Montreal researchers and developers
test tele-cardiology applications

By Andy Shaw

You could say that Montreal is at the heart of Canada’s telehealth development, if you’ll pardon the pun, because nowhere else are there more research and pilot projects involving the remote monitoring and care of cardiac patients.

Long known for their telehealth investigations at local universities, Montreal researchers and developers have recently taken their telesmarts to the streets.

In suburban Laval, for example, they’re recruiting 500 residents with high blood pressure to be part of a $ 1 million pilot program to control hypertension at home via an automated telephone system. Across town meanwhile, officials at Andromed Inc., in St-Laurent are about to pilot their unique “Home Telemonitoring System” that enables caregivers to wirelessly monitor the vital signs of thousands of unseen cardio-vascular patients at once. Both developments promise to become powerful weapons in the war against cardiac disease, the one sickness that kills and costs us the most.

The leaders of the Laval pilot, the first of its kind in North America, have given their project the acronym “LOYAL” for its goal of: “Lowering blood pressure by improving cOmpliance with hYpertension therapy through the Assistance of technoLogy.” Supporting their nine-month undertaking is Pfizer Canada’s million-dollar contribution as well as the co-ordination of the Centre hospitalier de l’Université de Montréal Research Centre and the systems of Tagge Medical Solutions of Montreal, among other contributors.

Tagge’s interactive support system will link patients, physicians, and pharmacists with automated reports of how well patients are controlling their hypertension. If needed, automated voice calls will go out to patients reminding them to adhere properly to prescribed drug treatment.

Andromed’s wireless Home Telemonitoring System enables remote caregivers to monitor patients’ pulmonary artery pressure and breathing regularity. It also beams them graphic representations of the patients’ cardiac sounds and electrical activities. Using Andromed’s patented “electronic stethoscope”, physicians can also “auscultate” or listen to an individual patient’s lungs and heart with great clarity from afar.

Wearing non-invasive sensors, cardiac patients can go about their lives, confident their vital signs are being constantly transmitted to a telemonitoring centre. There, specially trained nurses can often catch changes in patients’ conditions before they feel any symptoms. When needed, physicians can access patients’ encrypted data via the Internet and also perform a remote evaluation – speaking to the patients directly through a built-in audiovisual link.

Andromed filed patent applications for the Home Telemonitoring System early this summer in the United States, aware that the potential market for patient monitoring there is huge. Over 112 million Americans suffer from some form of cardio-vascular disease and account for 14 percent of all hospital admissions.

Compared to Americans, Canadians are only about half as good at keeping their cardiac cool.

“Only about 13 percent of Canadians who are hypertensive have their blood pressure under control while about 25 percent of Americans do,” says Farzad Ali, a LOYAL project leader and outcomes research manager at Pfizer. “We’re not really sure why that is, but it’s clear that in both countries a very large percentage of people with the disease do not have it under control.”

What’s more disturbing, says Ali, is that too many people everywhere don’t remain faithful to their drug regimen. “Somewhere between 50 to 70 percent of hypertensive people in Canada, for instance, go off their drugs after the first year of treatment. They see it as a short term disease rather than a chronic condition.”

Better patient education would seem to be the answer, but it’s been tried and found wanting.

“Studies show that educating patients only helps a bit,” says Ali. “Having them monitor their blood pressure at home helps some more. So do telephone reminders to take their drugs. But the LOYAL project is the first to combine all three of those techniques.”

In effect, LOYAL is an attempt to take the discipline of private clinical drug trials and apply it to the general public.

“In clinical trials, you’ve got the best doctors working with the most suitable patients and their compliance is virtually guaranteed,” says Ali. “But once you move beyond trials and take your drug or treatment out into the public, normally you lose that kind of control.”

In LOYAL’s case, the control begins with the participating physicians, all general practitioners, who began recruiting 500 willing patients this summer. As they sign the patients on, the Tagge system divides them into two equal groups at random: an experimental group for LOYAL-assisted care and a comparison group that will receive normal medical care.

