Using Six Sigma principles, one health care system has significantly cut ventilator days

As hospitals and health systems around the country strive to become more efficient and obtain higher levels of patient satisfaction, more and more are turning to a comprehensive methodology called Six Sigma1 to facilitate those changes more quickly.

Different from traditional experimentation in academic settings and used primarily in the industrial and manufacturing industries, Six Sigma is a powerful tool that helps companies analyze data and statistics to improve operational performance. Six Sigma is a disciplined, data-driven approach to process improvement aimed at the near-elimination of defects from every product, process, and transaction. Using Six Sigma, companies can take practical problems, develop them into statistical problems, find statistical solutions, and then implement those into practical, sustainable improvements.

At BayCare Health System, a not-for-profit health system with nine hospitals in the Tampa Bay area, Six Sigma has provided noticeable improvements. BayCare launched Six Sigma approximately 1 year ago to help improve patient care and reduce costs. The edge that Six Sigma has over other quality initiatives and the distinction that the methodology provides is the following:

• instills a common language and shared techniques

• demands and provides a rigorous control mechanism

• requires leadership support while driven by those closest to the process

• targets variability instead of aiming at averages

One of BayCare’s first 12 Six Sigma projects was to reduce the number of days intensive care unit patients remained on ventilators at a hospital. Reducing the number of ventilator days for ICU patients decreases their risk of ventilator-associated pneumonia and other infections and shortens their length of stay. Any reduction of time on a ventilator improves the clinical outcomes of critically ill patients.

The results were very positive. In less than 6 months, the hospital used Six Sigma to cut its ventilator days by more than half. Previously, the hospital was above the health system’s average for ventilator days. Now, it has the best monthly average throughout the health system in this clinical area. Not only does the reduction in ventilator days translate to fewer complications and shorter length of stays, which promise to improve the quality of life for patients, it also improves the hospital’s bottom line.

The hospital achieved these results by using Six Sigma’s DMAIC methodology. DMAIC means define, measure, analyze, improve, and control. Each phase plays an important role in achieving a practical solution. Here is more on each phase:

Define Phase
Launching and setting the context and objectives for the project

The first step in the Define Phase was to assemble a team that was knowledgeable about the ventilator process. This team included members of senior management and clinical representatives from seven hospital departments: nursing, respiratory therapy, pharmacy, rehabilitation/physical therapy, case management/social work, infection control, and nutrition. Members of this team utilized root cause analysis tools  to develop and create the problem and objective statement.

Defining the project allowed the team to identify project metrics and narrow the scope of the project. Upon completion of this phase, an important deliverable is a SIPOC diagram (suppliers, inputs, process, outputs, and customers diagram) that is created (Figure 1).

Measure Phase
Understanding the process and gathering the baseline performance and capability of the project

The first task during this phase was for the multidisciplinary team to document a process flow for a patient from intubation through extubation. This exercise revealed areas that needed improvement with more consistent process flow and standardization, leading to some immediate “lean” techniques, discussed later in detail in the Improve Phase of the project. Based on the process steps, the team used other quality-improvement tools, such as a fishbone diagram, to assist in identifying the potential causes for too many ventilator days.

The team began to see that many of the factors that “caused” too many ventilator days were controllable by the clinicians at the hospital. Data from 427 ventilated patients in 2004 was collected for the project. There were 37 variables identified on these patients. These variables or possible reasons included, but were not limited to, age of patient, attending physician, pulmonologist, principal diagnosis, type of sedation medication, daily arterial blood gases (ABG), daily chest x-rays, nutritional or rehabilitation consult and day of stay this consult occurred, and many other clinical factors.

A capability analysis (how well the process is currently meeting the customer’s needs) showed areas for improvement. The team then looked for best practices to set a benchmark goal. By comparing similar patient populations and weaning protocols, they set a goal of decreasing ventilator days by 30%. This target would serve as the project’s goal. The goal was verified by running an analysis of the means to determine that the change in the number of days was statistically significant and outside of the confidence intervals.

Analyze Phase
Using data and tools to understand the cause-and-effect relationships in the process or system

With a baseline sample of data to work with, the team entered the Analyze Phase ready to identify key inputs that affect the number of days a person is on a ventilator. Due to the quantity of variables, the Six Sigma team leader, known as a Black Belt, divided them into three areas.

The first set of information consisted of patient-centered or demographic information. These variables included, but were not limited to, age, race, physician, pulmonologist, primary diagnosis, admit source, and total length of stay.

The second set of variables included items that occurred to the patient while in our care, such as type of sedation medication, type of pain medication, whether a sedation vacation was performed, nutritional consult, rehabilitation consult, daily ABG, and daily chest x-ray.

