When St Bartholomew's Hospital and the London NHS Trust decided to
integrate patient hospital records, they opted for hanging
applications off a common framework. Liz Warren reports
For many years, the National Health Service has struggled to
implement and use information systems successfully. So it is a
brave move by St Bartholomew's Hospital, London, and The London NHS
Trust to have invested £1.6m in an organisation-wide electronic
patient record system. This will integrate all the patient
information which the trust holds electronically into a single
virtual repository and make it available anywhere through a
browser-based interface.
Despite the legions of NHS IT horror stories, the trust knew it had
to make this investment because it faced a number of headaches
which made change a priority. Firstly, it was facing increasing
problems communicating information - especially patient-related
data - both within the organisation itself and with the wider
healthcare community, especially GPs. A particular difficulty was
to make patient case notes available whenever and wherever they
were needed.
"The speed of throughput has increased over the past five years and
paper records couldn't keep up with that," points out Michael
Glynn, clinical director of the medical and emergency directorate
of the trust and a consultant in general medicine and
gastro-enterology.
As a result, patient treatment was being compromised and resources
were often used inefficiently. For example, the trust had a high
rate of cancelled operations, primarily because patient notes were
either unavailable or missing key information such as the latest
pathology results.
On top of that, the NHS is moving towards an approach to care which
demands that hospitals harness clinical information more
effectively, both to improve the care of individual patients and to
allow doctors to audit and improve their overall approach to
treatment. Outside pressure has also come from the Government's NHS
modernisation agenda, which includes performance improvement
targets, and the NHS information strategy, Information for Health,
which mandates that all trusts should have a basic form of
electronic patient record in place by the end of 2005.
Barts and The London NHS Trust had to respond to these issues, but
any new information solution also had to incorporate the trust's
legacy systems, such as its main hospital information system, which
runs on the VMS operating system and has roots stretching back more
than 25 years.
"The functionality of our existing systems is as good as anything
on the market and we didn't want to throw out that investment and
success," explains Barry Elliott, the trust's director of finance,
"but we needed to find a way to bring together those islands of
information and eliminate the need for staff to access multiple
systems to get comprehensive patient information."
At the same time, the trust wanted to replace elements of its
functionality gradually over the next five years, with new accident
and emergency and laboratory systems to be rolled out by the end of
2001 and new systems to support diagnostic imaging to follow
shortly. So the electronic patient record needed to be able to
incorporate these new applications without major disruptions to
other parts of the system.
Paul Thomas, the trust's technical services manager, says the trust
identified four ways it could respond to these challenges:
- To do nothing and muddle on, with data
being entered into multiple systems and staff needing to visit
different terminals to get all the information they required
- To rip out its existing systems and replace them in a big-bang
project, an approach which, on past NHS experience, is doomed to
failure
- To evolve by buying new systems and integrating them with
existing applications. This is the trust's current approach, but
one that is technically complex and difficult to manage and which
does not provide the step change in efficiency the trust is looking
for
- To create an architectural solution which allows different
applications to be hung from a common framework.
It was this latter approach that Barts and The London NHS Trust
embraced, using an architecture based on Microsoft's Distributed
Internet Architecture (known as DNA or .net).
"The key aspect is the layering: the separation of data, business
logic and presentation," Thomas explains. "That allows our
25-year-old hospital information system to become a data source and
functional engine for a browser-based electronic patient record.
For instance, we could use the outpatient functionality in the
hospital system to provide an outpatient booking facility over the
Web to GPs' surgeries."
The architectural approach also allows the trust to create a
virtual repository: drawing together data on the same patient from
different systems so that staff can access the complete patient
record from just one screen. Similarly, the trust is able to use
the architecture to enforce a common approach to security and to
develop single sign-on for users, irrespective of which system
holds the data or functionality.
This virtual approach - which allows Barts and The London NHS Trust
to incorporate any existing system as well as any new ones it plans
to buy in the future - is also helping the trust tackle another
headache: negotiating the tricky political waters of getting the
users of the 68 existing systems to sign up to a trust-wide
solution.
Many hospital system projects have foundered at this point, but
Barts and The London NHS Trust recognises that users are not being
awkward when rejecting centrally imposed solutions. Most are senior
clinicians who have used part of their own departmental budgets to
acquire products which precisely meet the data collection and
management needs of their specialist areas - needs which are rarely
met by monolithic hospital-wide applications. The beauty of the
Web-based approach is that it allows staff to continue to input
information into their local systems but permits any subset of that
data to be made available to the rest of the trust.
