October 17, 2006

Implementation of the EPIC Electronic Medical Record/Physician Order- Entry System

By O'Brien, Mary S


JKL Healthcare System is a not-for-profit organization that operates three acute care hospitals with five off-site ambulatory locations, a research institute, a 450-employee physician group with 50 local physician offices, and a home care services company. In addition, JKL is affiliated with a top-rated university medical school in the area. Its three acute care hospitals are Hospital J, Hospital K, and Hospital L. This case report describes activities primarily at Hospital L, a hospital that merged with JKL in 2000 and is located in an affluent community of about 40,000 people.


In July 2001 JKL Healthcare embarked on a journey to implement a complete electronic medical record and physician order-entry system. Most of the physicians admitting to Hospital L were independent physicians who were also on staff at two or three other competing hospitals in the area.

The implementation of this new electronic medical record system needed to be perfectly orchestrated to prevent admitting physicians from redirecting admissions to the competing hospitals. The first challenge was managing this implementation to make it easier for physicians to practice at Hospital L so they would continue to direct their patients there. The second challenge was training 1,700 employees in a completely new work process. Both physicians and staff had to be trained adequately to avoid adverse patient outcomes during implementation and to meet the goals of improved quality of care.


JKL's decision to implement an electronic medical record and physician order-entry system arose from the need to reduce medical errors and costs while improving efficiency and quality of care. JKL chose to be a leader in this effort by implementing this new system that focuses on the patient instead of the provider.

The planned cost for the new system, called EPIC, was $35 million. The vision for the new paperless system was that every physician and every clinician would use the system. Only with 100 percent adoption would the system meet its goals.

The major challenges facing Hospital L at the time of the EPIC installation included the following:

* Training 450 physicians

* Training 1,700 employees

* Maintaining the business during implementation

Physician Training

All physicians were required to attend a 16-hour training session to be certified. The physicians had to have the certification to admit patients or to consult. However, Hospital L physicians did not look favorably on the 16-hour training requirement. Unlike the physicians at Hospitals J and K, Hospital L physicians did not have residents or physician assistants to round on their inpatients or to manage their office visits.

JKL clinicians were at various stages of proficiency with technology. Several physicians near retirement had spent 40 years writing patient notes in charts and orders on paper. Most of them did not use e-mail, let alone use an electronic medical record. The physician order-entry function required physicians to spend much more time typing orders. If they were not adequately trained, it could adversely affect patient care; yet if the training was too complicated, physicians would not spend time doing it and would instead send their patients somewhere else. Fifty percent of physicians admitting to Hospital L also admitted patients to competing hospitals in the area. If EPIC did not make medical practice easier for physicians, then they might take their admissions elsewhere.

Nonphysician Staff Training

Approximately 1,700 nonphysician employees worked at Hospital L at the time of EPIC installation. The nonphysician staff were required to attend a 16-hour training session and then pass a test to become certified. The training for EPIC , included several different modules, including scheduling, registration, outpatient documentation, and inpatient documentation. Most employees required training in more than one module. The training included functionality of the software, new workflow patterns, and dramatic new ways to deliver patient care. Learning these new processes required time away from patient care and the significant expense of overtime and agency nurses. However, if staff were not adequately trained, patient care might be adversely affected.

Maintaining the Business

Fifty percent of the average daily census of patients at Hospital L had physicians who admitted to more than one hospital. Because these physicians were already on staff at the other hospitals, it was no bother for them to redirect patients. If the EPIC system created too much extra work for the physicians, they could redirect their admissions and outpatient ancillaries to the competing hospitals. Nurses and other staff would have a learning curve on the new system. However, patients needed the same timely and safe care while training was ongoing.


The general conclusion reached by the administrative team at Hospital L was to "overcompensate" with training physicians on the EPIC system. The goal was to quickly make the physicians experts in EPIC so that the functions that made their practices more efficient would be recognized immediately and the business would grow. For employees, the decision was made to cover the expense of adequate training and replacement nurses but to provide incentives to staff to take their training on days that they were not scheduled to work.

Physician Training

The training for physicians at Hospital L began in October 2003. Because all physicians at JKL were on staff at all three hospitals, some physicians who admitted to Hospital L had already been trained. This was especially true among the house physicians, such as the radiologists, pathologists, anesthesiologists, and emergency physicians, who rotated across all three hospitals.