The last set of variables were associated with post-extubation: items that did not fall into the prior two categories—discharge disposition, temporary or permanent tracheotomy, and reintubation rate.

Each of these sets of variables was run through a series of analytical tools for both continuous and discrete data. These tests allowed the Black Belt to narrow focus on which inputs had the most impact on the number of ventilator days. Once the analysis was complete and a statistical solution was evident, the team moved to the Improve Phase.

Improve Phase
Determining, validating, and implementing solutions to achieve the objective statement

Based on the statistical solution, the team began thinking of ways to implement practical solutions. They focused on a five-prong approach for these improvements. The improvements are outlined below (Figure 2):

1. Standardization

a. Process flow—Input on what flow or pathway of care would be ideal for a ventilated patient was gathered from team members and physicians.
b. Weaning criteria—Specific clinical outputs that are necessary for a physician to determine if a patient is ready to wean off a ventilator were collected.
c. Daily rounding sheet—The specific clinical indicators for weaning were accumulated and a standard sheet was created. This sheet identifies which department is accountable for the specific criteria. Additionally, goals are set for the patient each day.

2. Hospital-driven improvements

a. Interdisciplinary rounding—Seven departments (nursing, respiratory therapy, infection control, pharmacy, case management, rehabilitation, and nutrition) round at 8:00 am every morning on each ventilated patient utilizing the standardized form.
b. Daily sedation vacation—Nursing is now required (unless medically not safe for the patient) to perform a sedation vacation on every ventilated patient during the night shift to adequately assess if the patient meets weaning criteria and document on the standard form.
c. Ventilator-associated pneumonia (VAP) bundle administered—As part of the 100K lives Campaign,2 through the Institute of Healthcare Improvement, implementation of the VAP bundle is now documented.

3. Physician-driven improvements

a. Updated weaning protocol—Team members edited and reworked the existing protocol by adding updated techniques.
b. Pulmonologist acceptance—ICU medical director and respiratory therapy director are rolling out the protocol for medical staff acceptance.
c. Regained confidence in clinical staff—By using the rounding sheet, the clinicians are prepared and organized to answer physician questions concerning their patient’s status on the ventilator; additionally, special attention by seven departments rounding on their patients fosters improved confidence.

4. Information system support and documentation

a. New critical care orders—Current orders have been reformatted, and specific clinical measures found from data to be important in the reduction of vent days have been added:

• ABG on admission and daily x 8 days; PRN
• Daily chest x-ray
• Daily sedation vacation
• Nutritional, rehabilitation, and case management consult within 24 hours
• Daily weaning assessment

b. Daily departmental notification of ventilated patients—Information systems (IS) automatically prints a list of ventilated patients via departmental printers at 5:45 am.
c. Weekly dashboard—This is a standing IS report to show average ventilator days.
d. Policies and procedures to support new processes—Supporting policies and procedures to hardwire new process changes have been put into place.
e. Team map—Each of the seven rounding departments will track key indicators (yardstick of success). These seven measures will make up a “ventilator team map” so process owners can monitor project success and accountability.

5. Team member education and training

a. RN in-service is conducting ongoing education on topics such as pain management
b. Ventilator-specific annual competencies are being done for the RT department
c. Interdisciplinary “Lunch and Learn”—These are monthly educational opportunities for the seven rounding departments (and open to entire hospital staff) to learn about each other’s departments and the specific clinical terms related to the vent patient. Lunch and Learn is led by key members of the rounding team.

Control Phase
Establishing plans and procedures to ensure the improvements are sustained

The most important phase in this methodology—and what sets Six Sigma apart from other quality initiatives—is the Control Phase. During the Control Phase, the team focuses on putting processes in place to fully monitor the input improvements that have been implemented. This phase focuses on the key “x’s” or variables identified in the Analyze Phase. If these components are under control or within the specifications set by the team, the output or number of ventilator days should, in theory, always stay close to or under our target goal. All seven departments signed off on this control system (Figure 3).

The five-prong approach mentioned in the Improve Phase was piloted in July 2005. Results from this pilot have shown steadily decreasing ventilator days; this decrease has continued throughout realization. Since the second month, average ventilator days continue to be below our target goal. There has been a 64% decrease in ventilator days since the inception of the project with a financial realization of $554,390 through December 31, 2005.  

Angi Jennings, who has a master’s degree in administration, is a Six Sigma Black Belt with 15 years of experience in health care quality improvement, and is a member of the newly created Six Sigma Department at BayCare Health System, Tampa.

1. Six Sigma. What is Six Sigma? Available at:

2. Institute for Healthcare Improvement. 100K lives Campaign. Available at: Accessed January 24, 2006.