|
Inside the November/December 2003 print edition of Canadian
Healthcare Technology:
Feature Report: Clinical decision support systems
 Organizations work
to improve coding of clinical data
Summing up the diagnostic discrepancies found
in a data re-abstraction study done by the Ontario Ministry of
Health on 10 randomly selected hospitals (four teaching, six
community), one statistical slide in the presentation indicated
there was a “% Error” among the coded clinical data
of 65 percent!
Canadian software helps kids with prostheses
The amputee team at Bloorview MacMillan Children’s
Centre, in Toronto, has developed a software solution that tracks
how kids use their prosthetic arms at different stages of their
lives.
Consolidation and adoption of business outlook spurs e-commerce
As more Canadian hospitals make the transition
to a centralized purchasing model, there’s a growing trend
to adopt electronic commerce strategies at the same time, says
Nils Clausen, Country Manager, Canada, for Global Healthcare
Exchange Inc. (GHX).
Decision support systems can help hospitals reduce costs
and improve patient care
The term ‘clinical decision support system’
sounds mysterious, but these computerized solutions are yielding
clear-cut results for many healthcare centres.
MOXXI on the move
Quebec’s innovative MOXXI project gives
family physicians and other doctors access to electronic patient
records and enables them to send prescriptions electronically
to pharmacists. It’s now making use of wireless, mobile
technology.
Computerized CDM
Computers and electronic tracking are ideal
tools for Chronic Disease Management (CDM). But in order for
physicians and patients to use them, the solutions must be easy
to use and tailored to their workflow or lifestyles.
READ THE STORY ONLINE
The Atlas project
Dartmouth College, in New Hampshire, famed
for its computer science school, is home to a data mining project
called Atlas that’s assessing the quality of medical care
in the United States. Using the data, analysts are even able
to detect fraud.
Show time
November is a busy month for healthcare trade
shows and conferences. Two prominent examples are the OHA Convention
in Toronto, and the Medica 2003 show in Duesseldorf, Germany.
PLUS news stories, analysis, and features and more.
|
Organizations work to improve coding of clinical data
By Andy Shaw
One of the presentations at a recent Ontario
Hospital Association conference in Toronto could have been extremely
alarming to an outsider not familiar with the complexities of
coding clinical data. Summing up the diagnostic discrepancies
found in a data re-abstraction study done by the Ontario Ministry
of Health on 10 randomly selected hospitals (four teaching, six
community), one statistical slide in the presentation indicated
there was a “% Error” among the coded clinical data
of 65 percent!
If this sampling were true for Canadian hospitals
in general, one might hastily and erroneously conclude that a
major foundation of our healthcare system – the gleaning
of reliable data that guides how we practice, fund, and manage
care – was dangerously cracked and flawed.
For those in the know, however, what was revealed
by the Pilot Clinical Data Quality Audit, as the study was called,
was not nearly so alarming. First, they point out, the choice
of the word “Error” in the study summary was an unfortunate
one. In the medical context the word error has a fatal ring to
it, as in “medical error”. These were not medical errors
that were being studied by Ontario but rather differences in
the way hospital coders interpret the written records of what
caregivers and clinicians did to whom when.
“What the study was really dealing with
are what some call discrepancies but what we prefer to call variations
in the way events are coded,” says Caroline Heick, head
of data quality and classification for the Canadian Institute
of Health Information (CIHI) in Ottawa. “And we know from
our own CIHI studies that the discrepancy or variation rate in
coding nationally is much, much lower.”
To be fair to the Ontario researchers and
presenters, it should be made clear that the Ontario study was
never meant to reveal any sort of statistical average for the
province. Rather, it targeted coding connected with the charting
of a limited number of complex clinical practices – instances
where it was suspected that the discrepancy rate might be highest.
The purpose, in part, was to see where remedial action was needed
most.
What can be safely concluded from the Ontario
study findings is that coding is less a science and far more
of a craft.
Yet the study still does raise some serious
questions about discrepancies. Among them: Why do they occur?