Hospital L physicians were invited to be members of the JKL physician advisory committee. This advisory committee made all the clinical decisions about order templates, history/physical smart sets, and all medication alerts for contraindications. Very early in the training sessions, the physicians commented that the training was too long because some sections of the training were not relevant to them. Although the trainers believed it was important to instruct physicians on what their nurse colleagues were doing on EPIC, sections that were not specifically related to physicians were eliminated to reduce the training time to 12 hours.

Every physician who attended the training class and who became certified in EPIC was allowed to waive their professional staff dues of $275 for the year. All physicians received 12 hours of continuing medical education credit for the training sessions. The trainers helped each physician develop his or her personal templates for documentation and order entry, which gave physicians the opportunity to develop personal outlines, ordering strategies, and medical language. Once the physicians began to see the new capabilities of the system, they became much less resistant to the training. On each of the Saturday "go-live" dates, each physician was assigned a personal trainer to ensure that the physician could use the system proficiently on his or her own. These personal trainers were available 24 hours a day, seven days a week, for 30 days.

Nonphysician Staff Training

The implementation of the electronic documentation system was planned for December 2003. Training of staff for certification began in October 2003. The training team completed a train-the-trainer course and achieved full proficiency in every function of EPIC before classes began. Almost all of the 1,700 employees attended the 16 hours of formal classroom training.

Specially trained employees called "super users" were available on each shift for each department. Super users floated among formal classes and acted as advisors to each class. The formal classes were supplemented by additional training strategies, including the following:

* Review sessions

* Validation sessions

* Practice sessions

* Biweekly newsletters

* Super user back-up meetings

Hospital L planned for substantial expenses in the overtime and agency nurse categories because classes were held during regular hours and the cost of replacement nurses to substitute for those receiving training was high. To help reduce the cost of agency nurses, if not the overtime costs, an incentive program was developed to encourage staff to fill in for colleagues who were at training or to attend a training class at night or on the weekend for full pay. The agency costs during training at the other hospitals had been very high.

By December 2003, all Hospital L staff were trained for the first go-live date, which was when all electronic documentation was expected to be functional. After the first go-live date, staff began preparing for the second go-live date in April 2004, which was when the physician order-entry system would be functional. Each staff person attended four hours of refresher courses. Hospital L considered bringing both electronic documentation and physician order-entry systems live at the same time, but based on lessons from the first two go-lives at Hos\pitals J and K, the decision was made that separate go-live dates would create less risk of patient mistakes.

By April 2004, 1,700 employees had trained for the second phase, and the physician order-entry system went live. In the six months following the implementation of these two systems, two major upgrades have occurred, which have entailed online training for staff. An ongoing training program has been implemented for training new or transferred employees.

Maintaining the Business

To minimize the risk that Hospital L physicians would redirect business to competing hospitals, the decision was made to put extra effort into serving the physicians during implementation. The physicians were intrigued and excited by the functionality of the system, which they had learned about at the training sessions. On the first go-live date for electronic documentation, each physician was met at the front door by a personal trainer. The trainer escorted the physician to the Command Center to obtain a security code and then rounded with that physician to the patient units. Super users were stationed in every patient unit, the operating room, the emergency department, and all ancillary departments. Super users were available for the staff, and special super users were available for the physicians.

Every physician was provided a key fob, which allowed the physician to access the EPIC documentation system from their physician office, their home, or any remote site where they could access the Internet. At any time from any of these locations, physicians could write histories and physicals into EPIC or view radiology images or diagnostic test results. In April 2004, Hospital L went live with physician order entry. Saturdays were chosen for the go-live dates because physicians who admitted to Hospital L usually did rounds on Saturday mornings, and Saturdays were not as busy as Mondays. All personal trainers and Super users were available 24 hours a day, seven days a week for 30 days after each go-live date. Hospital L maintained the concierge service and the super user service for two months after each go-live date.


Nine months after the documentation go-live date and four months after the physician order-entry go-live date, the physicians who admitted patients to Hospital L said they would never want to return to a paper chart system. All 450 physicians attended training, even the physicians who were nearing retirement. Peer pressure and word- of-mouth from colleagues about the state-of-the-art system helped encourage all of the physicians to get training.

