As Black Belt Angi Jennings learned, applying DMAIC tools in the
classroom is a piece of cake compared to the first time you try to
use the tools to solve an actual problem. On her first Six Sigma
project for BayCare Health System, Jennings was faced with
16,000 data points, and a team that was new to Six Sigma. This
presented an opportunity for everyone to learn experientially –
Jennings learned about managing change, and how to apply ANOVA
and regression analysis to narrow down a huge data set. Her team experienced firsthand the effectiveness of Six Sigma, and how to
use DMAIC tools to affect positive change.

The Challenge

The standard treatment for patients who are unable to breathe on
their own is to sedate them and put them on a ventilator. In a
hospital’s Intensive Care Unit, patients on ventilators are typically
weaned off the device after a few days, when it’s determined that
they are capable of breathing on their own.

“The longer you’re on a ventilator, the less your lungs will work
naturally,” says BayCare Black Belt Angi Jennings. There’s a
“honeymoon” period when it is advisable to be on a ventilator,
she says. After that, however, patients who are not weaned off
ventilation soon enough are in danger of developing any number of
related complications — ventilator-associated pneumonia, internal
infections, oral ulcers, and bed sores or other conditions caused by
prolonged immobility.

Jennings was assigned a DMAIC project for St. Anthony’s Hospital,
one of nine hospitals in the BayCare Health System in Tampa Bay,
Florida. In addition to the facility’s 20-bed ICU, St. Anthony’s has
a 10-bed cardiovascular ICU. The project goal was to reduce the
number of ventilator days in both ICUs by 30 percent. A secondary
goal was to reduce the overall length of stay in the ICU for vented
patients.

In addition to improving patient care, the number one goal at
BayCare, the project’s forecasted savings were $775,000, attributed
to the projected decrease in ICU length of stay and a decrease in
cost per case.

The Process

The project was BayCare’s first clinical Six Sigma project and the
team took advantage of several Six Sigma tools in the Define phase.
In addition to a SIPOC map and a stakeholder analysis, Jennings recalls
that the affinity diagram was a helpful tool for producing a list of the
many clinical variables that affected the process.

In all, eight departments interacted with vented patients on a daily
basis. Thus, process mapping was also important. “When I interviewed
each department separately, they each had their own idea of the way
the process flow was supposed to be,” Jennings says. “I asked each
department, ‘If you could pick an ideal situation for this patient, what
would it be?’ And that’s the way we came to a consensus on the future state.”

A variety of statistical tools, including ANOVA and regression analysis,
helped Jennings pare down the data bit by bit. The analysis revealed
an interesting trend: 96 percent of the process variability could be
attributed to clinical variables. Treatments such as AGBs (Arterial Blood Gases), chest X-rays and “sedation vacations” were not always
performed, or the results were inconsistently reported on charts.

The data also showed a lack of standardized processes and protocols,
and an opportunity to improve communication between the different departments involved. Jennings says there was little incentive for each
area to note what the other areas were doing in the care of the patient.
Thus, the status quo became a collective lack of accountability for
weaning the patient off ventilation sooner rather than later.

Proposing improvements presented a challenge for Jennings, a non-
clinician. “I didn’t ask anyone on the team to do anything that they
didn’t already do,” she says. “I asked them to be accountable, and
to document what they were doing and have that information available
at certain times.”

On the first day, the team took 2.5 hours to complete rounds for only
six patients — not a very timely result. Jennings attributes the difficulty
to a lack of preparation, but without assigning any blame. “When we
went through the rounding sheet and asked very specific questions about specific care elements, people didn’t know the answers because nobody
had asked them those type of questions before.”

The team adapted quickly, however, and within two weeks was seeing
positive results. At that point, Jennings says, “Everybody began to
change their work habits. It was probably one of the best exhibits of
culture change I have ever seen, because people realized they made a difference.”

In addition to the daily rounds, the team agreed on several other improvements, many of which were essential prerequisites to the daily
rounds. These included a standard process flow, a weaning protocol
and a rounding sheet, which was fine-tuned using a DOE. The team
also came up with the idea of interdisciplinary “lunch and learns” to
help establish a more cross-functional view of the process.

The Results

As a result of this project, St. Anthony’s ICUs achieved a 38 percent
reduction in ventilator days, and a 23 percent reduction in length of
stay for vented ICU patients. Since the project was completed, the
process has remained in control and the hospital has realized $650,000
in hard savings (less than initially forecasted due to a decrease in ICU
patient volume).

Perhaps more impressive is that St. Anthony’s recently received a
Beacon Award for excellence in its ICUs. The project also helped
BayCare’s Performance Excellence program get off to a good start.
It demonstrated perfectly how the DMAIC process can reduce a huge
number of inputs to a critical few, and how important data can be to supporting process improvements.

Plus, it proved to be a good training ground for the team. “The most
rewarding thing,” says Jennings, “is hearing clinicians say to me,
‘Participating on this project is one of the best things that’s happened
to me in my career.’”