What kind of discrepancies are most prevalent? And what must
be done to avoid them in future?
For Heather Berry, the reason coding variations
occur are largely human. Berry is health records manager for
the South Bruce Grey Health Centre, a four-campus hospital centred
in Walkerton, Ont.. She attended the Toronto conference and was
an active participant in the audience give-and-take following
the study’s presentation.
“The information the coders are working
with is extremely complex, and yet, for it to be of high enough
quality to be useful it has to meet some very demanding criteria,”
says Berry, who then ticks off those criteria. “It has to
be specific, complete, accurate, timely, relevant, current, comprehensive,
and appropriate all at once.”
And that takes training. The preferred coders
being hired today, Berry points out, are graduates of a two-year
professional health records management program. They codify information
according to the World Health Organization (WHO) classifications
of disease and morbid conditions. But the complexities come when
patients bring with them, as so many do, more than one disease
or condition. Getting the coding right for that individual is
not only a personal challenge to the coder, but also crucial
to the tasks of hospital bean counters, who must re-coup the
costs from government of diagnosing and treating that patient.
“You could have a patient come in with
the simple condition of a fractured hip,” says Berry, “But
what if that patient is also a severe diabetic, a paraplegic,
and morbidly obese? The treatment becomes very much different
than for a simple hip fracture.”
Such complexities create what Berry calls
“grey areas” in coding practices, where the basic standards
are not quite specific enough to guide the coders in their interpretation
of what was written down as done to the patient.
Consequently, Berry applauds the efforts of
the CIHI, in particular, in conducting workshops for coders,
providing them with online interpretation assistance through
the CIHI website, and other initiatives the organization has
taken to help standardize coding practices nationally.
There’s another source of coding inaccuracies,
however, says Berry, that can only be addressed down at the hospital
level.
“You’ve got to work with the doctors
to ensure that what they are entering on the records is specific,
complete, accurate, and meets all the other criteria I mentioned
earlier in order to produce quality data,” says Berry. “They
have to be educated to the importance of all this, and that’s
challenging. What we have found that works here is to explain
it to doctors in terms of money. We say to them that just as
they must have their information complete on what they do with
patients in order to be paid, so does the hospital, as well.”
A long-term solution, suggests Berry, is to
make an introduction to coding practices part of a physician’s
education while at medical school – showing them how to
document their work accurately and thus conveniently for coders.
“It’s something that will help them run their own offices
once they graduate,” adds Berry.
Doctors alone of course are not the only challenge
to discrepancy-free coding. Other caregivers must also be brought
up to speed on the techniques for accurate record keeping.
Berry says she would also like to see more
rigour on the part of hospital management in applying the standards
of coding, perhaps adopting an ISO 9000 approach to data quality
that other organizations, in other industries, voluntarily qualify
for and adhere to. Also, as privacy regulations take hold in
healthcare, there will be even more pressure on hospitals and
coders to keep more and better records, she says, and hopes that
government funding of coding work will be boosted appropriately.
Most of all, she adds, she’s looking
forward to what will come of a data quality workshop and gathering
next month in Toronto involving major players from the CIHI,
the Ontario Hospital Association, the Ontario Ministry of Health
and the Canadian Health Records Association among others.
“There’s a very good spirit of cooperation
among these groups and a strong, shared desire to improve the
quality of our data and coding,” says Berry. “It will
be a very good meeting, and we will learn a lot about where we
should go from here.”
New software tracks how children use their prosthesis over
time
By Neil Zeidenberg
TORONTO – The amputee team at the Bloorview
MacMillan Children’s Centre has developed a computer-based
software solution to track how kids use their prosthetic arms
at different stages in their lives.
Unlike previous studies that simply compared
body-powered equipment to myoelectric devices, the PUFI software
addresses the functional benefits of children’s prosthetic
devices over time. It’s important, because without a thorough
investigation, it’s hard to prove that one prosthesis is
better than another.
