 
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
21-February-2003, Vol 116 No 1169
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Integrated care information technology
Ian Rowe and Phil Brimacombe
While information systems are widely used in the health
sector, they are seldom focused on the process of delivering care. Rich
information is available about the costs of care and historical allocation of
resources, but even in the most advanced hospitals some information is only
collected and recorded in written and narrative form.
The traditional visit-based funding process deters non-visit
information-based interventions, and discourages the sharing of information. The
long history of competing for scarce funding has left a legacy of
misunderstanding, distrust and a culture of not sharing information.
This, in turn, contributes to gaps within the system and our
patients fall through these gaps on a regular basis. Counties Manukau has a
population with the highest deprivation and lowest health status in the country,
and with a high genetic and environmental propensity for chronic illness and
high health resource consumption. We are under-funded and under-resourced and
there is no sign of this situation improving.
The challenge for IT in this context is how to implement
systems that facilitate dramatic change in the closing of these gaps, through
the sharing of information, within the critical constraints of privacy, funding,
and clinical safety.
This paper records how Counties Manukau District Health
Board (CMDHB) is using such an approach to deliver tangible benefits for its
population.
BackgroundFor those tackling this problem, Don
Berwick’s “Escape fire” video is compulsory
viewing.1 It is a compelling description of the
problems within the health system and provides key direction for the style and
nature of the approach that is required to fix them. In Berwick’s words,
“we must stop doing harm”.
Berwick draws an analogy between the health system and a
group of smoke jumpers fighting a relatively small fire, which circled round and
trapped them in a gully. The foreman recognised the potential for disaster and
ordered the crew to evacuate, but the only route open to them was up the
opposite hill face. Forced to run up a 70 degree slope, the fire was rapidly
overtaking them.
The foreman dropped his pack and tools, stopped and lit a
fire in front of him, creating a burnt area into which he then stepped into and
survived, thus inventing the ‘escape fire’ – a technique which
is now taught to all fire fighters.
Fourteen of his colleagues ignored his pleas to join him and
perished. Their failure was not through any lack of application or effort but
what Berwick refers to as the collapse of sense-making.
Berwick argues that the fire in the health system is now
raging out of control, more and more dollars are being consumed, without
improving the overall quality of care, medical errors are occurring with
increasing and alarming frequency, and that the doctors and nurses within the
system are running faster and faster under increasing pressure. The current
system, in which staff run faster and yet experience increasing pressure, no
longer makes sense to the people within it.
Future stateThe future health system will be
patient focused and will integrate providers to deliver consistent, continuous,
evidenced-based care. CMDHB is committed to delivering “the Right Care in
the Right Place at the Right Time” (Figure 1).
Figure 1. In this model, the patient is surrounded by a
seamless continuum of care between primary, secondary and community care. The
activities of providers are integrated and their boundaries are not of concern
to the patient.
 The patient interacts with this seamless continuum of care.
These interactions reflect the needs of the patient (including information
needs), not just the interests of the funders and providers.
If the goal of the organisation is to do the right thing and
deliver interactions toreflect the needs of the patient (including right care at
the right place and the right time), then the role of technology is to make it
easier for the Right Thing to occur. Dr Thomas Payne (MD of Clinical Informatics
at the Veterans Association Puget Sound Health Care System) has extended this
concept, and we have adopted his philosophy of “making the right thing the
easiest thing to do” as the mantra for our projects.
The above diagram shows the boundaries of the ‘sphere
of responsibility’ for the District Health Boards. However the health
system is not constrained by these boundaries, there are additional interactions
as depicted in Figure 2.
Figure 2. The health system facilitates interactions
with external agencies such as WINZ, Housing New Zealand, Pharmacies, private
laboratories and radiology services.
 In our theoretical future system, the patient has a good
understanding of the access points to care, and is fully informed about their
health status and the decisions that are being made for the delivery of
care.
In our escape fire, the provider has access to all the
information required for the delivery of effective care, without incurring a
burden of knowledge of how and where to access this information. All the
required information, and only the required information, is delivered where and
when required, without limiting clinical freedom, impinging on patient privacy,
or threatening financial viability for the providers.
