This article is the first in a series that will detail the journey to implement an electronic medical record (EMR) at UT Southwestern Medical Center University Hospitals in Dallas. These articles will feature the "front-line" perspective of clinical and IT professionals as we work to transform the delivery of patient care in our integrated health system. This series will detail our progress, our challenges, our dreams and our realities.

IT background

Our enterprise clinical IT vendor is Verona, Wis.-based Epic Systems. Over the past five years, we have implemented the Epic ambulatory EMR, scheduling, registration and operating room modules. Our inpatient environment has numerous disparate systems: We currently have two different pharmacy systems, two different order-entry systems, and limited electronic documentation.

Suresh Gunasekaran

We hope to extend our Epic investment into our inpatient environment. Major components of the new hospital EMR will be an enterprise pharmacy system, results repository, and clinical orders and documentation system. Specialty documentation modules will include emergency department, labor and delivery, as well as intensive care unit. Underpinning our entire implementation will be the interface of medical devices into our electronic documentation.

Why we are implementing an EMR

In 2005, with the arrival of a new hospital management team, the University Hospitals refocused around three organizational goals: patient safety, patient satisfaction, and operational efficiency.

We began numerous efforts to improve patient safety in the areas of medication administration, discharge education, clinical protocol development, and surgical services. Spurred in part by regulatory pressures as well as our own performance improvement goals, we significantly increased our data collection of clinical indicators.

Every clinical improvement effort seemed to confirm the same basic themes:

1. Clinical practice variation dramatically increased our risk of errors.

2. Our disparate systems drove our disparate processes which drove our disparate clinician accountabilities.

3. Accurate clinical data is crucial to drive performance improvement and change clinician practice. It's almost impossible to collect this data from disparate systems and paper chart abstractions.

For these reasons, our leadership committed to implement a single-vendor EMR suite to serve as a foundation to improve patient safety. Though other EMR vendors were considered, we recognized the value of extending the same Epic tool that was used in our ambulatory environment. The promise of a single medication administration record, a common results repository, and single physician user interface were just too compelling to deny. We have committed almost $20 million over two years to implement this suite of applications in our inpatient environment.

Many EMR projects are doomed from the beginning with unrealistic expectations, incomplete budgets, and lack of operational planning. We've worked diligently over the past three months to prepare our organization for the kick-off of our EMR project.

Developing the implementation plan

Developing the plan was one of the truly most difficult analytical exercises that we have conducted in our department. We struggled with many competing realities:

1. Our organization wanted to implement the EMR very quickly. However, our inpatient nurses have little computer experience today.

2. Some IT staff favored replacing all systems at once to minimize interfaces, while others favored using an incremental approach to allow better support of each implemented module.

3. We knew that the more modules that we implemented at once, the more staff that would be required.

Ultimately, we decided on an implementation plan that balanced the organizational impact and minimized the IT risk to reasonable levels. We decided on a two year plan: in year one, we will implement results reporting, the emergency room module, pharmacy and electronic MAR, as well as unit clerk order entry and basic nurse documentation. In year two, we will implement CPOE and specialty documentation modules.

Lessons we learned while developing our plan include:

1. Acknowledge project risks. Our IT department has never implemented an inpatient EMR. Our nurses have never used an electronic documentation system. Everyone pressured me to make the timeline and scope more aggressive. I've stood my ground so far.

2. Factor vendor product maturity. There are many Epic EMR customers. Still, there are more Epic customers currently live on the products we plan to implement in year one than in year two.

3. Avoid the Clinical Hot Zone. We have consciously decided not to implement any system during the November-March higher census months.

Developing the budget

The cost of implementing clinical information systems is a very legitimate concern for hospital management teams. Most of the project expense is labor related for internal staff, vendors, and consultants. The inability to make decisions, to build effectively, and to get user acceptance in a timely manner can quickly delay the project and drive up project costs.

Lessons we learned during our budget development include:

1. Hidden Cost: Infrastructure. We included the cost of network wiring to the patient room and redundancy in our network closets. We also budgeted for ergonomic devices to go with our PCs, as well as redundant hardware.

2. Operational Backfill. We budgeted for the backfill of clinical personnel in key areas to participate in design and testing. We also budgeted dollars to offset the impact of staff attending EMR training sessions.

3. Pay your Physicians. We have decided not to fund a single physician. Rather we are funding "a piece" of several physicians to create an inter-disciplinary physician champion team.

Let the clinical leadership lead

The CIO is not the leader of the EMR effort at the University Hospitals. Rather the leaders of our EMR effort are our chief medical officer and the director of clinical education/quality. We have named a nurse informaticist in our clinical education department as our program coordinator. We have created a physician team as a leadership council to drive clinical content development and system usability.

This is going to be a very difficult challenge for our IT department, as we are very accustomed to leading system implementations. We have had to do some internal soul searching on where we are to fit into this "New Frontier." We have decided that our role is as partners, and that our mission will be to drive getting the job done right.