“PUFI helps to determine how well kids
use their prosthesis over time, as well as how valuable they
consider it to be for a certain activity,” explained Sheila
Hubbard, manager of the amputee team at Bloorview MacMillan.
For example, a two-year old may find it difficult
to use a prosthesis for tying shoelaces, but over the course
of several months or years, he or she may become quite adept
at the task.
PUFI (Prosthetic Upper Extremity Functional
Index) is a multiple choice computer-based questionnaire that
asks how children between the ages of two and 18 fare when performing
a lengthy list of two-handed tasks, such as holding a hockey
stick, tying their shoelaces and doing up a zipper.
While answering the questionnaire, an animated
puffin named PUFI offers encouragement, word clarification and
assurance to parents and children that “they’re doing
great.”
Questions asked range from very general to
more task-specific, such as how they actually use their prosthetic.
The data are automatically saved in the PUFI database, enabling
clinicians to analyze how kids are using their prosthetics and
whether more training is required.
“PUFI records all of the data and instantly
gives you a full graphical report. Therapists can then sit down
with parents and children, look at areas where they may be having
difficulty and offer both encouragement and training,” says
Hubbard.
Having such detailed analysis before them
may also encourage children to use their prosthesis to perform
tasks they’ve never tried before. For example, some kids
may have never attempted to tie their shoelaces because their
parents hadn’t encouraged them.
Moreover, having solid data to back up their
findings gives developers of prosthetic devices the information
they need to improve their products and to understand what’s
most important to kids at different stages.
With the data accumulated so far, researchers
have begun noticing certain trends such as:
• Some children start out with a myoelectric
device, but due to an inability to perform certain activities
in the playground, they opt for a secondary prosthesis.
• As kids grow, their needs change, so
they use their prosthetics for different purposes and switch
to and from functional, durable or purely cosmetic devices.
Recently, PUFI has garnered international
attention with participation from 12 hospital sites around the
globe. Five of the sites are in Canada, three in the United Kingdom,
and one each in Australia, Holland, Sweden and Slovenia.
Last year, The Shriners Hospital for Children
in the United States, a network of 22 facilities throughout North
America, began using PUFI to track juvenile prosthetic usage
in about 300 of its patients.
Researchers hope to add up to 100 children
to the database each year, and re-evaluate the patterns from
children who are already participating in the project. As well,
they hope to make PUFI available on the Web. Going online would
simplify the sending and receiving of data as well as provide
therapists with access to the most current data.
“It’s important to not just capture
the data for these kids once, but rather, to capture it over
time,” says Hubbard. “When you look at someone over
a period of 20 years, you’re likely to see an interesting
pattern. However, if you do the same study over a shorter period
the changes will be subtler.”
Consolidation and adoption of business outlook spurs e-commerce
By Dianne Daniel
As more Canadian hospitals make the transition
to a centralized purchasing model, there’s a growing trend
to adopt electronic commerce strategies at the same time, says
Nils Clausen, Country Manager, Canada, for Global Healthcare
Exchange Inc. (GHX). Already GHX’s web-based exchange has
over 190 member hospitals in Canada – 95 of which are live
and trading with supplier members – and, according to Clausen,
all signs are pointing to increased growth in the future.
“By the end of this year we should hit an annual run rate
of around $115 million in transactions for Canada,” he says.
“And we see continued growth on a monthly basis as we get
more hospitals trading with more suppliers.”
Clausen credits the current focus on consolidation as a key driving
factor behind the push towards e-commerce, since it’s easier
for hospitals to connect electronically with suppliers when product
item masters and contracts are already streamlined in one central
database. Equally important, though, is the opportunity for increased
cost savings related to improvements in the entire materials
management process.
Recent benchmarks from GHX indicate those hospitals participating
in the exchange – an industry neutral utility that connects
providers, suppliers, distributors and group purchasing organizations
via the Internet – can expect a 25 to 40 percent reduction
in errors related to the ordering process as well as a 40 to
70 percent improvement in order cycle times. There’s also
a connection to enhanced patient care, by getting the right product
to the right place at the right time. As well, any money saved
can be used in other areas to improve patient care, he adds.
