Six Sigma at Academic Medical Hospital

Academic Medical Hospital and Six Sigma

AMH was a teaching hospital associated with the medical school of a state university located in a small rural community. As the largest employer in the area, AMH was also an important contributor to the community, and, although the hospital had a positive reputation in the community, life inside the hospital was not without turmoil and problems.

The portal of entry for a majority of AMH patients was the ED where AMH Chief of Staff Gerry Elbridge had struggled for 10 years to streamline processes to reduce ED patient wait time. Although long wait times seemed to be readily excused by many ED physicians owing to the complexity of emergency room processes, Dr. Elbridge was frustrated with the situation: “This is absolutely unacceptable. Patients with potentially life-threatening injuries or illnesses are waiting for over an hour for treatment. And that’s far too long,” he said.

Over the years, streamlining emergency care processes to reduce patient wait time had become Dr. Elbridge’s obsession. His first efforts to improve the situation utilized a 10-Step Quality Assurance Program, but he had encountered resistance from doctors who were unwilling to change their habits and routines based on long experience, believing that their emergency care processes ensured the best ED medical treatment.

In the 1990s, the quality programs at AMH had been geared toward finding out what kind of service patients wanted, but often these programs were not supported by the facts-based, root-cause-analysis methodologies that generate improvement. In 2000, Dr. Elbridge was invited to an introductory seminar with

U.S. corporate executives and introduced to Six Sigma. From that introduction, he became convinced that Six Sigma had the statistical rigor and change-management elements lacking in past programs tried at AMH. Dr. Elbridge spearheaded the introduction of Six Sigma at AMH and chaired a kick-off session that was attended by 80 of the university’s top administrators, executives, and medical staff members from both AMH and the medical school.

These leaders were uniformly enthusiastic about the Six Sigma approach, and they developed a preliminary draft for implementing Six Sigma at AMH, picking talented AMH staff members who were not physicians to serve as Six Sigma Black Belts. Following an introductory training session in May 2001, each phase of the rigorous Six Sigma methodology was taught to the selected Black Belts and Green Belts, who were trained in a subset of the Six Sigma methodology and usually worked with projects having a more limited scope. The leadership at AMH and the medical school had selected six training projects as the first wave of improvement activities: ED Wait Time, Discharge Cycle Time, Operating Room Case Carts Accuracy, MRI Turnaround Time, Orthopedic Clinic Access, and Physician Coding. AMH intended to add more Six Sigma projects as the process became the cornerstone of a cultural change at AMH. Six Sigma Foundations Teams were established, and each undertook an application practicum on an assigned AMH project with corporate trainers acting as coaches.

Dr. Elbridge expressed the consensus of the senior executive leadership group at AMH:

Six Sigma methodology appears to have all the elements of a very comprehensive and systematic approach. It has the foundations of statistics to support it, as well as the structure and oversight necessary for us to not only understand and improve our processes but also to achieve long-term gains by focusing on the acceptance of change. That is where we have historically dropped the ball in the past.

This leadership group agreed on the following objectives for instituting Six Sigma at AMH:

• Creating a safer environment for patients and staff by reducing medical errors, and improving patient and staff satisfaction;
• Becoming the Best Place to Care for health care professionals demonstrated by increased retention and reduced turnover;
• Becoming the preferred place to receive care demonstrated by increased market share both regionally and nationally;
• Continuing to reduce variability and waste in processes.

Understanding Six Sigma Roles and Responsibilities for the ED Wait Time Project

There were critical roles and responsibilities that ensured Six Sigma methodologies became ingrained in everyday work. These roles applied to the entire Six Sigma process—project selection to full implementation and assessment of lessons learned. Each improvement project was organized with the same structure. In addition to strict adherence to this structure, it was vitally important that an organization have unwavering support and commitment from the executive leadership when implementing Six Sigma. (See Appendix 1.)

