MEDINFO 2007

K. Kuhn et al. (Eds)

IOS Press, 2007
2007 The authors. All rights reserved.

Principles-Based Medical Informatics for Success How Hong Kong Built One of
the Worlds Largest Integrated Longitudinal Electronic Patient Records
Ngai-Tseung Cheunga, Vicky Funga, Wing Nam Wonga, Anna Tonga, Antonio Seka, Andre
Greylingb, Nancy Tseb, Hong Fungc
a

Health Informatics Section, Hospital Authority, Hong Kong SAR

Information Technology Division, Hospital Authority, Hong Kong SAR
c
New Territories East Cluster, Hospital Authority, Hong Kong SAR

or patient documentation [1]. The Electronic Patient

Record (ePR) was developed in 2000 to provide both a
standardized repository of all the clinical data collected
throughout the HA, as well as a clinician-friendly view
into a comprehensive longitudinal lifelong record. The
CMS and ePR are integrated across all settings inpatient,
outpatient and emergency, and have also been the platform
for development of all subsequent clinical modules,
including modules for different clinical specialties, the
allied health disciplines and nursing.

Abstract
Since 1994, the Hospital Authority has been developing
and deploying clinical applications at its constituent 41
hospitals and 121 clinics. The Clinical Management System (CMS) is now used by over 4000 doctors and 20000
other clinicians on a daily basis to order, document and
review care. The territory-wide Electronic Patient Record
(ePR) has given clinicians an integrated, longitudinal, lifelong view of a patients record.
Today the CMS and ePR form an essential clinical and
management tool to the Hospital Authority. The CMS handles two million clinical transactions per day, and the ePR
has over 6TB of data covering 57 million episodes for 7.9
million patients.
This paper describes how the HA has taken a principlesbased approach to Medical Informatics to achieve its success in the enterprise-wide deployment and deep
utilization of a comprehensive clinical information system.

By the turn of the century, the CMS had already been

rolled out to all HA sites, and today the CMS and ePR are
an essential clinical and management tool, handling over
two million clinical transactions per day. The ePR contains
the records from 57 million episodes for 7.9 million
patients in a repository of over 6TB (not including radiology images) and receives 300,000 views per day. The total
expenditure on system development and implementation
of the entire clinical informatics portfolio has been under
US$200million.

Keywords:
electronic patient record, medical informatics

The wave that never breaks

Introduction

Such a wide and deep penetration of computerized patient

records is far from the norm. Despite promises of the
paperless or electronic hospital for over 30 years,
comprehensive adoption of such systems has not materialized in most countries electronic medical records are
the wave that never breaks [2]. The seminal Institute of
Medicine report on the Computer-based Patient Record [3]
released in 1991 shaped the thinking of a generation of
systems, but implementation remained low. A decade later
another pair of IOM reports on medical errors [4] and quality systems in medicine [5] has stimulated a fresh round of
interest in the electronic medical record as a key enabler
for quality improvement and error reduction. Other potential drivers include consumerism and the rise of personal
health records [6] (PHRs), and estimates of positive
returns from investments in electronic records, both at the
institutional [7] and national levels [8].

The Hong Kong Hospital Authority (HA) was formed in

1990 to manage all public hospitals in Hong Kong. Today
the Authority manages a Head Office, 41 public hospitals/
institutions, 47 specialist clinics and 74 general (primary
care) outpatient clinics. Each year in the HA there are over
one million inpatient episodes, 2 million emergency attendances and 13 million outpatient visits.

Clinical systems in the HA

In 1990 the HA was essentially a greenfield site, with
very little computerization of any sort, let alone clinical
computing. From 1991 onwards HA developed its basic
information technology infrastructure, including financial,
HR, patient administrative and departmental systems. In
1994 the HA began developing the Clinical Management
System (CMS), an integrated clinical workstation giving
clinicians access to all available electronic clinical information as well as providing direct entry of orders and care

Perhaps the age of ubiquitous clinical computing truly is

imminent. However there are many indicators that a long

307

N. Cheung et al. / Principles-Based Medical Informatics for Success

and hard road remains ahead. The Annual HIMSS Leadership Survey has shown an striking incongruence between
intent and achievement in installing electronic medical
record systems. In 2003 the survey showed that 19% of
healthcare organizations in the US had fully operational
electronic medical record (EMR) systems, and that 32%
had begun installation. [9] In 2006 the same survey found
that 24% now had fully operation EMR systems, and 42%
had begun installation. [10] That is to say, in the last three
years only a 5% increase in fully operational systems was
seen.

