Five Lessons From Seattle On Adopting Electronic Medical Records

Third of an occasional series on health information technology.

SEATTLE – Atop a hill here, three of Washington state’s pre-eminent hospital systems sit within blocks of each other, equipped with state-of-the-art electronic medical record systems that track test results, send warnings about dangerous drug interactions and provide medical histories.

But a patient crossing the street from one hospital to another would be wise to bring paper records: The systems, made by different manufacturers, can’t talk to each other.

For much of the country, linking the electronic records of doctors, hospitals and clinics remains an elusive goal. Even in this tech-savvy city, “no one is quite there yet,” says Jim Bender, medical director for health information at Seattle’s Virginia Mason Medical Center.

Among the reasons: cost, computer systems that aren’t compatible with rival systems, resistance among physicians and privacy concerns. Overcoming the obstacles, Bender says, “will take federal will and money.”

Money is on the way. Under the federal-stimulus legislation, the government plans to spend $32 billion on health-information technology over the next 10 years, and projects $13 billion in savings by doing so. Most of the money will go to doctors and hospitals.

But there are risks. Unless the money is doled out carefully, the money “may go down a rathole,” says Janice Newell, chief information officer for Swedish Medical Center, another major hospital here.

That’s one of the lessons of Seattle’s experience, here are more:

1. Make sure records can be shared.

Swedish spent four years and $120 million rolling out an electronic medical record system, which now links all three of its hospitals, along with the majority of its 40 clinics. But it can’t share data with either Virginia Mason or the University of Washington, which is the third big hospital here.
Manufacturers have been slow to create systems that work together because they have wanted to emphasize their uniqueness to gain market share, Bender says.

To work around compatibility problems, hospitals have figured out ways for their staffs to see some of each other’s information. Virginia Mason staff, for example, can view the electronic records of patients served by Group Health Cooperative of Puget Sound, a large HMO, but “it’s not a true data exchange,” says Bender.

Change is coming. One of the largest makers of electronic medical records has created a way for its systems to communicate with rival products, says Tom Wood, the physician who oversees health-information technology at Swedish. That, he says, will allow Swedish and the other hospitals to eventually “talk to each other.”

Newell, the Swedish information chief, warns that money could be wasted if stimulus dollars are spent by doctors and hospitals on a hodge-podge of electronic medical record systems. Instead, she says, regulators should provide incentives for providers “to use existing systems” and then improve technology so those systems can easily share data.

2. Don’t move too fast.

Swedish Medical Center’s Wood warns there are risks in moving too quickly–especially in installing systems that allow physicians to enter medication or treatment orders electronically.

If hospitals rush into installing such systems, “it will kill people,” says Wood. “If you don’t do it right, physician orders can show up in the wrong place, be confusing or come at the wrong time,” he says.

Thomas Payne, medical director for information technology at the University of Washington, says his organization is now working on installing computer order entry for prescriptions.

Such systems, he says, have many benefits, but they “also have a risk for side effects.” Good programs help doctors quickly and easily distinguish between drugs with similar names, flagging differences between two similar drugs by capitalizing letters that differ in the names, for example.

But testing the systems and training the staff takes time-which raises concerns about the 2011 deadline for the first batch of stimulus money, says Payne. Especially in smaller facilities with smaller staffs, “we need to make sure we’re giving them time to do this properly, he says.”

3. Computers alone won’t slow rising health costs.

Advocates say that information technology can save money by making health systems more efficient.

Still, bigger savings require more fundamental changes, such as removing financial incentives to do duplicative tests and provide unnecessary care.

“As long as you will pay every time I do that test, the incentive is to do that test as many times as I can,” says Stephen Lieber, CEO of the Healthcare Information and Management Systems Society, a nonprofit group whose members include technology companies. Without changing the way insurers pay for care, “you won’t drive costs down.”

Payne agrees. “All of this attention on IT is terrific but it needs to be coupled with reform of the economics of health care,” he says.

4. Some progress is better than none.

Even with its limitations, Seattle – with systems in place in several of its major hospitals – is way ahead of many areas.

Nationwide, only 1.5% of hospitals have a full electronic medical records system, according to a recent report in the New England Journal of Medicine. Another 7.6% have a basic system in at least one area of the hospital. About 12% of doctors currently use electronic medical records, estimates the Congressional Budget Office.

“We don’t have to thumb through a paper chart to find things,” says Virginia Mason’s Bender. “All phone messaging is done electronically, so we don’t have waste and harm from lost messages. We’ve taken the hour and a half delay of getting medications or treatment to patients down to minutes.”

So, he says, even though the systems are far from perfect “there is a tremendous amount of value to create on the way.”

5. Even so, a national network is needed.
David Brailer, who served as National Coordinator for Health Information Technology from 2004 to 2006, says that the lack of a common network – not the compatibility of existing medical record systems – is the biggest hurdle to widespread adoption.

He likens a network to telephone lines, which tie together various types of phones and communication equipment so information can move around. Until such a network is well developed for health information, demand for electronic medical records will be slow, even with stimulus money, Brailer says.

“Imagine if companies tried to sell cell phones connected to some networks, but not others or that would call only certain area codes,” says Brailer, now chairman of Health Evolution Partners, an investment management firm in San Francisco. “The equivalent is trying to get doctors to switch to electronic medical records when they have, at best, a patchwork network to connect to.”