In the experimental group, each patient receives advice on how to take better care of their condition, a home blood pressure monitor, and access to Tagge’s integrated telephone and database support system. Once notified by the physician that the patient is a LOYAL subject, the Tagge system in turn automatically notifies the monitoring centre as well as the patient’s pharmacy.

Ali says that previous studies have shown that patients are not terribly accurate when reporting their pill-taking. “So in our case we’re putting compliance in the hands of the pharmacies, instead. The pharmacy will flag LOYAL patient records so the system will know if those patients are renewing their prescriptions appropriately.”

The telephone system will also keep watch on the experimental patients by polling them regularly for their blood pressure readings with automated calls. When the phone is answered, the system will know whether it is the patient on the line, thanks to a voice-print the system made of the patient at the outset.

The LOYAL project will continue for nine-months after the last patient is recruited. The organizers will release the project’s results some time in 2004.

But Ali admits it could be all for naught if LOYAL makes no significant impact on patient compliance. But if it does, given the ubiquitous nature of the telephone, similar automated prompting could be available to virtually all hypertensive patients.

“If we can eventually bring compliance up to say 25 percent or even 40 percent of all hypertensive patients this way, just think what that could mean to the number of strokes that will be prevented across the country and the consequent savings there will be to healthcare costs ,” says Ali.

Ali says Pfizer has agreed to be the financial backer of the LOYAL project for something other than pure commercial reasons. “It fits with our long term mission to be a ‘most valued company’ by investing in projects that may lead to long term solutions to healthcare problems,” says Ali. “So, in the LOYAL project, for example there are no Pfizer drugs directly involved.”

The drugs involved in LOYAL will be whatever the physician has already prescribed for his or her patients.

Once the patient has signed a consent form in the doctor’s office, the physician or nurse calls the Tagge system and then passes the phone to the patient for a conversation with a remote computer.

“The system asks for details about the patient including their name, their pharmacy, when they’d like to be contacted as well as other preferences and then automatically makes a random decision assigning the patient to either the experimental or the control group,” explains Stephen Maislin, Tagge’s president. “For those in the experimental group, the first call out to the patient does the full enrolment in the project.”

Again, in conversation with a computer, the experimental patient answers questions about when and how often they would like to receive reminders and other information about their treatment, as well as what day each week they would like to report their blood pressure readings.

“The physician’s office has already informed the system about what the patient’s current and target blood pressure readings should be. So each week, the system looks at the readings as they are phoned in and can make its own decisions,” explains Maislin. “If the readings are somewhat elevated, the system will go into more active coaching of the patient. Or if the readings are clearly high, it will notify the physician’s office, all automatically.”

In other words, the LOYAL project aims to improve overall patient care by automating:

• more patient awareness of their disease and its proper treatment,
• better patient compliance with their drug regime, and
• timely physician interventions.

And while at the back end of that automated process will be handled by Tagge’s state-of-the art communication platform, the front end will largely eschew modern technology such as the Internet in favour of a much older piece of equipment.

“There is a small Web element in LOYAL that will enable physicians to track the progress of their patients, but otherwise we are relying completely on the telephone,” says Maislin. “We’ve found in our other work that people really do prefer a verbal interface.”

Tagge’s software is at work for clients elsewhere, including an automated telephone support system for a post-surgery discharge program at a large hospital in the United States.

Rising American demand for such remote patient care has Montreal’s Andromed also betting its future on home telemonitoring systems. U.S. studies have shown that home monitoring of patients would reduce cardiovascular patient hospital stays by eight days on average, thereby slashing hospitalization expenses by $20,000 US per patient. Home telemonitoring would also cut readmission frequency in half.

Before invading the U.S. market, however, Victor Lonzo, vice president of research and development for Andromed, says the company is launching a pilot study this fall in Quebec. The pilot will test all the elements of the Home Telemonitoring System, including its scalable sensing and telecommunications components. The system’s key hardware component is Andromed’s Biological Sound Monitor (BSM).

