MODELLING A PORTABLE PERSONAL HEALTH RECORD
J. Santos
1,2
, T. Pedrosa
1,3
, C. Costa
1
and J. Oliveira
1
1
DETI/IEETA, Universidade de Aveiro, Campus Universitário de Santiago, Aveiro, Portugal
2
DEE, Instituto Superior de Engenharia de Coimbra (ISEC), Rua Pedro Nunes, Coimbra, Portugal
3
Instituto Politécnico de Bragança, Bragança, Portugal
Keywords: Personal health record, Electronic health record, Patient empowerment, Portable personal health record,
Personal health in formation.
Abstract: Active and responsible involvement of patients in their own health is accepted as an important contribution
towards an increased quality of health services in general. Management of Personal Health Information by
the patient can play an important role in the improvement in quality of the information available to health
care professionals and as a means of patient involvement. Electronic Health Records are a means of storing
this kind of information but their management usually falls under the responsibility of an institution and not
on the patient himself. A Personal Health Record under the direct control and management of the patient is
the natural solution for the problem. When implemented in a storage hardware portable device, a PHR,
allows for total mobility. Personal Health Information is very sensitive in nature so any implementation has
to address security and privacy issues. With this in mind we propose a structure for a secure Patient Health
Record stored in a USB pen device under the patient’s direct management and responsibility.
1 INTRODUCTION
Personal Health Information (PHI) is generated
dispersedly (Tang, et al, 2006) and stored under
different formats. Modern diagnose and treatment
techniques generate great amounts of PHI and
population ageing can only increase it. This poses
the problem of efficient usability and management.
Many PHI management strategies have been
proposed (Costa, 2004) namely Electronic Health
Record (EHR) systems (Classen, et al., 1991, 1992;
Evans, et al., 1992; Schoenbaum, et al., 1992;
Tierney, at al. 1993) that have some limitations.
Interoperability between different systems is limited
(Mandl, et al., 2001; Schonberg, et al., 2000), most
contain only institution-generated data and
sometimes only at a departmental level (Hasselbring,
1997). Some information supplied by the patient is
incorporated in the EHR by healthcare professionals,
but making patients responsible and active in their
own health, is not a concern of such systems since
an EHR is usually managed by an institution. That
goal can only be achieved through the use of a
Personal Health Record (PHR). This is stressed by
the generally accepted definitions of EHR and PHR
provided by the Markle Foundation (2004).
People are increasingly mobile. A PHR on a
portable device that can also be a repository of
emergency data, integrated with an EHR system is
the best architecture (Detmer, et al., 2008), since it is
standalone, closely integrated with external data
sources and allowing for mobility (Román, et al.,
2006). In this paper we propose the structure for a
portable PHR (pPHR) with main objectives:
- Allow the patient full management capabilities;
- Populated by the patient from various sources;
- Availability to third parties in different contexts.
Such an implementation relies on:
- Strong user authentication;
- Secure data encryption;
- Separate data storage depending on data type;
- Electronic signatures for integrity.
2 BACKGROUND
Current PHR implementations include network-
PHRs with decision making tools such as the ones
offered by Microsoft and Google, both promising
the creation of systems and standards for automated
465
Santos J., Pedrosa T., Costa C. and Oliveira J. (2010).
MODELLING A PORTABLE PERSONAL HEALTH RECORD.
In Proceedings of the Third International Conference on Health Informatics, pages 465-468
DOI: 10.5220/0002746504650468
Copyright
c
SciTePress
data collection and updating, allowing for patient
mobility since data is accessible from any place with
Internet access. But they have some limitations since
concerns may be raised as to the security and
privacy of data contained in a system over which the
patient has no control. The patient controls which
information is on the system but not the security and
privacy issues pertaining to the system’s internal
design or the information disclosure policies
enforced by the owning companies.
A pPHR overcomes all the mentioned limitations
providing full mobility to its users through regular
backups, strict access control policies, cryptography,
and detailed audits of each activity.
Among the various candidates to a portable PHR
device (Costa, et al., 2003), flash USB pens present
the most advantages since they can be carried
around, their use has been proposed for developing
countries (Srinivasan, et al., 2007) and they can also
be used in developed countries. Security issues have
been raised concerning PHR implementations in
these devices (Wright, et al., 2007a, b) but they can
be overcome with careful design.
3 PROPOSED STRUCTURE
The basic paradigm of the pPHR is total patient
empowerment. Our model also allows patient
mobility and provides an emergency data repository.
