HEALTHCARE INFORMATION SYSTEMS IN A NETWORK
ENVIRONMENT
An Interoperability Model and a Real Case in a Pathology Network
Emanuele Baj, Paolo Locatelli, Ilaria Marzolla, Nicola Restifo
Fondazione Politecnico di Milano, Via Garofalo, 39, 20133 Milan, Italy
Silvia Bragagia, Gianni Origgi
A.O. Ospedale Niguarda Ca’ Granda – S.C. ICT, Piazza Ospedale Maggiore, 3, 20162 Milan, Italy
Keywords: eHealth, Healthcare Management Systems, Electronic Health Record, Semantic and Syntactic
Interoperability, Pathology Networks.
Abstract: Interoperability between Healthcare Management Systems is a fundamental requirement to improve the
quality of care, to assure patients continuous assistance and to improve clinical research. Pathology
Networks create a common virtual environment which allows to overcome single HIS’ borders, by
connecting different healthcare organizations adopting common formats to share clinical data and,
consequently, knowledge. Syntactic and semantic standards are paramount to enable interoperability in this
context, since common document formats ensure efficient communication, while internationally known
vocabularies allow the unambiguous interpretation of contents. This paper has the purpose to show a real
experience of participation to a Pathology Network, providing a model based on open standards which
could be taken as a reference model for similar initiatives in other healthcare environments. Benefits
resulting from such an initiative will be presented, in terms of improvement for clinical research and quality
of patients’ care. The real case we describe is an example of how Pathology Networks can provide scientific
features to Electronic Health Records, by adding information not usually reported on standard clinical
sheets, but necessary to perform new analysis with the purpose of improving clinical research.
1 INTRODUCTION
Interoperability between clinical information
systems represents one among the major
opportunities of health informatics. The spreading
of many solutions in the field of Electronic Health
Records provides a great amount of digital data
about patients, potentially available to feed scientific
research in order to improve the efficiency of
clinical processes (Chauldry et al., 2006). An
interoperable infrastructure to enable data sharing is
paramount to best exploit the potential represented
by digital information: daily activities performed by
physicians and nurses are sources of a great amount
of data, whose value in terms of clinical and
scientific research could be enhanced if handled and
stored in a standard formats. In order to allow this
interoperability, it is necessary that syntactic and
semantic standards are provided: the former are
needed in order to assure a common representation
to the data which are transferred, while the latter are
required to assure unambiguous interpretation of
their meaning.
This direction is being followed by initiatives
known as “Pathology Networks”, whose aim is
creating a cross-organizational virtual database for
patients affected by a same disease.
This paper analyzes the concept of Pathology
Networks, their features and purposes, also by
presenting a reference model which could be taken
as a basis for such initiatives in different healthcare
contexts. The case study we describe is the solution
implemented at “Fondazione IRCCS Istituto
Neurologico Carlo Besta”.
461
Baj E., Restifo N., Marzolla I., Restifo N., Bragagia S. and Origgi G. (2010).
HEALTHCARE INFORMATION SYSTEMS IN A NETWORK ENVIRONMENT - An Interoperability Model and a Real Case in a Pathology Network.
In Proceedings of the Third International Conference on Health Informatics, pages 461-464
DOI: 10.5220/0002744404610464
Copyright
c
SciTePress
2 A REAL CASE: EPINETWORK,
THE EPILEPSY NETWORK OF
LOMBARDY REGION
In line with interoperability objectives fixed by
Italian Ministry of Health, the Region of Lombardy
has laid the foundations for the creation of four
Pathology Networks, concerning Rare Diseases,
Epilepsy, Haematology and Oncology. These
networks enhance contact among clinical and
research centres working on the same pathologies,
provide a coordinated assistance to patients, helping
to ensure the delivery of high quality care processes,
empowering clinical research and spreading the
outcomes among the different actors.
Among the various initiatives, adherence to
Epinetwork, the Epilepsy Regional Network, is
being analyzed at Fondazione IRCCS Istituto
Neurologico Carlo Besta in a research supported by
the Italian Ministry of Health within the project
“Web Integrated Information System for the
management of clinical and research activities in the
field of the neuroscience and the implementation of
individual electronic health record”, coordinated by
the Institute and with the participation of A.O.
Ospedale Niguarda Ca’ Granda and Fondazione
Politecnico di Milano. Istituto Besta is an
internationally renowned centre of excellence in the
care of neurological diseases and, among them,
Epilepsy: physicians and researchers of the Institute
worked in a team of experts coordinated by the
Region of Lombardy in order to define semantic
standards needed to identify an epileptic patient and
to provide all information required to follow his care
process. A.O. Ospedale Niguarda, together with
Fondazione Politecnico di Milano, a research and
innovation partner connected to the Politecnico di
Milano technical university, is working at a
technical solution for the Institute, by exploiting the
methodology already followed during previous
experiences in the development of a reference
solution for the Hematologic and Oncologic
Regional Networks.