Sarah Friesen, general manager, Shared Healthcare Supply Services
(SHSS) in Toronto, agrees e-commerce is catching on in Canadian
hospitals. SHSS, which serves as a central purchasing facility
for University Health Network, Mount Sinai and Sunnybrook &
Women’s College hospitals, is currently using GHX for product
purchases in its Catheterization Lab as well as in Medical Imaging.
“The system has worked well for us. We’ve achieved
what we hoped we would in terms of error reduction and better
control over our ordering,” says Friesen. Additional gains
are being made in the lab, where handheld personal digital assistants
(PDAs) are being used in conjunction with barcode technology
to speed the ordering process, as well as ensure accuracy, since
many items purchased for the lab are implants that need to be
tracked.
“For us, the other primary objective was to get the patient
care worker doing patient care, as opposed to spending their
time sitting on a computer doing non-healthcare-related work,”
notes Friesen. “We’re trying to find innovative ways
to reduce the amount of time our patient care workers spend on
material management – it’s our job, not theirs.”
Under the handheld pilot, when healthcare workers scan a product,
an order is automatically created and sent to the appropriate
supplier via GHX. Order information is linked to the hospital’s
information system, so when the product is shipped, it can be
matched against the original purchase order, receipt and invoice
to ensure no errors are made.
The ultimate benefit is better control over the entire supply
chain process, says Friesen. “From the patient-care side,
it’s allowing us to be a lot better prepared and to be able
to respond to issues a lot faster than with the manual processes
we had before.”
Another group benefiting from e-commerce is the Capital District
Health Authority in Halifax, which currently has nine of its
major vendors trading on GHX. According to vice-president of
administration Calvin Crocker, one of the main advantages of
moving to an e-commerce model is the ability to pick up on pricing
discrepancies, sooner rather than later. It also provides access
to up-to-date information on order status, which in turn helps
to improve communications between materials management staff
and internal hospital customers.
In addition to joining GHX, Capital District Health Authority
has also signed on with a national group-purchasing program called
Healthpro. Dealing with major vendors and national contracts
through Healthpro makes it easier to engage in e-commerce activities,
says Crocker, because those are the suppliers who are investing
money to get their catalogues and pricing information on-line.
Like SHSS, Crocker would also like to trial barcode technology
at some point. “What we ultimately want out of this whole
e-commerce initiative is to move to a better materials management
process in terms of stock and just-in-time quantities, and better
sourcing,” he says. “On the finance side we want to
get to a point where everything is electronic and we’re
not cutting cheques.”
Both Crocker and Friesen say it is getting easier to convince
executives to buy into the concept of investing in e-commerce
because there’s been a shift in focus towards running a
hospital like a business.
“In terms of bringing the region together and consolidating
our operations,” said Crocker, “we’ve made a commitment
to be as efficient as possible. We want to spend as much of our
budget on the patient-care side as we can, and the only way we
can become really efficient is to ensure we take advantage of
the technology.”
Decision support systems can help hospitals reduce costs
and improve patient care
By Andy Shaw
The term ‘clinical decision support system’
sounds mysterious, but these computerized solutions are yielding
clear-cut results for many healthcare centres – perhaps
nowhere more so than the University Health Network (UHN), St.
Michael’s Hospital in Toronto, and at the McGill University
Health Centre (MUHC) in Montreal. All three institutions are
producing both clinical and financial advantages for themselves
and their patients through decision-making that’s supported
by a computer system that aids and abets better decision-making.
One can think of a decision support system
as an “informational” application rather than simply
an operational or transactional system. It can help both clinicians
and healthcare managers make easier and better decisions by interpreting
the information that flows from the hospital’s day-to-day
transactions. And it does so with varying degrees of intelligence
– either by monitoring, comparing, analyzing, forecasting
or modelling the data.
At UHN, for instance, Katherine Henning, head
of performance measurement, and project manager Janine Kaye,
use a Misys Data Warehouse, an advanced clinical decision support
application that makes sense of reams of data from an assortment
of databases, including hospital records, pathology, pharmacy,
physician order entry and patient satisfaction.