While the sectors within health operate in separate
information silos, from a CMDHB perspective, most of the silos function well.
What is missing is the plumbing between the silos to enable transfer of
information, and the taps and valves to ensure that the flow is appropriately
controlled.
Some of our GPs have old silos that will not withstand the
maintenance required to connect the plumbing. Our approach is to encourage them
to use more modern silos or to join with their colleagues in order to afford
better silos.
The real power of the Internet is the ability to quickly
access information from multiple locations, without the need for a single
monolithic repository. The health sector has one huge systemic advantage, which
it generally fails to exploit. The customer (patient) can only be in one place
at a time, and with today’s technology there is sufficient time between
their arrival at the registration desk and their arrival in the consulting room,
for a physician to complete an Internet search and retrieve any relevant
information from other providers.
Current stateThe current health system is broken.
Using the metaphor of the wheel, the gaps occur between the traditional silos of
the health system structure, primary, secondary and community, creating breaks
in the rim of the continuum of care (Figure 3).
 Figure 3. Health system current state
The gaps in the health system cannot be addressed purely at
a technology and information level; there are fundamental shortages of clinical
resources associated with each of these gaps.
The formation of district health boards (DHBs) with a wide
sphere of responsibility, has improved understanding and measurement of the
gaps, providing an opportunity to allocate resources to them. CMDHB has led the
way in committing the funding and resources to closing the gaps, and there is
insufficient room in this paper to give full credit to the enormous amount of
talent and effort that is being applied in their various projects.
The difficulty, and IT opportunity, is the provision of
mechanisms to target resources to the gaps without creating additional
organisational units, and without risk of the additional resources being
consumed within existing fires.
To focus attention on the gaps and the need for information
sharing as a solution, we chose a case study that was representative of the
problem. The case study was James Whakaruru. James’ untimely death is a
reflection on the communications failings of the system rather than the failure
of the hard-working individuals within it.
In his short four years of life, James was seen 40 times by
health practitioners, made four presentations at the hospital emergency
department, was admitted twice and attended one outpatient clinic, had three
face-to-face Plunket contacts, and made 30 visits to general practitioners at
four practices.2
After investigating this case, the Children’s
Commissioner Roger McClay stated that, “Collectively the health sector had
available a telling picture of James’ circumstances. This picture was
never put together because of poor communication between
practitioners.”
While we can no longer save James, and we do not claim that
any system would have, we will know that we have made real progress when we have
in place a solution that could have worked for James.
James’ story is a classic tale of child abuse
characterised by multiple visits to a large number of different practitioners.
If each of the practitioners simply had access to a high-level summary of what
had gone before, their approach to James’ care may have been
different.
It also provides an insight into the type of information
sharing that is required. In order to have changed their behaviour, each
provider did not need access to all of James’ information. All they needed
to know was that James had been seen and a brief description of the purpose of
the visit. It was not necessary to have invaded James’ privacy in order to
save his life.
The journey to the futureIn Berwick’s smoke jumper
story, the foreman was the first to understand the problem. Traditional methods
(running like hell) were not going to work – something new was required.
However, even having created the escape fire, the foreman was unable to
communicate the problem or convince his colleagues of the merits of his
solution. In health, understanding and articulating the problem is even less
easy – there are so many fires in so many places.
The first and most obvious question was where to start?
While recognising that the starting point is less important than the act of
starting, there was in fact a logical place to start, and perversely that was
not at the beginning.
CMDHB found that the discharge summary was the best example
of the kind of information exchange that should pervade the health system. The
intentions of the discharge summary are to convey clinical information about an
episode of care from one provider to another as well as to the patient, and to
convey recommendations for continuation of care.
The Electronic Discharge Summary Project was one of the
first sparks of our escape fire. Automating and improving this existing process
demonstrated:
In a similar vein, status messages
(informing the GP of the status of their patient secondary referral, booking,
etc) can be made available through a relatively simple process that derives
significant benefits at low cost. SAH now sends over 6000 electronic discharge
summaries and over 12 000 electronic status messages per month to over 300 GPs,
who receive the messages integrated into their Practice Management Systems
(PMS).