The pPHR is a complete and fully functional PHI
management platform where a repository of the
lifelong patient’s history is kept based on close
interoperability with all the participating institutions
where an individual’s health data is generated. The
pPHR is the most complete repository of an
individual’s PHI. In the limit, any external EHR
system can have as much information, but none is
meant to have more. Any access to the pPHR and its
data depends exclusively on patient’s authorization.
In order to allow the patient to both manage the PHI
contained in the pPHR and to make it available to
other healthcare actors in a scalable manner, various
operating modes are available upon patient’s option.
Population of the pPHR is done by the patient in two
different situations, in person and remotely via an
Internet connection.
Patient authentication for account provisioning is
required for pPHR issuance, authentication method
depending on the prior relationship of the issuing
institution with the patient. The device is supplied
with a built in master-password that will be used for
the initial setup of pPHR native applications, upon
which the user is prompted to create his working-
password. The system only allows three tries of
logging. Upon failure of the three attempts, the PHR
and all its functionalities are blocked and will only
be unblocked via the master-password. Whenever
the patient accesses the PHR, he will be prompted to
supply the password and to select an operating
mode. An inactivity timer blocks the system after a
time interval of inactivity that is preset but
configurable by the patient. This inactivity lock can
be unlocked via the password. One of the key
concepts of our model is the existence of different
conceptual data types. There are five different data
types:
- Confidential pPHR Data: information considered
to be extremely sensitive and to be kept secret.
- Normal pPHR Data: information to be disclosed
to users in general.
- Transfer pPHR Data, recently downloaded into
the PHR waiting for patient’s management.
- Prescription pPHR Data, typically generated by
physicians to be read by other users.
- Emergency pPHR Data, publicly available in
case of emergency.
Figure 1 shows the data types in usage context and
Table 1 expresses the different information access
privileges by all the actors involved. Through the
implementation of this working scenario we achieve
free and easy access to emergency information.
Prescription PHR Data is still freely available but
not displayed because it might be relevant in an
emergency situation. Access to prescription
information and the possibility of writing Tranfer
pPHR Data are given. All actions taking place in the
pPHR are recorded and become part of an audit
report that keeps track of all information transactions
and the respective dates in which those actions took
place. The data management functionality is
implemented as document classification or
document drag and drop strategies.
All data items are electronically signed by their
author for information integrity but, in case they
aren’t, the pPHR provides a functionality that allows
the patient to act as a delegated signer. For efficient
storage space management, the individual files
placed inside each container are compressed with a
choice of “deflate” or “bzip2” algorithms. A
compression level can be chosen for each file, no
compression being also a possibility.
HEALTHINF 2010 - International Conference on Health Informatics
466
Table 1: Information Type Access Privileges of the various pPHR users.
Information Type
Actor
Confidential Normal Transfer Prescription Emergency
Patient Read/write Read/write Read/write Read/write Read/write
Physician - Read Read/write Read/write Read
Social Institution - Read Read/write Read Read
Researcher - Read Read/write Read Read
Laboratory - - Write Read Read
Pharmacy - - Write Read Read
Non-qualified - - Write Read Read
Figure 1: Data types contained in the pPHR.
The pPHR is completely pen-resident and
consists of a stand-alone application with various
functionalities, one of which allows access to a
virtual container where a database and the various
data items are stored. The virtual container is
implemented in a zip-like manner (ZipArchive) with
compression and file encryption via a soft-coded
encryption key. AES encryption algorithm and the
fact that, upon access by the application, the files are
decompressed to computer memory only are
important features contributing for enhanced
security. The database is built upon SQLite and
contains entries that are pointers to individual data
items.
Automatic Zip64 extensions allow upgrade of
maximum size, maximum size of individual files in
an archive, and the number of files in an archive.
Each file inside an archive is encrypted by the
Advanced Encryption Standard (AES) algorithm
with a 256-bit key encryption key. The access to any
individual file stored inside a virtual container is
subject to authentication and the decompression of
the files is made to computer’s memory only for
improved security.
4 CONCLUSIONS
PHI has been dramatically increasing raising the
issue of efficient information management. EHR
systems have been implemented but they have some
limitations that can be overcome by the use of a
portable PHR in a USB pen device that
complements existing EHR systems.
Our model totally empowers the patient with
respect to his PHI. He decides which data becomes
part of the record, who, and under what
circumstances, has access to it. The pPHR is fully
interoperable and populated with information from
various external sources as well as self-entered PHI.
Data is kept in separate conceptual data containers.
There are five different data types, each of each
of which with different security and confidentiality
characteristics.
