This article is the eighth in a series detailing the journey to implement an EMR at UT Southwestern Medical Center University Hospitals in Dallas.

We continue to make progress towards our implementation of an electronic medical record at UT Southwestern University Hospitals.

This success could not have been achieved without the hard work and diligence of our clinicians and information technology professionals. Specifically, Thomas Lacour, M.D., our medical director for anesthesiology and Tsedey Melaku, our lead nurse informaticist for our EMR implementation, provided us the necessary clinical expertise and operational guidance to make this project a success. Within the IT department, Josh Youngblood and Donna Fitzgerald led the hard work of system build and testing for this project.

Our work is far from over and multiple initiatives are entering different phases of development. Our pharmacy implementation is nearing completion of the design phase, and moving fully into the build and test phase. Our ED implementation is also near completion of the design phase. Our unit clerk order entry implementation is just beginning the design phase.

As we have so many projects moving through the design phase, I thought I would highlight some of the lessons learned from our design work for this EMR.

Don't let the vendor drive the design session

All too often, organizations treat the EMR design process as a field trip. The vendor sends the hospital a list of people that will board the bus. The vendor will provide each of the attendees a schedule and list of activities. The vendor will also list what items each person will need to bring on the field trip. The vendor will then lead the people on the bus through the list of activities. At the end of trip, the bus driver will have the necessary information to build the EMR and the attendees will definitely have been taken for a ride.

One cannot really blame the vendors for taking this simplistic approach. In a perfect world, the vendors would have a magic wand that stops time. After they stop time, they would bring together an inter-disciplinary group of clinicians who could work continuously for weeks to design the EMR. In most organizations, this magic wand doesn't exist, so vendors must settle for the field trip approach.

Clinicians should provide the documentation of current state

The most important contribution from clinical operations can be to help the organization determine what operational processes are going to be affected as a result of the implementation of the EMR. Often, this is best led by a nurse informaticist who can both understand the capabilities of the new technologies and the demands of current clinical operations. The use of clinicians to document current processes is far superior to having technical staff perform this activity. The IT staff may have more time, but the clinical staff has the necessary depth and detail. Superior documentation and understanding of current operations can save the design team months in the end.

Large groups do not design EMRs

More than a few implementations have been derailed by design groups that have made decisions that weren't accepted by senior management or by line staff clinicians. For that reason, many organizations attempt to design an EMR by committee, including staff from all departments and at all levels. The rationale is that the more people that participated in the decision, the greater the acceptance.

We have found this not to be the case. Most importantly, larger groups are harder to keep focused and to achieve consensus. More often than not, these groups either stay at too high a level to make real decisions, or get bogged down in minute detail. We have found that forming a much smaller design group, with key clinically and operationally credible staff, is the best approach. These groups, with six to eight members at the maximum, are large enough to present diverse viewpoints and small enough to achieve a shared understanding of the issues. The outcome of the design group can then be presented to larger groups or senior level management by design group members.

Develop guiding principles

It is important to develop a set of guiding principles that will help drive the overall design and implementation of your EMR. Guiding principles help determine the key principles that must always be followed in order to achieve overall system goals. Examples would include: all results must be available within the same results- viewing capability, all orders must be built with standardized nomenclature, all physician user interfaces must follow the same style palette, all core measures will be tracked electronically, etc.

The guiding principles become particularly important during the design sessions, because they provide strategic guidance. It is all too easy to develop the ICU nursing flow sheets in a manner that is most easy to use for the ICU nurse. However, in order to meet overall goals, it may be important to design the flow sheet in a slightly different way so that the flow sheet looks consistent across the enterprise. By doing this, we are able to capture core measures in a more standard way. All too often, these enterprise guiding principles are sacrificed too early.

Remember there is never a blue sky without clouds

The promise of new technology is to re-invent virtually everything that we do. The reality of technology and particularly EMR technology is to incrementally improve the way we deliver care if we understand how to best take advantage of it in our hospitals. EMR design must never work from an entirely blue-sky approach. Though it is helpful to understand the best practice process for medication reconciliation in the abstract, it is equally important to understand the realities of unit dose packaging in our hospital and the limitations of our IV pumps.

Enterprise change management is one of the most onerous and difficult tasks in an EMR implementation. Most organizations cannot break years of care patterns through a new computer screen overnight. It is difficult to help clinicians accept the changes if they're radically different from the processes of today, and are fraught with operational challenges. One nurse told us that she understands the importance of documenting at the bedside, but if that means that every time she performs an intervention that she must wheel a heavy PC on a cart into the room, she will just document when she gets back to the nursing station instead.

Formalize design oversight

One final recommendation is to make the last design decision first: who will ultimately approve the EMR design. This group should optimally gain input from real users who understand operations and the technology; however, the group itself should consist of organizational leaders with enough vision to understand the potential of the EMR and enough authority to address the practical realities of current clinical operations. The proper EMR design will not guarantee success, but a poor EMR design is difficult to overcome.