Volume 19, Number 5
Deborah Compeau's research shows the value of a bottom-up approach to building an electronic medical health system.
The UK's National Health Service recently announced the scrapping of a decade-long program to build a national electronic health record system. The government acknowledged that its approach, based on a centralized model of planning, was not getting results.
Nobody questions the benefits of electronic medical records. With health costs rising steeply, electronic medical records hold the promise to help control costs and improve quality of care. But the failure in the UK shows that it's important to go about it the right way.
In Canada the federal government formed Canada Health Infoway in 2001 to develop a country-wide electronic health information system. It's a huge undertaking, but progress is being made. Still, there are many who claim that the entire industry is moving too slowly.
Deborah Compeau is a professor of management information systems at Ivey. She understands the enormous challenge and complexity of creating one electronic health record for the entire country. She's not surprised by the experience in the UK. "The accepted model is to get all the stakeholders together, create a strategic plan, set the mission and vision, and then execute the plan to create a single integrated system," says Compeau. "This approach sounds logical and systematic, but it's been shown not to work in situations like this."
Compeau sees a national electronic health network as an ecosystem - a collection of large and small organizations with disparate goals, all interacting as one dynamic entity. She argues that an ecosystem can't be planned and constructed from the top down. "Policy makers need to think about how to cultivate it, how to have it emerge from a more iterative and evolutionary process." Rather than envisaging a massive infrastructure, the cultivation approach focuses on small building blocks in the system, and tries to figure out how to connect a few at a time.
Compeau recently completed a study of a small electronic health system that demonstrates the success of the cultivation approach. The system was the idea of one hospital and a small group of physicians, who agreed to share lab tests and other data. This data was previously being sent from the hospital by fax, and then scanned into the physicians' records. The partners were looking for a more efficient solution.
The physicians and hospital had their own technologies, but they cobbled together something that allowed them to share some data. The system worked, but had flaws. There were times, for example, when the system simply stopped sending, and others when the physicians received duplicate reports or reports for unknown patients. Getting support was sometimes difficult, since problems involved multiple vendors' products. Physicians also had very limited choice about what reports they received - they had to take the standard set without the ability to customize.
Despite this, by 2012 the system had expanded to include three hospital systems and 300 primary care physicians. Compeau was asked to evaluate the technology to help understand the clinical benefits and provide a framework for future projects.
She began her study with selected in-depth interviews, and followed up with a survey to all users (which received a 44 percent response rate). She found to her surprise that most users liked the technology. "This finding was counterintuitive, as previous research suggests that for people to like it there needs to be a high degree of system quality and reliability." Clinicians reported a number of benefits that improved quality, such as better follow-up with patients, staff ease of response to patient inquiries, and improved ability to search for patient information. The time saved in faxing and scanning enabled physicians to either reduce staff or redeploy them to other more proactive activities.
This technology is a good example of the cultivation approach to building electronic systems. "The partners opportunistically started doing things to see what would work," says Compeau. "The resulting system wasn't perfect, but it got the job done." Her findings confirmed the clinical value in continuing to implement the technology and make additional investments to improve it. The system is now being redeveloped and expanded throughout the province.
Compeau acknowledges that the bottom-up approach carries risks. One risk is ending up with building blocks that simply can't be connected, and must be rebuilt. "This wouldn't happen with a centralized approach, but centrally planned systems tend to be so large that they can't get implemented," she says.
Compeau likens the process to that of the Internet. "Nobody set out to build a global communications network," she says. "The Internet is a lot of little pieces that somehow connect together. If you look at it from a top-down mindset, you would say it won't work - but it does."