In the context of the broken wheel, these projects did not
increase resource capacity but were more a signal of intention, connecting a
couple of spokes with string and providing a context around which further
discussion about information sharing could be based (Figure 4).
Figure 4. Closing the gap
 Having created the spark, it was now time to set fire to the
grass and invent our own escape fire. In our case this took the form of the
Integrated Care Server (ICS). This capability was also used as the line that was
drawn in the sand to maintain the separation between the primary and secondary
elements of the gap.
The ICS collects data about patients, uses a set of rules to
interrogate the data, and then automatically sends alerts, reminders and other
messages to the caregiver. It was used for two pilots, the Well Child and
Diabetes Disease Management Projects.
The ICS also addresses one of the other key issues in health
– while the patient can only be in one place at a time, you can never tell
where they will turn up next. We have already considered and discarded the
concept of a single health database for the country. The problem is therefore
how to get access to relevant (right) information at the right place and the
right time.
Very few primary provider systems are available on a
twenty-four-hour-a-day basis, so the ICS provides a repository of key
information that can be delivered on request.
The technology (Figure 5)Patients are enrolled on the
Diabetes programme within the primary setting, but Well Child enrolment occurs
at birth within the secondary setting. Clinical information is captured and
recorded in CMDHB’s specified electronic templates within the GP’s
existing PMS. Where information already exists within the PMS from previous
visits or from electronic laboratory results, this is automatically included in
the template without need for re-keying.
Messaging occurs via Healthlink, using the New Zealand RSD
version of HL/7, in which message types previously existed. New message types
are being accommodated via XML, and we have now converted the entire RSD
specification to XML for future use. We are working towards making all messaging
compliant with the recently released Code of Practice for the secure electronic
transmission of health information.3
On receipt of these messages, the ICS has the ability to run
business rules over clinical values. The results of business rules are fed back
to the GP through messages and are formatted as suggested advice. The GPs are
free to act on system-provided advice as they see fit and the system produces
reports to the GPs on their outcomes.
The system works with the GP offline or online. If the GP is
offline (dial-up connection to Healthlink), message exchange and receipt of
advice takes around three minutes; if the GP is online (ADSL connection through
the Health Intranet), then messaging is virtually instant.
The ICS uses Orion Systems’ Concerto Disease
Management application, part of their Soprano suite running within their
Concerto framework. The server is a Hewlett Packard LPR Server with a Microsoft
platform including NT4, IIS, and SQL Server 7.
The ICS can be accessed via Internet connections using IPSec
encrypted pathways using a standard browser interface. Providers’ access
to information is controlled by their profile and governed by an access
agreement. Providers will only access information about a patient when they are
providing care to that patient.
Well Child The
Kidslink system notes when a child is born, records the linkages between the
child and the nominated GP and Well Child provider, and ensures that the GP and
Well Child provider are also informed. When the provider completes the
immunization or check, the ICS is updated. For those (12) GPs with current
versions of a PMS with messaging capability, this happens automatically.
Otherwise, the GP or provider delivers the information manually (eg, by fax),
and it is keyed into the system centrally.
The system calculates when the various immunizations and
Well Child checks are due, and automatically sends reminders to the GP or
provider if the relevant event has not been recorded on schedule. The system
assumes that the provider’s system has also calculated a schedule that at
least meets this minimum requirement, and intends ‘chasing’ only the
exceptions.
Diabetes While the
Well Child pilot was the first flame of our escape fire, the Diabetes project
took this technology and created a fire. Under this project, we extended the
capabilities of the ICS to collect a wider dataset of quarterly clinical
observations and reason on them, returning to the GP evidence-based
recommendations for care. This was coupled with additional funding and
resources, so that the provider was able to deliver the additional care. In the
context of our mantra, we are reminding the GP of the right thing and providing
the system, funding, and resources in order to make it the easiest thing to
do.
