PA hospital says results of CPOE reduce drug errors

A Pennsylvania hospital reports that it is seeing tremendous results from a computerized order entry system (CPOE), with an 85.7% decrease in improper doses and a 76.8% decrease in medication omissions.

To make the results even more satisfying, the new system was introduced with almost complete support from the hospital's physicians.

The CPOE system was introduced from day one as a risk management project, not an IT project, says Joel Diamond, MD, chief medical information officer at UPMC St. Margaret, a community hospital within the University of Pittsburgh Medical Center umbrella. The hospital worked closely with risk management to introduce CPOE as part of an overall quality improvement initiative at the hospital, he says.

The difference in that approach is that the CPOE plan was introduced to physicians and promoted as a tool for improving quality and patient safety, as opposed to announcing it as a new technological improvement "and now you have to enter your orders electronically," Diamond explains.

St. Margaret used the Millennium system, a CPOE product from Cerner in Kansas City, MO.

Physicians supportive from the start

The hospital's "go live" date for the CPOE system, the date on which the system would be fully functional and everyone was expected to use it, was Sept. 12, 2004. The effort to introduce the plan began about a year earlier.

CPOE was the final piece of a quality improvement effort that included other electronic enhancements enabling the review of lab results and radiology reports on the computer. Despite the inherent skepticism of some physicians who will resist any changes in how they perform daily tasks, the physicians at St. Margaret were almost unanimously on board with the plan two weeks before the go-live date.

"We did a lot of preparation to get people ready for this change, which included countdown posters showing how many days were left before we went live with the CPOE," Diamond says. "Our original goal was to have a pilot program about six weeks before the real go-live date, but the posters were all counting down to the date for the start of the pilot program. We had prepared the staff so well and got them so psyched up that they assumed the countdown date was for the full hospital start-up."

On the day the pilot program was to start, a Sunday, Diamond went to the hospital to help get the system running. Physicians kept coming up to him all day to tell him how excited they were to put in orders electronically.

"Everyone started putting in orders, so we either had a hospitalwide pilot or we jumped right to full participation on the first day," he says. "Either way, it speaks well to the fact that the staff was primed and ready to go."

Focus on quality, not technology

The introduction of some of the other electronic improvements helped pave the way for the CPOE system, says William Fera, MD, medical director of wellness services. Physicians had seen that the other systems were implemented smoothly and yielded good results, so they were receptive to the idea of another system that could improve efficiency and quality, he explains.

To educate the physicians about the CPOE system, St. Margaret sponsored a social gathering off the hospital campus.

"We made it clear that the reason we were introducing CPOE was that it would improve quality at our hospital — not for any other reason," Fera says. "That was sort of our mantra, that CPOE was going to improve the quality of care we give to our patients."

Fera also was up front about the fact that there would be a learning curve and that CPOE might initially slow down order entry as people got used to the new system. But he promised that improvements would come quickly.

"As we went forward, we identified physician champions in each department and did pretraining with them so that they would be familiar with the product when we started training all the house staff," he says. "It worked so well that everyone was ready to go earlier than we thought."

Two weeks after full implementation of the CPOE system, the hospital eliminated all paper forms for order entry, Fera says. There were no complaints by that point.

In the year since the CPOE system was implemented, St. Margaret has seen impressive reductions in the kind of errors that can threaten patient safety. Diamond and Fera offer these examples of how CPOE has improved quality and safety at the hospital:

  • medication omissions: 76.8% decrease;
  • improper dose: 85.7% decrease;
  • extra dose: 50% decrease;
  • unauthorized drug: 46.7% decrease;
  • drug clarification: 79.1% decrease.

Diamond says the results are better than he thought they would be after one year.

"But now, seeing how well it works, I expect that six months from now they will be significantly better even than what we're seeing at this point," he says. "These numbers are absolutely astounding to us, and they fly in the face of study results that show slower results in patient safety."

There's a reason for that, Fera says. St. Margaret's multidisciplinary approach to rolling out the CPOE system resulted in a level of acceptance and confidence that other organizations might not have achieved before going live with the new technology, he says.

"We involved nursing, and pharmacy, and radiology at every step," Fera says. "So when there were questions or problems with processes and work flow, people were still communicating verbally. Everyone knew exactly what was happening, as opposed to it just being physicians who were on board and knew what was going on."

High rate of physician usage helps

Diamond also suspects that other health care providers have not seen such good results with CPOE in such a short time because they did not have as high a rate of physician usage. When St. Margaret officials first began planning the transition to CPOE, they figured that usage by about 40% of the physicians in the first six months was a benchmark for success. Diamond disagreed and pushed for much higher participation.

That high level of participation is why the hospital reaped such striking benefits so quickly, he says. Using a mix of CPOE and paper order entry introduces the opportunity for confusion and other errors, Diamond says.

Fera also notes that St. Margaret's system is true CPOE, with the physicians actually entering the orders themselves. Some facilities using CPOE actually have someone other than the physician entering the order into the system, he notes, and that can only slow down the operation while also introducing the potential for more mistakes.

Two weeks after going live with the CPOE system, St. Margaret's underwent a JCAHO survey, which some people thought would be a disaster. Instead, the survey was a big success, and JCAHO raved about how well the CPOE system improved patient safety.

"This is absolutely a reproducible model," Diamond says. "Others can do this same thing. The keys for success are to have a unified team that introduces this as quality improvement, not IT, and to have multidisciplinary support so that everyone is in this together."