Doctors at the University of Illinois Medical Center in Chicago used to hide patients' medical charts under hospital beds. Better that than let a precious chart descend to the bureaucratic black hole known as the medical records storage department. Let the chart go and a doctor might never see it again. So the doctors would hide it until they were done with it. "We called it hoarding records," says Dr. Bill Galanter, who used to hide his patients' records in his office rather than under their beds for fear the mattresses would get changed. Now those records are electronic and accessible from anywhere in the hospital or the Internet. Hard copies don't exist anymore. When the medical center built its new outpatient center in 1997, it did not include a records storage room. "We burned the bridge. No paper," says Joy Keeler, the IS leader of the medical records conversion, drawing out those last two words while boring a friendly hole in your forehead with her intense brown eyes.
Keeler is a born campaigner, a small, thin wisp who burns very brightly. For five years she has been pounding the halls of the motley scattering of old and new buildings that compose the medical center, trying to convince doctors to change the way they practice medicine. If that sounds like a grandiose goal, it's because she had no choice. She could not do her job without first changing how doctors did theirs. Change management is important to any system implementation, but it's everything in a hospital. There is no value to an electronic medical record if doctors don't use it. It is an all-or-nothing proposition.
The University of Illinois Medical Center won an Enterprise Value Award because of a brilliantly executed change management strategy, through which it managed to get the toughest users on the planet to lift their heads up long enough from the work of saving lives to change the way they deliver care to patients. Most believe they can do their work better and more cost-effectively than they did before the system. Computer system value doesn't get much better than that.
A Big-Bang Implementation
Change is hard. Despite their complaints about paper medical records, doctors really don't want to give them up. Only 4 percent to 15 percent of hospitals nationwide have electronic medical records today, according to various industry estimates. It's still faster to scribble a note than it is to sit at a computer and enter data.
Speed is an obsession with doctors because they see between 16 and 30 patients a day. "I have a maximum of 20 minutes to do everything when I see a patient," says Dr. Patrick Tranmer, head of the Department of Family Medicine at the medical center. "I have to find out what's wrong, get their history, do a physical exam, make a phone call, write a prescription, instruct the patient, make a follow-up appointment and then educate a student doctor about what I've just done."
Seconds matter when you're in that 20-minute zone, says Tranmer, who stares impatiently at his notebook computer while waiting for Gemini, as the medical center's assemblage of applications is known, to come up on his screen. Ten seconds pass. He pushes the computer away. "It's slow today," he says.
Gemini was balky because a chip inside one of the Compaq Alpha computers that runs the system burned up the day before, bringing down the entire system. Ironically, it had been so long since the last outage that the doctors forgot how to access the read-only backup database that Keeler had built for such emergencies. Angry calls streamed into IS. One doctor demanded, "Do you want me to waste time asking my patients about their medical histories?" That's how valuable the electronic records have become to the doctors--many think they can't do their job adequately without them.
The doctors' characteristic impatience explains why Keeler chose a big-bang implementation--the riskiest possible strategy. There would be no pilots. The system went online in two big gulps: the outpatient clinics in early 1999 and the inpatient hospital later that year. Rollouts to selected groups would have collapsed at the intersection of the paper and electronic records, says Tranmer. "Doctors would have said, Now this is twice as hard. Forget it." Worse, he says, they would have rejected the electronic system because the old way was more familiar.
Getting those impatient people to accept a new system required extraordinary patience at the highest administrative levels. When a financial crisis hit the hospital in 1998, outside consultants recommended outsourcing IT and with it the Gemini project, which had begun in 1996. But in 1999, Dr. Charles Rice arrived. Rice, the vice chancellor for health affairs, moved Keeler's IT group from operations into his administrative department to subsume the IT budget and take outsourcing off the table.
"This is an educational institution," says Rice. "Part of our job is to educate the next generation of medical workers. I don't know how you do that with an [outsourcer]." During those difficult days, Rice insists that the board never questioned the investment in the electronic medical record project--which is now at US$11.2 million and counting, with a $1 million annual maintenance budget.