“The BSM is less like a medical device and more like an electric guitar,” says Lonzo of the BSM’s acute acoustic sensitivity.

Patients in the pilot will wear a lightweight wireless transmitter at home that connects using Bluetooth wireless technology with small BSM sensors on their chests. The sensors will constantly measure pulmonary arterial pressure and lung ventilation while relaying the findings to the transmitter. ECG sensors will similarly relay data for phonocardiograms of heart sounds as well as electrocardiograms of heart rhythms. The transmitter, in turn, will encrypt both the BSM and ECG feeds before sending the data on from the patient’s home over high-speed DSL phone lines.

“The hard part was not the hardware, however” says Lonzo.“The hard part was deciding how to move the data out. So first, we settled on Bluetooth for the short-range transmission because it works very well and consumes the least power. But then we had to figure out a new way to transmit data over the long range to the monitoring centre.

The centre is designed to eventually handle thousands of patients, so we have to move the data cheaply if the system is going to be economic.”

What Andromed developed, as a result, was a modified version of the TCP/IP protocol that can move encrypted medical data faster and at less cost than ever before.

As the data from each patient stream into the monitoring centre, the patient’s readings can be monitored by nurses before piling up neatly in Andromed’s SEQUS e-Survey database.

In the case of an abnormal reading, an alert message automatically appears on screen, prompting the nurse if needed to call the doctor. With his or her electronic stethoscope plugged into a computer, the physician can remotely listen to the patient’s vital signs and also communicate with the patient in real time via streaming audio and video.

If the situation is critical, the doctor can tell the patient to go to the hospital or have an ambulance dispatched – then lift the patient’s clinical information from the database and transmit it to the hospital in advance of the patient’s arrival.

In the end, if such patient telemonitoring techniques and technologies developed by the likes of Andromed and the LOYAL project truly work, then a lot more people than just Montreal researchers can take heart.

BACK TO TOP OF PAGE

Academics to produce new e-Learning technologies

By Jerry Zeidenberg

A consortium of university researchers is seeking to create new technologies and methodologies for e-Learning using broadband networks.

Called the Pan-Canadian Health Informatics Collaboratory, the two-year-long project is focusing on the development of new, online learning techniques for health informatics.

The researchers believe the technologies they produce could be applied to many other types of e-Learning, involving a wide variety of content. (See the web site at www.health.bcit.ca/hic/index.html)

“We’re re-tooling e-Learning, to move the envelope forward,” said Dr. Francis Lau, associate professor and director, School of Health Information Science at the University of Victoria. UVic is the lead contractor for the $1.6 million project, which obtained $750,000 in funding from CANARIE, the Ottawa-based economic development agency.

“We’re experimenting with innovative features that have not been tried before,” he said. “Currently, there are systems for e-Learning that have been around for six to eight years. We’re attempting to build a second or third generation platform for online education.”

The new technologies will emphasize interactivity, including students working together as groups – linked by their computers – and making use of various forms of conferencing and videoconferencing systems. The researchers are also building systems that bring advanced Internet search engines into the courseware, making use of sophisticated technologies such as object repositories and metadata tags.

(Object repositories are storehouses of coded routines that can be shared by developers across the country, meaning they don’t have to re-invent the wheel every time they build a new application. Metadata is data that describes data, and is important for locating information on the Internet.)

Perhaps more importantly to the end-users, the emphasis is on creating systems that stimulate students much more than previous on-line courseware.

“We’re trying to engage the learners in different ways,” said Dr. Lau. Instead of the rote learning that’s found in many e-Learning systems today, the Collaboratory researchers are taking a more hands-on approach, called “constructionism” in educational theory. For example, to teach students about establishing telehealth networks, a course might ask the student to go through the steps of setting up a telehealth project.

“We’re going to build simulations, some of which will use animation,” said Dr. Lau.

Other members of the consortium include the British Columbia Institute of Technology, Dalhousie University, the University of Calgary, the University of Alberta, the University of Sherbrooke, the University of Western Ontario, the University of Waterloo, and the Michener Institute, in Toronto.

