The Key to CPOE

When Summa Health Systems of Akron, Ohio, began considering a computerized physician order entry (CPOE) system a few years ago, there was little question that we would improve patient care, safety and efficiency. We knew, however, that those benefits would only be ours if we could find an implementation approach that would work for our clinical staff at all facilities.

Thanks to careful planning, we now find ourselves a model of CPOE success, with 100 percent physician order entry, acceptance and efficiency that has contributed to a reduced length of stay by 0.2 of a day in less than a year.

Our implementation covered our two campuses at Akron City Hospital and St. Thomas Hospital, where combined staff includes 200 residents and 1,200 physicians. In total, the two facilities comprise approximately 1,000 beds; and handle 35,000 admissions, 21,000 surgeries and 100,000 emergency department visits.

A number of elements made our implementation a success at these two facilities. First, we were fortunate enough to have strong support from leadership, which made everything we accomplished possible. In addition, we chose a rollout strategy that was adapted to our facilities—we developed order sets in advance, involved physicians and nurses in the implementation at many levels, and relied on a blended training and support strategy.

An Implementation Strategy
We worked with First Consulting Group (FCG) to select a system. After a 26-week selection process, we chose an Eclipsys system and began planning the implementation.

One of the first significant questions we faced was whether to roll the system out in a phased approach, or all at once. In the end, our plan actually involved doing both. Besides myself (CMIO) and Pamela Banchy (director of clinical information services) our executive steering committee included the VP of medical affairs, the chief nursing and financial officers, a physician board member at Summa Health Systems, the director of finance, and the directors of pharmacy, radiology and laboratory services. These members, who were chosen for their ability to provide project oversight for care quality and financial outcomes, met with nursing and medical staff leadership.

The committee decided to conduct a phased rollout, beginning with a pilot on a single nursing unit at the Akron City campus, and then phase-in the other units. Once the Akron City campus was fully implemented we would then carry out a “big bang” implementation across the entire St. Thomas facility.

One of the greatest challenges of implementing any CPOE system is ensuring that automating the clinical order entry process is safe, efficient and adopted by the medical and nursing staffs.

One reason we chose the “big bang” approach for our second facility was that there were a number of physicians and specialists who saw patients at both campuses and who would already be familiar with the system. Also, by the time we were ready to roll CPOE out at St. Thomas, with the exception of two locations, we would already have implemented electronic order sets for all clinical departments at Akron City.

We took a clinical approach in selecting a pilot site, looking at the number of beds and the types of services that admitted patients to each unit, as well as patient flow. We wanted to build a strategy that would maintain patient care and ordering efficiency along with medical staff adoption, while causing the least amount of change to the current clinical workflow.

We eventually chose the Akron City acute care of the elderly unit as our pilot site because of its homogenous patient population, consistent orders, and limited number of admitting physicians.

The pilot gave us the opportunity to evaluate the chosen implementation strategy, clinical order sets, clinical utilization and overall success. The pilot also enabled our clinical and information technology teams to evaluate and make recommended changes for more successful implementations in the future. One important lesson we learned was to limit the scope of our implementations. We initially implemented CPOE, the electronic medication administration record (eMar), and nursing documentation all at once. We learned that the degree of change for the clinical staff was more than anticipated and was not providing the desired outcomes. It was collectively decided among clinical leadership, to temporarily suspend the further rollout of the eMar. Collaborative meetings with clinical staff and leadership was done for a period of one month, and recommended changes were implemented prior to progressing to our remaining clinical units.

Physicians Involvement
One of the greatest challenges of implementing any CPOE system is ensuring that automating the clinical order entry process is safe, efficient and adopted by the medical and nursing staffs. To involve them and create acceptance, we formed a physician focus group at the beginning of the process to help provide clinical oversight and recommendations. This gave the medical staff the opportunity to provide input and decisions to help design the order entry process in a way that would work effectively.

But simply making sure that physicians felt invested and involved wasn’t enough. Education and training along with ongoing support was also felt to be important during the transition to a new workflow and clinical system. A combination of training two to four weeks prior to go-live, along with activation support on the units 24 hours a day, was our methodology and approach.

Physicians were supportive of CPOE, and the challenge was the ability for them to secure time from their clinical practice to attend the brief training sessions. Several opportunities and schedules were available, along with weekend and off-hour training. The flexible scheduling provided the medical staff with the time necessary to attend the training. Classroom training sessions worked well for residents. We trained approximately 80 percent in one-on-one sessions. The remaining physicians were taught in groups of eight to 10.

A Major Roll
Nursing staff played a much more important role in training and in our overall implementation than we initially anticipated. Early on, we were challenged by the fact that order sets on paper didn’t look at all like typical CPOE sets, and we solved this in a way that allowed us to involve nurses. We called on the light-duty nurses to help with testing and validation of the clinical order sets. Nearly all of the order sets were made available before go-live.

In addition to planning, the nursing staff took on a more centralized role. During the implementation, we trained a number of nurses from all clinical units as “super users” who could assist other clinicians during rollout. These super users assisted in our 24×7 activation support during our activations for the entire rollout. Since the nurses already knew the staff, our procedures and the facility’s practices and policies in general, we found the method particularly beneficial.

In the end, our training strategy was one of the most significant contributions to our success in gaining user acceptance—and to keeping our project under budget. By using our own nursing staff for training, we greatly reduced the need for outside trainers and consultants. In the phased portion of our rollout, we relied on FCG and Eclipsys for some areas of training and support, but by the time we reached our St. Thomas big-bang rollout, we no longer required outside consultants.

Today, clinicians use CPOE order sets 90 percent of the time, compared to 12 percent when we relied on paper. Another good indictor of our success—sliding scales for insulin were once used 5 percent of the time, but today clinicians take advantage of them 93 percent of the time. We’ve also greatly reduced duplication of medical, laboratory and radiology orders, and we’ve considerably reduced allergic reactions and drug interactions. And, from an administrator’s budgetary point of view, it may be possible to pay for the entire implementation within a year, given the significant reduction the system has made in an average patient’s length of stay.

The overall reductions in times of stay enabled by the CPOE system hold benefits for both patient and hospital. For Summa Health Systems, by lowering a patient’s stay even 0.2 to 0.4 days, we could save an estimated $3 million per year. With significant contribution to patient safety and newly gained efficiencies and savings, we’re proof that careful planning with staff involvement and training that accommodates physician’s schedules are real keys to CPOE success.

November 2007