Keeler is certain the project would have foundered without such support. "First and foremost this has to be the strategy of the organization," she says. "This isn't a project; this is a culture change. Transformation has to be the goal of the organization--not just of the chief medical officer or the CIO."
Finding a Better Way to Doctor
This emphasis on transformation--not implementation--drove Keeler's dealings with the project's startup software vendor, Kansas City, Mo.-based Cerner. As she pushed for applications to be built, Cerner built them: separate, discrete applications for all the tasks that doctors and nurses perform when they interact with a medical record. As Keeler watched the icons multiply across her system test screen like popcorn, she sent Cerner back to the drawing board. "We're not implementing applications; we're trying to transform the process," says Keeler determinedly. "Cerner didn't understand that. What they gave me in 1996 was a bunch of applications. That's why I went back to them and said, I need a desktop. I need an inbox."
Today, the system mimics the look and functionality of an e-mail inbox, except that it is more powerful, combining messaging capabilities with workflow from the different applications hidden under its skin. For example, when a doctor logs on, she'll immediately see her phone messages, reports needing her signature and alerts from other doctors who need her help. Keeler knew the powerful inbox metaphor would be an excellent way to draw reluctant doctors into the guts of the application--the medical record.
The screen immediately shows Keeler's primary intent with the system: improved communication. Doctors used to play phone tag for days to coordinate referrals and conferences. With Gemini, they can trade messages that have the specific content attached to them for action. Nurses used to pointlessly page doctors because they couldn't read their handwriting or because they were bouncing messages between doctors, patients and pharmacies. Now they can leave a message for the doctor in the system, and the doctor can automatically chart the prescription and send it to the pharmacy. Nurses tell Keeler that automating the prescription process saves them an hour each shift, or about $1.2 million annually.
Accessing information is the other process that has been transformed. Chasing down X-rays and lab reports devoured doctors' days. Doctors, nurses and patients endured "nonvisits," where doctors sent patients home because they didn't have the information to treat them or had to order duplicate tests because they couldn't find the original results. Now the labs simply access patient records through Gemini and attach their reports when they are ready. Although not all tests or reports are yet electronic, the stack of paper associated with one patient at the hospital has decreased from about 300 pages to about 75, says Keeler, and it continues to drop each year.
Today, the benefits of the electronic record are inarguable. Nearly 90 percent of doctors at the center believe communication and accessibility have improved since the paper days, according to surveys performed by the project steering committee. Though he has no hard numbers to prove it, John DeNardo, the medical center's executive director, insists that the system has made his business more efficient, allowing doctors to see more patients per day and reducing the average length of stay in the hospital. But when Keeler first rolled out the system, those benefits weren't immediately apparent. Initially, doctors resisted because they couldn't or wouldn't see past the extra time it took to create a medical record electronically.
This is why Keeler enlisted a powerful constituency--nurses--to help convert the doctors. Part of the reason doctors could put up with the inefficiencies of the paper medical record is that they passed much of the labor off to the nurses. Can't find an X-ray? Send a nurse after it. Can't locate a doctor? Ask a nurse to find him.
The Revolt of the Nurses
Revolutions start when the downtrodden see a way out of their predicament, so Keeler focused on the nurses. She handpicked those who were viewed as leaders by their peers. Then she trained them as power users. They came to know so much about the system that they didn't just train other nurses, they influenced the doctors' acceptance of the system.
Doctors were too busy to listen--at first. "I said the same thing to the nurses after the system was installed that I did before it went in," recalls Tranmer: "'Get me the X-ray.' Except I noticed that instead of 10 minutes it took the nurse one minute to get it. After that happened again and again, I finally said, 'How did you do that?'" Eventually Tranmer learned how to do it himself and became one of the biggest backers of Gemini.