Some concerns have been expressed regarding
the trust a physician can place on patient-generated
information. By the use of pPHR, the patient is
invited to better understand his own health and, by
being able to manage it, his increased health
awareness can only be beneficial. The pPHR is
meant to be used with an EHR, therefore the
Default
Emergency Info
-------------------
-------------------
-------------------
Prescription
Data
Click to View
Transfer Data
Click to Input
Social Ph
y
sician
Prescription
Data
View/Input
Emergency
Data
Click to View
Transfer Data
Click to Input
Normal Data
-------------------
-------------------
-------------------
Prescription
Data
View/Input
Emergency
Data
Click to View
Transfer Data
Click to Input
Normal Data
-------------------
-------------------
-------------------
MODELLING A PORTABLE PERSONAL HEALTH RECORD
467
traditional interactions the physician is used to
establish with his patients can still be enforced but
the later being more informed and collaborative.
There is some overlapping of disclosure and
write privileges concerning Emergency and
Prescription Data Types. From a technical
perspective it’s arguable that they could be grouped
into a single type but from an organizational and
usability perspective they are different. They both
are possibly disclosed to most actors but Prescription
data is to be generated by the physician and not
necessarily to be disclosed in an emergency situation
(although possible). The patient can write into that
area (although he shouldn’t) because nobody should
have better access privileges to any part of his pPHR
and since every data item is to be electronically
signed, there is no confusion possible between
physician-generated data and eventual patient-
generated data.
REFERENCES
Tang, P., et al. 2006. Personal Health Records:
Definitions, Benefits and Strategies for Overcoming
Barriers to Adoption. Journal of the American Medical
Association, 12, pp.121-126.
Costa, C., 2004. Doctoral Thesis: A Security Dynamic
Model for Healthcare Information Systems, in
Departamento de Electrónica e Telecomunicações,
Universidade de Aveiro.
The Markle Foundation, 2004. Connecting Americans to
their Healthcare: Final Report.
Classen, D. C. et al., 1992. Computerized Surveillance of
Adverse Drug Events in Hospital Patients. Journal of
the American Medical Association, 266, pp.2847–
2851.
Classen, D. C., et al., 1992. The Timing of Prophylactic
Administration of Antibiotics and the Risk of
Surgical-wound Infection. New England Journal of
Medicine, 326, pp.281–286.
Evans, R. S., et al., 1992. Computerized identification of
patients at high risk for hospital-acquired infections.
American Journal of Infectious Control, 20, pp.4–10.
Schoenbaum, S. C., Barnett, G. O., 1992. Automated
Ambulatory Medical Records Systems: An Orphan
Technology. International Journal of Technical
Assessment in Health Care, 8, pp.598–609.
Tierney, W. M., et al., 1993. Physician Inpatient Order
Writing on Microcomputer Workstations: Effects on
Resource Utilization. Journal of the American Medical
Association, 269, pp.379–383.
Mandl, K., Szolovits, P., Kohane, I., 2001. Public
Standards and Patient’s Control: How to Keep
Electronic Medical Records Accessible but Private.
British Medical Journal, 322, pp.283-287.
Schonberg, R., Safran, C., 2000. Internet Based
Repository of Medical Records that Retains Patient
Confidentiality. British Medical Journal, 321, pp.
1199-1203.
Hasselbring, W., 1997. Federated Integration of
Replicated Information Within Hospitals. International
Journal on Digital Libraries, 1(3), pp.192-208.
Román, I., et al, 2006. Demographic Management in a
Federated Healthcare Environment. International
Journal of Medical Informatics, 75 (9), pp.671-682.
Detmer, D., et al, 2008. Integrated Personal Health
Records: Transformative Tools for Consumer-centric
Care. BMC Medical Informatics and Decision
Making, 8, p.45.
Costa, C., et al, 2003. A New Concept for an Integrated
Healthcare Access Model, Proceedings of MIE 2003
(Medical Informatics Europe). The New Navigators:
from Professionals to Patients, Studies in Health
Technologies and Informatics, 95, pp.101-106, IOS
Press Book Series (ISBN: 1-586-03347-6).
Srinivasan, U., Datta, G., 2007. Personal Health Record
(PHR) in a Talisman: An Approach to Providing
Continuity of Care in Developing Countries Using
Existing Social Habits. 9
th
International Conference on
e-Health Networking, Application and Services,
pp.277-279.
Wright,A., Sittig, D., 2007. Security Threat Posed by
USB-based Personal Health Records. Ann Intern Med,
146, pp.314-5.
Wright,A., Sittig, D., 2007. Encryption Characteristics of
2 USB-based PHR Devices. Journal of the American
Medical Association, 14, pp.397-399.
HEALTHINF 2010 - International Conference on Health Informatics
468