The Canadian Health Records Association and COACH, the Canadian Organization for the Advancement of Computers in Healthcare, are also members of the team, and will help test the applications.

In terms of technology, the partners are building on e-Learning systems provided by Open Text Corp. of Waterloo, Ont. They include products such as LiveLink and MeetingZone. Konova Solutions Inc., of Montreal, is also involved as a collaborator, and is assisting with software development.

Dr. Lau noted that university groups are devising pilot courseware that will incorporate new learning techniques. In brief:

• Researchers at BCIT are working on the telehealth module, in which students will learn how to set up a telehealth project.

• At the University of Western Ontario, an introduction to electronic health records is being produced.

• The Michener Institute is developing a course dealing with Picture Archiving and Communication Systems (PACS).

• At the University of Alberta, the thrust is on patient research.

• And at Dalhousie University, researchers are producing a module on HL7, a standard for transmitting healthcare information.

All of this is being designed to work on broadband networks – in particular, CANARIE’s new CA*net3, the organization’s latest computer network that boasts very high speeds. It will make real-time video conferencing, and other technologies such as ‘telepresence’, available to the application developers.

However, Dr. Lau mentioned that while many universities and some hospitals have access to CA*net3, not all do. “One of our challenges consists of how we will connect to some of the learners who don’t have access to this high-speed network,” he said.

BACK TO TOP OF PAGE

Easy Pax provides cost-effective messaging system for images, reports

By Jerry Zeidenberg

TORONTO – A Canadian company has started to score international successes with a low-cost messaging system that enables healthcare professionals to send, receive and analyze medical images and reports over the public Internet or private intranets.

The system boasts a mixture of security measures, including a patented feature that prevents images from being sent to anyone but the intended recipients.

Easy Pax Inc., which has developed the application since 1998, started marketing the technology earlier this year. So far, it has sold two large systems, including one to a radiology group in Houston which reads diagnostic images for 30 clinics and hospitals in rural Texas that have no radiologists on staff.

Instead of waiting for a radiologist to visit their clinics and hospitals, or sending their patients to Houston, the rural medical centers can now transmit computerized ultrasound, X-ray, CT and other images directly to the specialists for interpretation. The Houston-based radiologists, who are reading 1,200 cases a day, have promised a two-hour turnaround for reports, once the images are received.

The cost of the Texas implementation, which includes a server and 50 licences, is less than $500,000. That compares with Picture Archiving and Communication Systems (PACS) that can require an investment of $5 million or more.

While the Easy Pax solution doesn’t come with its own archive or certain specialized functions that are found in full-scale PACS networks- such as pre-fetching – it does connect with all types of DICOM-standard radiology and computer equipment and can query archives when a radiologist needs to see a patient’s previous exams.

It also has sophisticated viewing tools built in – such as zoom, contrast and calibration of structures contained in images.

David Koff, a French-trained radiologist who practiced medicine for 20 years before turning his attention to creating and leading Easy Pax in Toronto, noted that other types of information can be easily appended to files – such as pathology reports, endoscopic or dermatology images, or even a complete electronic patient record. This gives the physician more context when interpreting the images of patients.

Radiologists can annotate images, using text that’s superimposed on the pictures, so details are highlighted for referring physicians.

“We’ve developed a very low-cost, secure messaging system that can help referring and consulting physicians, and also helps patients,” said Dr. Koff, president of the 11-person company.

He explained the system not only provides faster delivery of readings for patients and physicians, saving them time, trouble and traveling expenses, but it also reduces costs for hospitals, health regions and clinics.

Even the costs of supplies can be greatly reduced. For example, after reading a study, a radiologist will typically print out the images for pick-up by the patient or shipment to a referring physician. This amounts to $20 to $30 per study, when using high-quality printers and paper, not to mention the expense of a courier, if required.

Such charges are eliminated by Easy Pax, since images and notes can be electronically transmitted directly to the consulting physician – there is no need for paper-based printouts.