The Health Informatics Team in close collaboration with

the IT division runs the Clinical Informatics Program
Office (CIPO), which is the executive arm of the CIPSG,
and all developments, enhancements, collaborations and
other activities are filtered through CIPO.
This constitutes the complete top down / bottom up
approach to customer engagement in the HA, from
detailed consideration of needs at the working groups and
CIPO, to implementation decisions at CIPSG, to investment and strategic decisions at ITGC with constant
coordination and communication between the levels of
governance (figure 1).

Outside the US, some very well funded high profile

national initiatives (such as the UKs 6billion Connecting
For Health [11] and Canadas C$1.2billion Health Infoway
[12]) are making headway, but are still years away from a
nationwide EMR implementation at both hospital and outpatient settings.
So why is that in Hong Kong, with its comparatively modest expenditure on health IT, the wave has apparently
broken? Success in clinical information systems is multifactorial [13] , [14] and without doubt the nature of the
healthcare system in Hong Kong and the leadership of the
HA have played a major role. However this article will
focus on one aspect that has been constant throughout the
history of clinical systems development in Hong Kong
the integration of medical informatics into the leadership
and execution of the clinical systems programme, and a
consistent adherence to certain key principles and processes throughout this period.

Figure 1 Clinical Systems Governance Process

Win - win (and win - win)

Seven principles for medical informatics

success

It is often easy to create a IT-based program that provides

benefits to patients or the organization, but at a cost to the
clinicians asked to implement the program - often in the
form of extra work, changes in process or simply loss of
flexibility. Avoiding this problem by ensuring systems
benefit, or at least are cost-neutral, to all stakeholders is
vital to ongoing success. The four wins in this section
heading refer to the four major stakeholder groups 1) clinicians, 2) patients, 3) management and other secondary
data users and 4) health informatics and ITD who need to
ensure integrity of data and architecture.

The customer is always right (well, almost)

This old adage from the retail industry is of course not to
be taken literally, but points to the reality that meeting clinician needs is paramount [15] and that clinician resistance
is often the critical factor leading to the failure of clinical
system [16]. Focusing on the needs of the customer has
always been of prime importance in clinical systems
development in the HA. But to meet the needs of our customers, we must know who they are.

Key considerations in enabling common benefits are: can

the needs of individual clinicians be met with a standardized system? Can structured data be captured by clinicians
at the point of care without using more time and impinging
on clinical autonomy? Can the system adjust for different
practices and workflows? Can the data gathered by the
system flow onto management, quality assurance and
other such uses?

The clinicians using the system for care delivery are obviously the most direct customer. However, clinical and
executive management, policy makers, payers, researchers
and ultimately the patients are all customers of the clinical
systems. To fully engage this large body of stakeholders in
a manageable fashion, the HA has instituted formal Clinical IT Governance structures and processes. The key
governance committee is the Clinical Informatics Program
Steering Group (CIPSG), comprising nominated representatives from different hospitals, specialties, corporate areas
and other constituencies. Each representative chairs a
working group that looks at the detailed needs of that constituency, and reports back to CIPSG. The CIPSG in turns
reports up to the board-level IT Governance Committee
(ITGC).

The next two principles have been the key to answering

these questions in the affirmative.
One step at a time (success breeds success)
In the complex domain of medical informatics it is
extremely difficult, if not impossible to envisage a comprehensive solution to any problem de novo, and it is safe
to assume that unexpected issues will arise [17]. In the HA
big bang implementations are avoided wherever possi-

308

N. Cheung et al. / Principles-Based Medical Informatics for Success

ble in favor of a rapid prototype design cycle, followed by

one or two pilot sites with evaluation and redevelopment
where necessary, followed by phased rollout to the rest of
the organization.

sudden switch from an informatics-driven push of clinical systems to a clinician-led pull of additional
requirements. Today there are far more requirements than
can be met by available resources, in terms of functionality, hardware or support for specific clinical initiatives.