The solution will exploit the features provided to
the Electronic Health Record (EHR) which the three
partners are already realizing for the Institute
(Locatelli et al., 2009) ; each feature was analyzed
and developed trying to produce as much structured
data as possible, with the purpose of creating an
operative tool aiming at supporting ordinary
activities in a complex system able to provide useful
information to feed scientific research. Among
Besta’s scientific pipeline, Epilepsy represents an
opportunity to create a structured methodology to
collect data, which overcomes the Institute’s
borders, thanks to Epinetwork regional initiative.
3 A REFERENCE MODEL FOR
PATHOLOGY NETWORKS
A Pathology Network is the concretization of an
“inter-corporate virtual department” in which
operators involved in the same pathologic treatment
can profitably contribute to the process of diagnosis
and care, even if operating at different geographical
sites. This way of cooperation envisages, within a
wide domain, that highly specialized medical hubs,
qualified for high complexity treatments – as
intensive chemotherapy, stem cells transplant and
haemorrhage emergency management - and smaller
peripheral institutions could not only share and
exchange knowledge and opinions but, above all,
publish on the Pathology Network all medical
reports produced after patient’s treatment.
By following well known international standards,
it will be possible to study homogeneous data
collections, i.e. to analyze a wide range of disease
cases monitoring the results of new clinical
experimentations, or to compare different treatment
schemes with the purpose of defining common care
pathways for specific diseases and establishing
prevention practices (Tang et al., 1994).
The keystone in a Pathology Network is
represented by the implementation of a
technological platform linking clinical databases of
all hospitals involved in the network and enabling
the exchange of clinical data (e.g. patient cases,
scientific knowledge, etc.). This platform should be
carried out in a flexible way, in order to interconnect
the heterogeneous clinical data repository of each
nodes, realized with different technologies, various
information systems and data management tools.
Data sharing for each clinical episode takes
place only at a logical level, not at a physical one:
this means that clinical reports are not sent to or
received from hospitals but what is really shared is
the logical link by means of which each document
can be univocally retrieved by a HIS at its physical
location.
Italian Region of Lombardy is bringing forward
initiatives in this sense, with a centralized structure
represented by the extranet SISS (Italian acronym
for Social Healthcare Information System) and a
central domain embodied by an EHR called FSE, a
database collecting data related to the whole clinical
history of a citizen registered to the National
Healthcare Service (SSN). The cooperative logic
HEALTHINF 2010 - International Conference on Health Informatics
462
characterizing the FSE structure lets each subject
involved in the network act both as supplier and user
of clinical documents, at the same time.
FSE database feeding is achieved by publishing
clinical electronic documents (called DCE) produced
by qualified healthcare organization as PDF files to
which physicians’ digital signature is applied. In
order to give legal value to digitally signed reports,
the repository keeps a clear reference of the
versioning of the document collected.
Various DCE documents exist, concerning
outpatients’ examination results, specialized
radiology or laboratory reports, accidents and
emergency reports, inpatients’ discharge letters,
evaluation notes, patients’ summaries, phone contact
reports and tele-consulting results.
Figure 1: Interoperability model for Pathology Networks.
Based on these assumptions, the Pathology
Network database is fed with a DCE subset, made
up of predefined sets of data, codified into XML
tags, suitable to collect various data formats.
Epinetwork solution is based on a syntactical
structure defined inside a XML CDA2 (Clinical
Document Architecture Rel.2) standard. This
standard, besides being the reference solution for
exchanging healthcare information at international
level, helps us adding efficiency to interoperability:
persistence and human-readability assure that all
different recipients of the information are able to
receive and properly interpret stored information;
moreover, CDA specification is so rich and flexible
that doesn’t place any limitation to future definition
of new document typologies (Dolin et al., 2001).
The XML document allows physicians to fill in
specific pathology-related fields, such as diagnosis,
codified with the international classifications ICD 9
and ILAE (International League Against Epilepsy)
(OREp, 1996), causal factor and assessment over
Epilepsy’s severity, both semantically codified.
Other information are required, especially
concerning drugs, EEGs and other exams or
demanded performances.
Pathology Networks imply some specific needs
in terms of privacy respect and data security: in
order to ensure that only authorized subjects can
access data, some fundamental concepts regarding
report’s publication have been introduced, which are
DAO, GAN and “consolidated CDA”. A DAO -
Italian acronym for “Authorization and Obfuscation
Document” - defines if the DCE is authorized, i.e.
freely and autonomously consultable by the patient
and if it must be obfuscated, i.e. not visible by
healthcare operators due to a voluntary patient’s
deny or in case of special legal cases (HIV,
voluntary termination of pregnancy, drug addiction,
abuse), except in case of emergency assistance.
GAN document - whose acronym stands for
“Authorization and Designation Management” -
compiled by the patient at his first contact with the
hospital and then digitally signed by a clinical
responsible, specifies who inside the Pathology
Network is authorized to consult the DCE, with the
understanding that access to patient’s data are
restricted to clinical information correlated to the
specific pathology treated.