Roseanne McBride manages the nurses in the "stepdown" clinic, a 27-bed unit that monitors patients who are coming out of intensive care until they can be transferred to regular rooms in the hospital. She has been around the medical center long enough--15 years--to identify the primary difference between the old paper-based system and Gemini: doctor compliance. "This was our opportunity to say to the doctors, 'We're not doing your documentation for you anymore; you do it,'" she says. "It put ownership of the information back where it belonged."
Selling the Doctors
To get doctors to use the system, Keeler had to make one major concession: allow them to enter their clinical notes and observations any way they wanted. They can type free-form or dictate data into the record and aren't constrained by specific computer-generated forms. Galanter, who is leading the next phase of the project of adding KM capabilities, lives with the effects of that concession. "We made a conscious decision at the beginning that we would not force people to use the system in a way that would make decision support easier down the road," he says. "If we had forced them to enter information in such a way that it would generate more discrete data, we would never have gotten everyone to use the system."
Until document management systems can make sense of the free-form data that doctors enter, Galanter is focusing on keeping doctors and nurses from making mistakes with medications by programming alerts into the system that warn a doctor when a medication conflicts with another that the patient is already taking, for example. According to David Bates, M.D., chief of general medicine at Harvard's Brigham & Women's Hospital in Boston, computerized medication order-entry systems can reduce medication errors by 55 percent. Though medication errors don't always harm a patient, when they do the costs are high. In a study, Bates found that two out of every 100 hospital patients suffer an adverse drug reaction--at an average cost to the hospital of $5,857 each--that is preventable through better medication management.
Yet the biggest value the system provides is simply its ability to deliver all the data that exists about a patient to any doctor, anywhere, anytime, so that the doctor can apply his own knowledge management to the data in the system. When Galanter was on call recently, another doctor's patient called him saying she had lost her medications and couldn't remember anything about them except their color. "I was able to get on the Internet, get into her record, see who the doctor was, see what he had prescribed and call in the order to her pharmacy," he says.
It's a story he repeats in his practice every day. "I know that I am providing infinitely better care every day because I have all the information in front of me," says Galanter. "I make fewer mistakes in care because I know so much more than I used to."
The Battle not Won
Despite the value Gemini has demonstrably delivered, a minority of doctors in the hospital still refuse to have anything to do with it. Two departments, ophthalmology and otolaryngology, have never used the system for various reasons, says Tranmer, and among the individual resisters (about 20 out of 250 doctors who see patients at the hospital), the majority are surgeons. Acknowledging the opposition, Dr. Larry Ross, head of the department of urology and a practicing surgeon, says, "Surgeons tend to be very results-oriented and want to get in and out as quickly as possible, as opposed to the more introspective specialties like internal medicine." In his department, he sees older surgeons and those who dislike using computers as the biggest resisters. But Ross is a backer of the new system and says all new resident surgeons are being trained to use it.
Overall, 20 percent of doctors described themselves as "reluctant users" of the system in a survey in 2000. Though Keeler has not surveyed the doctors again, she is certain that number has decreased now that the system speed has been goosed up and more computer terminals have been added.
Keeler, who became associate vice chancellor in 2002 to focus on the development of strategic initiatives for the next phase of Gemini, has brought the hospital a long way. "We're about 20 percent done," adds Rice. The other 80 percent will come through a gradual introduction of more discipline in the order-entry methods, a tighter link between the medical record and the billing system, and making records available to patients over the Internet.
"Getting everybody electronic is the Holy Grail," says Enterprise Value Awards judge John Glaser, vice president and CIO of Partners HealthCare System in Boston. "[Illinois Medical has] laid the foundation, but they can't stop there and say they've transformed patient care. This system puts them in the position to begin that transformation."
UNIVERSITY OF ILLINOIS MEDICAL CENTER
CORE BUSINESS Patient care
REVENUE $300 million; 450 beds
KEYS TO SUCCESS
- Designed system to resemble an e-mail program to ensure simplicity
- Addressed most frustrating problem first to show immediate results
- Enlisted powerful constituency before tackling toughest users
- Imposed few constraints to data entry to reduce resistance
-- CIO US
Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.