Success stories are vital to the long-term health of clinical

IT, and taking small steps allows continuous demonstration of success, developing a cadre of clinicians who have
seen the benefits and understand the issues. A stepwise
approach has also allowed us to make our mistakes earlier
and on a smaller scale, providing continuous learning
whilst minimizing the risk of catastrophic failure.

Demand management is now crucial, and all requests must

be evaluated, developed and either discarded or moved up
the governance chain until the appropriate level is reached
where a yea or nay decision can be made. At each level of
governance evaluation criteria and processes have been
put in place to enable an equitable approach which takes
into account the needs of patients, clinicians and the organization as whole.

Medicine is an art and a science (and so is medical

informatics)

Embrace your Informaticians (and feed them well)

Clinical systems sit at the uneasy interface between the

vagaries of human biology/disease and the scientific
requirements of quality management and evidence based
medicine. There is a constant tension between the need for
standardized, codified data and the clinical requirements
for flexibility and expressivity. It has been shown that
acceptance of structured entry is not uniform [18] and in
the HA the optimum balance between the two is a key
decision for any new module.

Implementing sophisticated clinical systems is a difficult

endeavour, and there is no shortage of literature examining
stumbling blocks [20], problems [21] and case studies of
failure [22].
The principles listed in this paper all require constant
attention, and a team that is able to navigate between clinical and organizational needs, technical limitations and
possibilities, standards and architectures, policies and
guidelines, change and demand management, and to do
this over long time periods. In the complex HA environment, and in any large care delivery setting, this is not a
task that can be done properly without dedicated, trained
informaticians.

Although the degree of structure is negotiable, the way the

data is codified is not all systems must conform to the
standards laid down in the HAs Information Architecture
[19]. However the design of systems must always try to
allow for flexibility whilst preserving structure.
Use it or lose it (data use begets data quality)

Conclusion

In the HA environment there are tens of thousands of users

of varying levels of IT sophistication entering data from a
variety of modules in different clinical settings. How do
we ensure the quality of this data? The main strategy taken
in the HA is to reuse the data wherever possible write
once, use many.

Healthcare operates in a sea of change social, environmental, scientific and technological change. Electronic
medical records are an increasingly important tool to
enable healthcare delivery organizations stay afloat and
perhaps prosper, but they can only do so with the necessary
commitment, skill sets and a adherence to disciplines such
as have been presented in this paper.

One obvious reuse of data is secondary uses, such as quality management, planning and analysis. Where clinicians
run their own quality or analysis programs they will soon
see the need for good data, and we have developed sophisticated, clinician friendly data mining tools to allow
clinicians to run their own ad hoc queries on the massive
data warehouse that has accrued over the years.

Acknowledgements
We would like to thank the members of Clinical Informatics Program Steering Group and all many others at the Hospital
Authority who have contributed their time and efforts to shaping
the CMS and ePR.

References

However opportunities also abound for reuse of data in

direct clinical care. The medications list on a discharge
summary can be copied from the discharge prescription.
The reason for doing a test can be drawn from the problem
list. Key lab results can be automatically copied into a diabetic documentation template. Data reuse of this sort is a
form of semantic interoperability albeit within but not
between complex clinical systems environments and the
Information Architecture provides the framework and
standards to support this reuse.

[1] Cheung NT, Fung KW, Wong KC, Cheung A, Cheung J,

Ho W, Cheung C, Shung E, Fung V, Fung H. Medical
informatics--the state of the art in the Hospital Authority.
Int J Med Inf. 2001 Jul;62(2-3):113-9
[2] Berner ES, Detmer DE, Simborg D. Will the wave finally
break? A brief view of the adoption of electronic medical
records in the United States. J Am Med Inform Assoc.
2005 Jan-Feb;12(1):3-7
[3] Dick RS, Steen EB (1991). The Computer-based Patient
Record: An Essential Technology for Health Care,
Washington, D.C., National Academy Press
[4] Committee on Quality Health Care in America. To err is
human: Building a safer health system. Washington, DC:
Institute of Medicine; 1999
[5] Committee on Quality Health Care in America. Using
information technology. Crossing the quality chasm:

Prioritize ruthlessly (first things first)

In the early days of clinical systems in the HA, overcoming clinician resistance and finding areas of benefit were of
prime importance. However as a critical mass of informatics aware clinicians developed, there was fairly

309

N. Cheung et al. / Principles-Based Medical Informatics for Success

[6]

[7]

[8]

[9]

[10]
[11]

[12]
[13]

[14]

[15]

A new health system for the 21st century. Washington, DC:

Institute of Medicine; 2001
Detmer D, Steen E. Learning from Abroad: Lessons and
Questions on Personal Health Records for National Policy.
Available at American Association of Retired Persons at:
http://www.aarp.org/research/health/healthliteracy/
2006_10_phr_abroad.html. Accessed 19 Apr 2006.
Garrido T, Raymond B, Jamieson L, Liang L, Wiesenthal A.
Making the business case for hospital information systems-a Kaiser Permanente investment decision. J Health Care
Finance. 2004 Winter;31(2):16-25.
Walker J, Pan E, Johnston D, Adler-Milstein J, Bates DW,
Middleton B. The Value Of Health Care Information
Exchange And Interoperability. Health Aff (Millwood).
2005 Jan-Jun;Suppl Web Exclusives:W5-10-W5-18
14th Annual HIMSS Leadership Survey, retrieved Nov
10 2006 from: http:/www.himss.org/2003survey/ASP/
healthcarecio_final.asp
2006 Leadership Survey, retrieved Nov 10, 2006 from:
http://www.himss.org/2006survey/healthcareCIO_final.asp
NHS Connecting for Health. History of Our Organisation.
Available at http://www.connectingforhealth.nhs.uk/about/
history. Accessed 10 Nov 2006.
Catz M, Bayne J. Canada Health Infoway A pan-Canadian
Approach. AMIA Annu Symp Proc. 2003; 2003: 807.
Scott JT, Rundall TG, Vogt TM, Hsu J. Kaiser Permanente's
experience of implementing an electronic medical record: a
qualitative study. BMJ. 2005 Dec 3; 331(7528): 1313-1316.
Ash JS, Fournier L, Stavri PZ, Dykstra R. Principles for
a successful computerized physician order entry
implementation. AMIA Annu Symp Proc. 2003;:36-40.
Hier DB, Rothschild A, LeMaistre A, Keeler J. Differing
faculty and housestaff acceptance of an electronic health
record. Int J Med Inform. 2005 Aug;74(7-8):657-62

[16] Lapointe L, Rivard S. Getting physicians to accept new

information technology: insights from case studies. CMAJ,
2006 May 23; 174(11): 15731578
[17] Campbell EM, Sittig DF, Ash JS, Guappone KP, Dykstra
RH. Types of unintended consequences related to
computerized provider order entry. J Am Med Inform
Assoc. 2006 Sep-Oct;13(5):547-56
[18] O'Connell RT, Cho C, Shah N, Brown K, Shiffman RN.
Take note(s): differential EHR satisfaction with two
implementations under one roof. J Am Med Inform Assoc.
2004 Jan-Feb;11(1):43-9
[19] Cheung NT, Fung V, Kong JH. The Hong Kong Hospital
Authority's information architecture. Medinfo. 2004;11
(Pt 2):1183-6.
[20] Sittig DF, Krall M, Kaalaas-Sittig J, Ash JS. Emotional
aspects of computer-based provider order entry:
a qualitative study. J Am Med Inform Assoc. 2005
Sep-Oct;12(5):561-7.
[21] Hendy J, Reeves BC, Fulop N, Hutchings A, Masseria C.
Challenges to implementing the national programme for
information technology (NPfIT): a qualitative study. BMJ.
2005 Aug 6;331(7512):331-6.
[22] Ornstein C. Hospital Heeds Doctors, Suspends Use of
Software. LA Times, Jan 22, 2003.

Address for correspondence

Dr N.T. Cheung
Rm 121N, Hospital Authority Building
147B Argyle St
Kowloon
Hong Kong
email: cheungnt@ha.org.hk

310