“We decided about two years ago to make
a fundamental structural change at UHN and create a group of
analysts that would support both management and clinical decision
making,” says Henning. “Specifically, they help our
quality teams look at data and make decisions from new and different
vantage points.”
The analysts can mount those vantage points
from the heights of transactional data that UHN has been collecting
since 1987 on what was then a data system from Per-Se (since
bought out by Misys) and which has been flowing into its enterprise
data warehouse since 2000.
“We now have data coming in now from
as many as 15 transactional systems,” says Kaye. “And
we have about 40 people accessing that data through an Oracle
reporting tool.”
Henning says there will be increasing amounts
of decision supporting data available to those analysts as time
goes by. “We’re only about a third of the way through
the list of inputs we eventually hope to have streaming in. This
year, for example, we want to integrate the operating room data
set as well as nursing workload data. The nursing data, for instance,
will tell us things like how many hours nurses spend with patients
and what they actually do with patients during that time.”
And therein lies the biggest rub for implementers
of all such decisions support systems, says Beverly Delaney,
MUHC’s director of administrative informational systems.
“We use our decision support system (from
Eclipsys) for all of our medical records reporting, for financial
and performance analysis, for utilization management, for quality
indicators, and essentially for any function that needs to look
at integrated data,” says Delaney. “But the challenges
we have faced in implementing all that have not been technical.
We’ve rarely had a major technical issue. Most of our challenges
have been cultural.”
That’s a polite way of saying that, for
some time, those who input data or those who use it to interpret
the data often act like unwilling horses or worse. They can be
lead to the water, but they can’t always be made to drink.
“We’ve recently integrated with
four other campuses,” explains Delaney. “And most of
them have long histories of doing things their own way. In the
process, we’ve found that the ability and willingness of
people to standardize their data is the biggest challenge. And
they’ve got to be willing if a decision support tool is
to ever be truly useful to everybody.”
A corollary to that challenge is also the
willingness of managers to use the data once they get it. Too
often that a decision support tool’s output gets ignored
– albeit, sometimes for good reasons: suspicions about the
quality of the data.
“Auditing and validating your data is
a job you cannot afford to underestimate,” says Delaney.
“Data quality is a major issue with these systems.”
It’s an issue McGill has faced up to
particularly on the clinical side of its decision support strategy.
“We’ve been gathering clinical information for over
12 years now, and for those physicians who have been with us
longest, their acceptance of the data is excellent. The best
way we’ve found to build acceptance is to present clinicians
with new looks at their data – and then ask them to validate
the data for us. We say to them: Please tell us what is wrong
with this information and help us fix it. And that has proven
to be quite a successful strategy.”
Soon there will be more data to validate,
as Delaney and MUHC press ahead with creating new and more useful
vantage points for its decision makers.
“We’ve decided to roll out the latest
‘Gold’ version of the Eclipsys (Sunrise Decision Support
Manager) product,” says Delaney. “It offers a Web-based
piece of the application that will give our users faster access
to the data through our intranet and will also simplify the data
for them. We’re in the very early stages of this, but eventually
our quality advisors and other expert users of the system will
be able to create views of the information specifically for targeted
users. Take for example, our obstetrical mission team. They will
be able to access views of their particular obstetrical patient
population and that population’s use of things like diagnostic
services, or the number of their returns to Emergency after discharge,
for instance, and whatever else that interests that particular
mission team. And it will give them tailored information both
about the quality and the use of their services.”
Still, as Delaney is quick to admit, and as
Michael Stewart at St. Mike’s hospital in Toronto agrees,
the beauty of that data very much lies in the eye of the beholder.
“Pharmacy is a classic example of how people view information
differently,” says Stewart, who is the hospital’s decision
support director. “They may buy morphine from their vendor,
for instance, that costs them say 35 cents. But looked at from
a hospital-wide point of view, that morphine may in total cost
us 65 cents – taking into account the hidden costs of the
purchaser who bought it, the transport person and receiver on
the shipping dock who handled it.
