1999). Multiagent Systems: A Modern Approach
to Distributed Artificial Intelligence. Cambridge, MA:
MIT Press.
WEBIST 2008 - International Conference on Web Information Systems and Technologies
268