Dr. Koff stressed that the system has been made easy to use, with the assumption that family doctors and many other physicians don’t want to spend much time learning how to install and use specialized computer packages.

For that reason, Easy Pax works with common Internet browsers, such as Explorer and Netscape. “The software is browser independent,” said Dr. Koff. “The doctor can view it on any platform.”

The software can be used with any telecommunications system, including regular dial-up, although Dr. Koff recommends an ADSL or cable connection. Easy Pax contains its own compression technology, designed by the company, which squeezes picture files by ratios of up to 15:1 while still preserving DICOM information.

This means that an MRI study – often consisting of 100 slices – can be reduced in size from some 1.5 megabytes to 150K or less. Even dial-up modems are able to handle files of this size within a minute or two.

Easy Pax also offers an ASP solution, with the server software and processing located at a data center in Mississauga, Ont. Users can tap into the application from anywhere in the world, exchanging images and reports with others around the globe.

The company earlier this year landed a contract of this type with the Italian branch of a multinational insurance company that’s seeking second opinions for radiographic images of cancer patients. “They’re sending images on to radiologists in Paris and Chicago, for another opinion, to see if there’s agreement about the diagnosis and treatments,” said Dr. Koff.

If satisfied with the results, the insurance company may expand its use of the system to corporate offices in 37 different countries, said Dr. Koff.

When it comes to the ASP model, users are charged a fee of $49 per month, plus 45 cents for each megabyte that is uploaded. For the compressed, 150K MR study, that works out to about 7 cents.

Several health systems in Canada are currently examining the Easy Pax technology, including a cancer care network.

Dr. Koff said Easy Pax contains a patent-pending security system, which prevents images from being sent to anyone beyond the intended physician. “We’ve shown this feature to several radiology companies, and they’ve told us they’ve never seen this capability before.”

Along with radiology, the technology can also be used more broadly in telehealth, since all types of images and reports can be appended as attachments and sent securely over low-cost networks.

Jeff Vachon, president of telehealth company HealthworksTMS, has been providing consulting services to Easy Pax, concentrating on market entry strategies and broadening the capabilities of the software to create a clinical messaging system for store-and-forward telehealth applications.

BACK TO TOP OF PAGE

Measuring help desk effectiveness with strategies developed in call centre industry

By Stephen Tucker

Measuring the effectiveness of the help desk at Baycrest Hospital proved to be a far more difficult problem than we had originally thought. At conferences and in casual conversation with peers in healthcare IT, it became obvious that Baycrest was not alone. Others were exasperated with help desk measurement, analysis, and effective implementation of service improvement.

This article describes how Baycrest is teaming up with a call-center provider in order to develop analysis methods appropriate and beneficial to healthcare IT.

Background: The Baycrest help desk fields a wide range of calls. Users want help remembering a password, assistance with e-mail, or a solution to a phone problem. With more than 10 people handling approximately 1,800 help desk inquiries a month, we desperately wanted to be able to appropriately measure and improve not only service but also IT system performance. Measurement can serve as an early warning system helping to identify bugs and other system related performance issues.

Our first measurements took place a year ago. We called a percentage of those who used the help desk and asked them about their experience. The results were all over the map. There was no way of using the information to improve service or to develop an early warning alert for system issues.

Part of the problem rested on unfocused survey questions and our sampling method. Another was that we were asking the technicians to, in effect, evaluate themselves. There was bias in the data collection, as help desk personnel tried to justify rather than objectively gather information.

When technicians got busy, follow-up calls fell off. Sometimes one user was called multiple times by error. This caused understandable user frustration. At the same time, we were trying to develop a “bug scorecard” to help identify system issues. The scorecard effort did not give us consistent, useable data.

Subsequently we tried a periodic e-mail survey. The survey was sent to all users who had accessed the help desk in the preceding period – we tried various time frames from a week to a month. Only a small percent of users replied. Questions offered a range of answers from satisfied to dissatisfied. Most respondents went right down the middle or gave us the highest marks. The small sample size and the way people were answering did not provide us with anything useful.