“Costs, in other words, that don’t
appear on the invoice. So you have to have a mechanism built
into your decision support system, as we now, that allows you
to reconcile those two different views.”
More challenging yet, adds Stewart, is the
attitude of users that can so easily turn sour.
“There’s a particular problem in
healthcare when data is not as good as it can be. We have a lot
of scientists and scientific-thinking people in healthcare, so
when it comes to data, they don’t expect ‘nearly right’.
If they find one patient they think has wrong information, then
it tends to invalidate the rest of the data. In other environments,
people are happy with a 95 percent accuracy rate, but here in
healthcare, it has to be 99.9 percent.”
The challenge for the future, says Stewart,
is to establish “confidence limits” in the accuracy
of a decision support system’s data. Says Stewart: “Systems
can be made 100 percent accurate, but there is usually a very
high financial cost connected with that perfect accuracy.”
Stewart says a more productive approach, and
one which St. Mike’s is pursuing, is to offer decision makers
more views of the same information, so that one bad apple does
not spoil the whole data barrel.
One way to do that, advises Stewart, is to
provide more links to concurrent utilization review. “One
such tool we have, for example, is used by our nurses who can
come on shift and ask: Why is this patient still in hospital?
It allows them to see what are the lab test results, or other
bottlenecks that are holding patient care up. If you tie that
view into hospital management systems, then management can really
start to come to grips with workflow and costs. If MRI says they
need another machine, for instance, that creates a challenging
decision for management. They inevitably wonder: Is it just because
MRI wants to be state of the art? But if you can pull up data
for management that says: By not having that new MRI machine
we will have 50 patients a week who are waiting four more days
and that are each costing us $500 a day – then its easier
for them to make a more informed decision about spending several
million dollars on a new MRI.”
Providing more multi-faceted views of their
transactional data is where St. Mike’s is doing some real
decision-support groundbreaking.
“We are the only hospital I am aware
of, for instance, that does care costing for all of its outpatient
visits,” says Stewart. “So our fully-costed patient
visits include 30,000 in-patients, 60,000 emergency patients,
and now 500,000 ambulatory or out-patients whose visits are also
fully costed.”
It is with such cross-department numbers that
modern day decision support systems can truly shine, says a man
who should know, Kevin McCole, vice president of sales, decision
support for Eclipsys in Chicago. A veteran of decision support
systems stretching back to 1990, and work experience with PSI,
an Eclipsys predecessor, McCole says he has stayed in the field
because of a philosophy that to this day still rings true for
him and Eclipsys.
“In the 1980s, there was really no tool
that really could measure the hospital as a business,” says
McCole. “Hospitals were managed back then by departments.
But a hospital’s process, costs, and efficiencies really
run more along service lines that cross over departmental boundaries
as a patient progresses through its care. So if open heart surgery,
for example, went up by 25 percent, say, there was no real way
of knowing what that would mean to other departments.”
But decision support systems can make a serious
difference to not only the efficiency of the hospital, but to
the care of the patient.
“We’ve had the advantage in the
United States of being forced by our billing systems to pay much
more attention to cost per case analysis, and so we’ve been
quicker to adopt and embrace decision support systems that can
give us that information,” says McCole. “But it’s
clear now from numerous studies that such systems can produce
an extremely significant reduction in a patient’s length
of stay in hospital, as well as among other benefits their surgical
complications.”
McCole says politely that he thinks that Canada
is still lagging in its use of such tools, but was impressed
recently by a healthcare costing conference in Edmonton he attended.
“There were a lot of people there who
wanted to get a better handle on this whole topic and I think
it shows the Canadian market is becoming a lot more sophisticated.”
For Michael Stewart at St. Mike’s, there’s
also a gap to be closed in decision support systems between what
he terms his “Cadillac” system and the Volkswagen or
smaller systems of the community care centers his hospital’s
patients also encounter. And at McGill, the gaps that will always
need closing are between people.