A “consolidated CDA”, the real entity shared by
healthcare organizations, is made up of the digitally
signed DCE (P7M format), the clinical report in
XML-CDA2 format, and the XSL file containing
presentation rules; this set is then compressed into a
ZIP archive, encoded with Base64 system and
finally inserted as a field of an HL7 message, which
will then be shipped over the network through a
manager.
The procedure applied for feeding the Pathology
Network is carried out in a semi-automatic way from
structured clinical documents contained in the
Health Information System’s local repository,
following the guidelines “Integration Profile”
defined by IHE (IHE International, 2003).
When requesting storage of a Pathology Network
document, such as an Epilepsy report, a form is
opened soliciting the physician's digital signature,
realized with a card identifying the user as an
operator enabled to publish documents over the SISS
platform: signature enables the creation of the
Epilepsy structured report. Then, a first transaction
“Provide & Register Document Set” takes place
towards the clinical organization’s repository
producing the document; due to platform’s
constraints in terms of control over documents’
formal correctness, an employee owning a SISS card
HEALTHCARE INFORMATION SYSTEMS IN A NETWORK ENVIRONMENT - An Interoperability Model and a Real
Case in a Pathology Network
463
with super-user rights is assigned to collect all
reports stored into local database and not yet
published, control that related DAO documents are
authorized and not obfuscated, and finally mark
them with a timestamp. At this point, the “Register
Document Set Document Registry” transaction can
be processed, which effectively publishes the report
into FSE database.
Figure 2: Feeding and retrieval process of a Pathology
Network.
In case an Epinetwork member asks for patient’s
clinical documents consultation, a Query Registry is
invoked over FSE Document Registry: services
offered by SISS infrastructure control if requester is
allowed by the GAN and, if so, generate a Retrieve
Document transaction to show the epilepsy report to
the final user through his local visualization tool.
4 DISCUSSION
The aim of this paper is to show a real solution
adopted to participate in a pathology network and to
show the positive effects it could have inside a
healthcare environment. The solution realized at
Besta represents a particular implementation of a
general model already applied to the Hematologic,
Oncologic and Epileptic Pathology Networks in the
Region of Lombardy, and it could be taken as a
reference methodology for similar initiatives in
different environments. Generality and reusability of
the solution are ensured by technological choices:
first of all, the interoperable platform is supplied by
a unique provider (in this case the Region), which
takes care of all aspects concerning data security and
communication between repositories; moreover, the
adoption of international syntactic standards ensures
correct interpretation of all data published by various
participating systems.
Advantages experienced in taking part in a
Pathology Network can be felt both at clinical and
scientific level, since clinical researchers are
provided with a great amount of high reliable data,
on which more efficient analysis can be performed,
contributing to support clinical decisions, providing
a general improvement in diseases’ care processes
and a better service level to patients.
Pathology Networks could be seen as a
revolutionary concept in healthcare, since it moves
the focus from Health Information Systems to Web
integrated Healthcare Information Systems.
Networks allow to overcome single organizations’
borders, enabling extended co-operation via Web
between healthcare professionals working in
different structures: this represents an opportunity to
share experiences and to join together different
approaches towards the care of diseases, with
undeniable benefits for scientific progress and the
quality of assistance delivered to patients.
REFERENCES
Chauldry, B., Wang, J., Wu S., et al, 2006. Systematic
Review: Impact of Health Information Technology on
Quality, Efficiency, and Costs of Medical Care.
Annals of Internal Medicine, 2006;144(10):742-752.
Locatelli, P., Baj, E., Origgi, G., Bragagia, G., 2009.
Medical tutorial: porting of a clinical portal between
healthcare organizations - Reuse of an Application in
Health Informatics. Proceedings of HEALTHINF 2009
- Second International Conference on Health
Informatics, L. Azevedo & A. R. Londral, Eds., pp.
375-380, Porto, Portugal: INSTICC Press. January
2009
Tang, P.C., Fafchamps D., Shortliffe, E.H., 1994.
Traditional Hospital Records as a Source of Clinical
Data in the Outpatient Setting, Knowledge Systems
Laboratory.
Dolin, R.H., Alschuler, L., Boyer, S., Beebe, C., Behlen,
F.M., Biron, P.V., Shabo, A., 2001. The HL7 Clinical
Document Architecture, Release 2. The Journal of
American Medical Informatics Associations, Volume
13 Number 1 Jan / Feb 2006; 30-39.
Osservatorio Regionale per L'Epilessia (OREp), 1996.
ILAE classification of epilepsies: its applicability and
practical value of different diagnostic categories.
Epilepsia. 1996; 37(11):1051-9
IHE International, 2003. Integration Profiles: The Key to
Integrated Systems. Available at: www.ihe.net.
Accessed September 3, 2009
HEALTHINF 2010 - International Conference on Health Informatics
464