A physician survey was conducted two months after the physician order-entry go-live date. Of the 73 physicians who responded, 90 percent agreed that the EPIC system made it easier for them to do their work. Admissions at Hospital L have risen slightly, suggesting that the electronic medical record system has not caused physicians to redirect their patients away from the hospital. However, the system has not yet encouraged physicians to send many more patients to Hospital L either. One very positive spurt in census occurred during implementation because the physician support was so strong in those four weeks that many physicians admitted more patients to get practice while they had an assigned personal trainer.

Outpatient visits increased by 3 percent between May 2004 and August 2004. In July 2004, data showed that medication errors caused by illegibility and transcription were eliminated completely. In August 2004, patient satisfaction scores for overall satisfaction with care and overall nursing care at Hospital L climbed to their highest levels since January 2000.

Employees also have found that the EPIC system makes their jobs more efficient. Nurses and other healthcare professionals are not searching for charts, calling for results, or begging physicians for a written order. Everything they need is in the system, and it is all paperless. As many professionals as necessary can be looking at the same chart at the same time from many different locations. The security of the system is so comprehensive that only the professionals involved in the care of each patient can view the record.

Since April 2004, cost savings on office supplies have reached almost $50,000 and the number of medical records staff has been significantly reduced. Agency nurse costs during training and implementation were just slightly over budget (at 1 percent).

In 18 months of implementation there have been only two downtime periods of longer than two hours, and comprehensive downtime procedures have been developed in response. The staff and physicians have had a hard time returning to paper, even for only five or six hours. Both downtime periods have strengthened the commitment of staff and physicians to improve their knowledge of the comprehensive functions of the EPIC system so they never have to return to a paper system again.

In June 2004 JKL Healthcare was given an award for having implemented the most comprehensive electronic medical record system in the United States. In September 2004 JKL received the Davies Award, the information system industry's most prestigious award for accomplishments in implementing state-of-the-art systems.

Because JKL is one of the first implementers of a successful electronic medical records system in the country, it will serve as a consultation site for many other healthcare systems. JKL's overall message to other institutions undertaking such a comprehensive implementation is to involve physician leaders from the start.

Source Materials

Consultation with EPIC Systems Corporation in Madison, Wisconsin and with a group of experts provided direction in the analysis and resolution of this problem. In addition, the following source materials were used for this report:

Armoni, A. 2000. Healthcare Information Systems: Challenges of the New Millennium. Software Ltd. Publishing.

_____. 2002. Effective Healthcare Information Systems. Colorado Springs, CO: Cygnus Software, Ltd.

Breyer, G. 2003. Hardwiring for Excellence. Gulf Breeze, FL: Studer Group LLC.

Devaraj, S. 2001. The IT Payoff: Measuring the Business Value of Information Technology Investments. New York: Pearson PTR Publishing.

EPIC. "2002 and 2003 Users Group Meeting Manual." Madison, Wisconsin: EPIC.

Mary S. O'Bnen, FACHE, president, Evanston Northwestern Healthcare, Highland Park Hospital, Highland Park, Illinois

Mary O'Brien, FACHE, is president of Highland Park Hospital in Highland Park, Illinois. She started her healthcare career as a medical technologist at Evanston Hospital. Over the past 29 years at Evanston Northwestern Healthcare, she has held several leadership positions, including director of corporate training and development, vice president of performance improvement, senior vice president of hospitals and clinics, and executive vice president of Highland Park Hospital. Ms. O'Brien graduated with a Bachelor of Science degree in medical technology from St. Mary's College Notre Dame and obtained a Master of Science degree from Northwestern University. She is a member of the Policy Committee of the Illinois Hospital Association and a member of the Economic Club of Chicago. This case study represents a part of Ms. O'Brien's ACHE Fellow Project. It was voted one of the best case studies in 2005. To view this Fellow Project online, visit http://ache.org/mbership/AdvtoFellow/fellowproj.cfm.

Copyright Health Administration Press Sep/Oct 2006

(c) 2006 Journal of Healthcare Management. Provided by ProQuest Information and Learning. All rights Reserved.