Call centers seemed to offer a solution: The call center industry has performance measurements. We thought their procedures might be effectively applied by us. After talking with a few companies, we found that the measurements were not right for an IT help desk. They did not dig out the service and the system performance data we needed. Ultimately, we were introduced to Alchemy, a call center company located in Sault St. Marie, Ont.

Alchemy had for some time recognized an opportunity to expand services into the healthcare market, but they felt they needed a partner in order to fully understand the unique measurement requirements of an IT help desk. After our initial conversations it appeared that both sides would gain through a joint effort.

During several weeks earlier this year, Baycrest and Alchemy developed a simple, yet highly focused survey of five to seven questions. Alchemy suggested surveying each month a random sample of 5 percent of those served by the help desk. In our case, that would be roughly 90 people. Alchemy said the sample is small enough to be manageable and cost effective, while providing statistically relevant data.

The process: At the end of each month, Baycrest e-mails to Alchemy basic contact information on everyone using the help desk during that period. The information is contained in a Microsoft Excel spreadsheet. Alchemy reps call Baycrest Centre personnel and conduct a brief telephone interview. They stop calling after 90 surveys have been completed. The responses to the questions are coded into a database.

Subsequently, Alchemy forwards us a detailed report. Baycrest is charged a nominal cost for each completed survey. We are also charged a small monthly administrative fee. We had our first report roughly 60 days after beginning work with Alchemy. We do receive a few negative comments from those surveyed who are critical about an outside firm calling them. But for the most part, Baycrest staff is behind this. It requires only minutes, and the majority of the staff understand that this is being done to help improve service.

The findings: The report shows us how we are doing in key areas of user satisfaction, such as successful closure rate, average time to completion, and early warning on system bugs. Importantly it provides us a trend-line comparison with past months. This gives us a way of quickly jumping on problems rather than letting them stray too far out of acceptable limits.

Interestingly, the survey tracks the individual performance of our help desk technicians. This will be invaluable as a management and professional development tool. For the first time, we will have a measurement on performance that enables us to appropriately coach, mentor, and reward. The data will be invaluable at annual performance reviews.

One thing we learned was that our customer service skills could be improved. We subsequently contracted with an outside firm to provide customer service training. The help desk personnel enthusiastically received the training. After all, Baycrest was making an investment in their future and helping them to do their jobs more professionally.

We also realized that we are asked for help in areas that perhaps we should not be, for example, help with software applications such as Microsoft Word. We help the caller. However, this information has stimulated us to investigate ways of solving the problem with means other than help desk intervention.

We discovered that users are not always aware that we’ve solved their problem. This typically happens when a technician works on their computer after hours. We now have preprinted post-it notes to leave at their workstation. We also send an e-mail alerting users that the problem has been taken care of. These are small easily taken steps; yet closing the loop with our internal customers builds trust and confidence.

An insight we gained from this project is that we can affect help desk performance with the questions we ask. For instance, if we begin to develop questions on speed of response, the help desk staff will focus on improving scores in that area. Survey construction and feedback begin to be tools for innovative ways of directing continuous improvement.

Recently, we used the trend information and other data to prepare goals, objectives, and a budget for improving help desk operations for the upcoming year. Having hard data to prepare an accurate budget is an important benefit and one we did not initially anticipate.

For example, we now have facts on staffing and equipment needs to take before senior management. And as we start looking at what’s happening at the other institutions in terms of broad-based trends, we should be able to make even better, more informed decisions.

As Alchemy expands its number of healthcare customers, it will begin providing industry-wide comparative data on best practices and other performance indicators to its clients. It is the kind of data that effective help desk management requires.

Stephen Tucker is Director, Information Technology, for Baycrest Centre for Geriatric Care in Toronto, Ont.

BACK TO TOP OF PAGE

HOME - CURRENT ISSUE - ABOUT US - SUBSCRIBE - ADVERTISE - ARCHIVES - CONTACT US - EVENTS - LINKS