"To make it a real success, we need to ensure we continue to have
clinician buy-in and that what we are doing is fundamentally
meeting the requirements of clinicians," says Elliott. "If we do
that, we should by definition be meeting the requirements of
business managers and administrators."
The trust has recruited a number of "champions" who are consultants
at the hospital - including the lead clinician in the accident and
emergency department and the head of the medical and emergency
directorate - to evangelise the benefits of the virtual electronic
patient record amongst their colleagues.
However, getting the funding to create this grand vision was not a
simple task. "Just because we're large doesn't mean we have more
flexibility compared to other hospitals when it comes to investing
in IT," points out Elliott. "There is always pressure to spend
every last penny on patient care, with support services playing
second fiddle. But at Barts and The London NHS Trust, we recognise
that delivering high quality patient care does require good
information systems - and the infrastructure to underpin those
systems - and that we need to invest in that."
The problem has been compounded by the fact that, until very
recently, Barts and The London NHS Trust was running a budget
deficit. So any solutions the trust does implement must cost little
up front, irrespective of cash savings they may bring in the long
term. That is why the Web-based, evolutionary approach was so
attractive.
Of course, the trust has had to put some money up front and, with
the system still in its early stages, has seen little return so
far. The most important benefits expected in the future, but those
which are hardest to quantify, are improvements to patient care and
patients' experiences in the hospital, thanks to better access to
information.
"Even in these early stages, when we only have a snapshot of the
data and restrictions on where it can be viewed, we are still in a
better position, because most of the information in the record is
now available on screen if the patient notes aren't available,"
Glynn says.
The electronic patient record will also have a direct impact on the
bottom line by, for example, reducing wasted theatre time by
ensuring operations are not cancelled purely because a patient's
notes are missing or incomplete.
Elliott says that the strategic long-term objective is to get rid
of paper records altogether, although he acknowledges there are
issues relating to cost, staff attitudes to such a move and the
need to maintain the speed of bedside interactions which still need
to be solved. To this end, the IT team is now considering the
potential of wireless networks which will allow access through
handheld devices such as personal digital assistants. The cost of
such solutions will eventually be offset by savings in the revenue
budget because such systems should be cheaper to operate than
paper-based medical records and a film-based X-ray service.
Glynn adds that the system will also support moves in the medical
community towards evidence-based medicine, where clinicians use all
the available research to determine which treatment will deliver
the best outcome for a patient. The Web-based infrastructure will
be able to deliver the vast quantities of information involved in
this approach. Similarly, the system has the potential to deliver
various forms of telemedicine, such as the ability for a doctor at
the hospital to consult with a patient and GP while the patient is
in the GP's surgery.
Barts and The London NHS Trust
St Bartholomew's
Hospital and the London NHS Trust is one of the UK's largest and
most complex teaching and research hospitals. It has an annual
budget of £330m and more than 6,200 staff spread over three main
sites.
It is both the local district general hospital for Tower Hamlets
and home to a number of specialties: as well as being a major
trauma centre and the base for the Helicopter Emergency Medicine
Service for London and East Anglia, the trust also provides
advanced patient care in the areas of neurosciences and cardiac,
cancer, renal and oral-maxillofacial services.
What bonuses has the project delivered?
- Clinicians who have specialist needs can retain their own
preferred IT systems but that data will be available to others in
the hospital who need it
- Applications can be replaced incrementally with best-of-breed
solutions without disrupting the electronic patient records as a
whole
- Patient information can be made available anywhere - and all
the time - in the trust's three main sites, with a wireless network
being considered to allow deployment at the bedside using handheld
devices
- Not only provides electronic patient records for the trust but
allows the trust's systems and data to be shared in a controlled
way with other health and social care organisations such as
GPs
- Single sign-on to all the different systems a user is
authorised to access.
Clinical information systems
- More than 60, including: A&E, pharmacy, pathology, X-rays,
angio films, ultrasound, ECG, lung functions, cardiac, diabetes,
nursing, clinical documents, clinical notes, scanned documents
- Functions
- Data
- Common NT-based security infrastructure providing single
sign-on
- Delivered over hospital network, NHSNet and secure dial in (eg,
by GPs).