“There is really only one way to make
decision support systems effective,” concludes MUHC’s
Delaney. “You’ve got to go from department to department
and from person to person until everyone feels accountable to
each other for the data in the system and the way the information
is used.”
Wireless handheld computers enable doctors to prescribe at
point of care
By Neil Zeidenberg
MOXXI has gone mobile. The innovative Montreal-based
pilot project that has given physicians electronic access to
the medication histories of patients, and enables them to send
scripts directly to pharmacies, started off by using standard
desktop PCs.
But physicians are often on the go, and the
current phase of the ‘Medical Office of the 21st Century’,
aka MOXXI, is enabling doctors to use wireless iPAQs to receive
and send drug-related information.
“The beauty of this system is that doctors
can change the prescription on the spot, while the patient is
still present,” explains Dr. Robyn Tamblyn, director of
the MOXXI project.
The iPAQ Pocket PCs are equipped with Sierra
Wireless AirCards – high-speed wide-area-network cards that
connect to the Bell Mobility IX system. Physicians can use the
handheld computers to access patient charts and to send drug
prescriptions electronically to pharmacies, where the scripts
are automatically integrated into the pharmacy software.
Dr. Tamblyn explained that in Quebec, pharmacists
can accept electronic prescriptions. “Physicians and pharmacists
here established standards for what would constitute an acceptable
electronic prescription,” she said. “The digital signature
here is legal,” a development she believes will be adopted
by other provinces and by the United States.
The MOXXI project originally started in 1996
using standard PCs; the addition of wireless iPAQs began in January
2003. The mobile devices are said to require less training than
desktop solutions, and support in the field is better since any
problems experienced by doctors can now be monitored via the
wireless technology.
“We can actually see where they’re
having troubles entering data, how often they’re using it
and provide strategic support for physicians having difficulty,”
said Dr. Tamblyn. “It provides a technical solution that’s
much easier to support and it’s easier to use for doctors
in the field.”
MOXXI currently connects 30 physicians, about
12,000 patients and 31 pharmacies in Montreal. The pharmacies
are all located near the physician practices.
“We tried to set it up so that no matter
what pharmacy the patient visited, the system would be able to
retrieve electronic prescriptions and send information about
prescriptions dispensed,” Dr. Tamblyn explained. Patients
can essentially pick and choose on a moment’s notice which
pharmacy they go to. And since it’s electronic, pharmacists
don’t receive illegible handwritten prescriptions, but encrypted
documents with all of the information they need.
Initially, physicians enter prescriptions
onto a wireless keypad. The data is sent to the MOXXI servers
by cellular network where the drug and dosage is reviewed and
compared to the patient’s current medication. MOXXI then
sends information back to the physician regarding the medication
ordered, including proper dosage and any risks associated with
that drug. “Pharmacists are delighted they don’t have
to fuss over what the prescription is,” says Dr. Tamblyn.
“The physicians find the drug profile incredibly valuable,
especially for patients taking many different medications.”
The drug profile shows physicians a list of what drugs their
patients are on and from whom it was prescribed.
To date, MOXXI’s accomplishments include
a 19 percent reduction in the rate of medication errors; development
of a technology that can be readily used by busy primary care
physicians, and a technological model that makes it possible
to support them; the capacity to integrate information from both
the private pharmacies and the public insurer so physicians can
look at all drugs being prescribed and dispensed to their patients.
Despite all the achievements, there is room
for improvement. Number one is the memory capacity of the iPAQ
in terms of its performance. It needs to be faster. Electronic
prescribing is certainly faster when you’re writing many
prescriptions, but for a single prescription, it’s slower
than writing by hand. That’s because they’re currently
using a 56k modem.
“In order to make this a more functional
system, they need a more powerful handheld device and high-speed
wireless,” says Dr. Tamblyn. Additionally, doctors have
requested interfaces with the lab so they get their lab results
automatically on the iPAQ, and the capacity to dictate into the
iPAQ to record their notes.
|
