PASTE plus teams at core of award-winning program: safety initiative identifies four key subprocesses

HealthCare Benchmarks and Quality Improvement, Nov, 2004

Key Points

* PASTE is a customized version of the Six Sigma DMAIC model.

* Best practices should not be taken at face value; you must have your own data to analyze.

* Develop a specific measurement plan for each project to accurately determine level of success.

A medication safety improvement initiative at Heartland Health in St. Joseph, MO, employed a structure centered around PASTE plus teams and four key subprocesses to achieve significant process improvement and earn the Missouri Team Quality Award.

PASTE (problem analysis solution transition and evaluation) is a customized version of the DMAIC (define, measure, analyze, improve, control) model common to Six Sigma projects.

"We use it for all of our quality initiatives," says David B. Lloyd, RN, MBA, service leader for clinical services. "Plus, we brought in additional Six Sigma tools."

The four subprocesses were:

1. ordering the meds (which involved mainly physicians);

2. transcription into the computer systems;

3. dispensing (pharmacy);

4. administration (nursing).

"We used the PASTE plus teams for all but dispensing," says Lloyd. "When we looked at our internal data [on errors], dispensing ranks very low; pharmacy tends to have their act together."

He adds that subprocesses make sense, because you easily can get bogged down in projects with a bigger focus.

The award-winning initiative, which used 2002 as a baseline year, was preceded by several less successful attempts by Lloyd and his colleagues, which in turn all became part of a learning curve.

"In 1997, I was on a med/surg unit, and my boss told me to go out and do a quality project," he recalls.

He decided to look at wrong meds, and "took a few initial stabs," but realized he did not have good internal information. "We could not even get information on how many doses the unit gave every month," Lloyd explains, thus underscoring lesson No. 1.

Subsequently, he followed best practices in the literature from sources such as the Institute for Healthcare Improvement and the Healthcare Advisory Board, and "just went in and implemented, implemented, implemented." Lloyd notes:

"We did not argue with the best practices or measure if they were worthwhile; we just assumed they were good." Lesson No. 2: "Rather than take best practices at face value, we had to have our own data," he explains.

Each subprocess team was distinct, both in composition and goals. There also was another team that looked at the larger process, just tracking and measuring adverse drug events. Even this was a departure from the past, including a switch from a punitive to a nonpunitive report system.

"We streamlined [reporting] from a big, full piece of paper with check boxes to an orange 5" x 7" card that asked for very basic information," Lloyd notes. (See example, below.)

"We also put a dedicated nurse on as our medical safety nurse. All cards get put through her to be tracked and trended and root cause determined, and then put into a sophisticated database we built," he adds.

The order subprocess team was very unusual, since it was physician-driven, Lloyd notes. "We had about 11 physicians--including surgeons, GPs, specialists--and they took it very seriously, providing good suggestions and solutions," he continues.

"One that was very helpful was a revamped order form. It has a table built into it, which prompts you to put in a drug name, dose, route time, and PRN indications, which have really helped a lot," Lloyd adds.

The form ensured all required elements were provided in order, and have driven compliance upward, he reports. "There has been a 78% reduction in the number of adverse drug events related to ordering," LLoyd declares.

The transcription team engendered a huge process change, including a severely streamlined new pharmacy information system.

"In the past, a single order had to be entered into two systems, neither of which talked to each other," Lloyd says.

"This system takes things out of the hands of the secretarial staff and interfaces back to the nursing system, which creates the MAR [medication administration record]," he explains. The results: a 64% reduction in errors.

Administration "did not quite hit our goal," says Lloyd, while noting that his team had set tenfold stretch goals. They did achieve a 34% reduction, however. The No. 1 solutions? "The nursing staff were heavily involved in the report process," he adds.

"They created standard processes, changed training, tools, and references through standardization, which was significant," Lloyd says.

In addition, there was a change in the MAR and labeling on meds, and increased utilization of preprinted order sets.

"The new ones were vetted through nurses and physicians to make sure they were concise and clear," Lloyd explains.

"As we move forward with a complete move to CPOE [computerized physician order entry] in the next two to three years, these order sets will move right into it." At present, there are some 300 already in use, he notes.

The area of dispensing, which had no subprocess team, still saw a 17% improvement due to various internal PI efforts, including streamlining processes and training staff.