The dataset was gathered using templates within the
GP’s PMS, however modifications were undertaken so that many elements of
information were prepopulated from elsewhere in the PMS to minimise duplication.
The remaining information was entered into the template by the GP during the
quarterly or annual check (Figure 6).
All the information was then sent as a message to the ICS,
which was then able to apply a large number of clinical rules. If certain
indicators such as lipids exceeded preset thresholds, the system returned a
suggestion within minutes, providing clinical decision support using the
knowledge and rules built into the system by agreement between the GP and
secondary specialists.
If a patient presents to the Counties Manukau Emergency
Department, the information from their previous four checks is available to
clinicians.
Figure 6. The MedTech 32 interface for
diabetes
 ResultsAt the end of the Well Child pilot,
which comprised a sub-sample of 296 consented children for whom caregiver and
provider data were available, immunization eligibility was determined. Evidence
was obtained (by 23 July 2001) that 84% of children eligible for three
immunisations had received them, and that 91% of those eligible for at least
two, and 94% eligible for at least one had also received them.
No attempt was made to distinguish between checks done
before the follow-up process was initiated or after. Therefore, the project
cannot claim to have influenced the immunization rate to any specific degree.
What we can say is that with information sharing and motivated providers, it is
possible to record an accurate picture and demonstrate a high level of
compliance (in excess of 90%). This project has now progressed to implementation
in three districts within CMDHB, and one district within Waitemata
DHB.
While the Diabetes project only began six months ago, the
early results are promising and the first evaluation has recently been
completed.
The ICS trial now encompasses three practices and 250
patients. There is a consistent story of reduction in HbA1C, positive feedback
from patients and providers alike, and ever-growing interest in picking up this
approach.
HbA1c >9 showed some significant changes in the cohort
involved in the Diabetes project. The percentage of patients with HbA1c >9
pre-enrolment was 43%, and reduced to 16% post-enrolment at the first review
visit. Also, the average level moved from a value of 8.7 to 7.8. More
significantly, not only did the mean drop, but the distribution curve changed so
that high-end outliers came closer to the mean in the post-enrolment
population.
The independent project evaluation concluded, “The
results that have been obtained to date illustrate real and significant
reductions in HBA1c levels. If these are sustained, CMDHB could expect,
according to international evidence, an average 30% reduction in total health
care costs for these
patients.”4
Early analysis of secondary data for patients on the
programme indicates a reduction in secondary admission, however further time and
data are required before this can be quantified.
A better outcomeThe Well Child and Diabetes
projects have demonstrated the power of the systemic application of our mantra.
Providers represent the best-motivated and most compassionate people. They
always want to do the right thing, but it often is not the easiest thing to
do.
In many complex diseases such as diabetes, the right care is
difficult to understand and even more difficult to apply. Providers are expected
to remember ranges and values for a large number of observations and to be able
to assess the patient on a large number of combinations of these observations.
This is analogous to learning the road code – we have all learnt the
information and passed the test, but do we always remember to adhere to all of
the rules in our daily lives? The pace of advances in medicine means that the
rules are constantly changing, yet the increasing burden associated with these
advances means we expect doctors to drive faster and faster.
Our implementation of the ICS has demonstrated that there is
benefit to having a system that monitors the road and conditions, calculates the
appropriate speed and distances and relays its advice to the GP. The GP retains
full clinical freedom and control, but the evidence is that the additional
information improves compliance and reduces adverse events.
IT can provide useful tools to assist DHBs with targeting
resources to address gaps in the system. Information sharing is not the same as
information collection, and does not require all of the information all of the
time. It requires all of and only the relevant information required to deliver
the right care at the right place and the right time.
We have used technology to set a fire in front of us and we
have stepped into it. There is room in our escape fire for others and potential
for others to light similar fires.
Author information:
Ian Rowe, IS Enterprise Architect; Phil Brimacombe, Chief Information
Officer, Counties Manukau District Health Board, Auckland
Correspondence: Ian
Rowe, Private Bag 94052, South Auckland Mail Centre. Fax: (09) 262 9501; email:
ian.rowe@cmdhb.org.nz
References:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|