A Six Sigma program, Champion, created the vision for the Six Sigma initiative, defined the strategic goals and metrics of the organization, established business targets, and created an environment within the organization that promoted the use of Six Sigma methodology and tools. As program Champion, Dr. Elbridge’s support for Six Sigma was unwavering, and his influence as a physician leader helped open the door to communication with the medical school community. But the segregated reporting relationships between the AMH and the medical school clinical departments lessened his authority with attending physicians2 and residents3 working in AMH. And, like all pioneers of Six Sigma implementation, Dr. Elbridge was unaware of some of the pitfalls related to rolling out the program throughout the entire institution.

The project Sponsor was a local leader, who was generally the requestor of the project and committed to its solution(s). The Sponsor identified project goals, objectives, and scope; removed barriers and aligned resources; served as a key communicator of project progress and status; and approved final recommendations. But for each project in the first wave of Six Sigma projects at AMH, the Sponsor was not the project requestor. In the case of the ED Wait Time Project, Dr. Terry Hamilton was recruited as the project Sponsor. Dr. Hamilton, a busy vice-chair of the medical school’s Emergency Medicine department, was active in several other departmental initiatives and responsibilities, as well as an involved father and husband. Dr. Hamilton had been solicited by Dr. Elbridge to serve as the project Sponsor of the ED Wait Time Project, but Dr. Hamilton was reluctant from the onset.

The project Owner was a local operations leader who, as a member of the Foundations Team, participated in implementation and was accountable for sustaining long-term gains. Nancy Jenkins, a long-time ED Patient Care manager, was selected as project Owner. Clinically, she was an exceptionally competent care provider and performed well as an operations manager but the implementation of significant process and behavior changes was not her strength.

2 An attending physician or an attending is a physician who oversees and provides emergency medical care in the AMH, as well as supervising resident physicians. Emergency Medicine Attendings are faculty members of the Medical School’s Emergency Medicine Department and contracted employees of the hospital assigned to the ED. Their three primary responsibilities include education, research, and patient care. Attendings report to the Chair of Emergency Medicine who reports to the Medical School Dean.

3 A resident is a physician in training who rotates through various departments during their medical education. ED residents provide primary patient care in the ED under the supervision of Emergency Medicine Attendings.

Black Belts were full-time project leaders who were experts in Six Sigma methodology and statistical tool applications. For example, Jane McCrea had a 25-year career history of nursing leadership roles and had significant work experience in the industry. Project management, quantitative analysis, team facilitation, and building relationships with clinical staffs were her strengths. Like Dr. Elbridge, she was a great proponent of Six Sigma.

Green Belts were trained in a subset of Six Sigma methodology and usually worked with projects having limited scope, generally in their respective work area. They also partnered on a project led by a Black Belt, as was the case on the ED Wait Time Project. Dr. James Wilson, whose medical practice involved a high degree of interaction with the ED, served as the project’s Green Belt. An AMH physician leader in clinical process improvement and a faculty member of the medical school’s Surgery department, Dr. Wilson was a natural choice. His own clinical commitments in surgery at AMH limited his availability for a high level of project involvement. Still, he consistently interacted with the ED physician as needed per McCrea’s request. Dr. Wilson was also a long-time, personal friend of Dr. Hamilton.

The Foundations Team was a group of local experts who participated in project communication/education, problem analysis, and improvement design and implementation. Originally, seven team members from both the ED management and clinical staff ranks were selected and trained. Two dropped out owing to scheduling constraints and interest levels, however, and one resigned from AMH. The four remaining team members included Nancy Jenkins, Patient Care manager:4 Georgia Williams, ED Registration Manager,5 Bill Barber, Senior Clinician,6 and Steve Small, Senior Clinician and Quality Improvement Coordinator.7 All these individuals had a strong commitment to quality and performance improvement, but their previous efforts to effect change in the ED environment had not been successful.

Applying Six Sigma DMAIC Methodology to the ED at AMH

The Six Sigma methodology was a project-focused, step-by-step approach to improvement realization. There were five phases in the DMAIC methodology: Define, Measure, Analyze, Improve, and Control. (See Appendix 2.)

Six Sigma projects were defined from a business, operational, and customer-needs basis. Often, as was the case with AMH, projects were selected and defined by their link to an organization’s strategic plan. Dr. Elbridge and the Foundations Team narrowed the Wait Time Project to a manageable scope by limiting the study to Priority Level II walk-in and wheelchair patients. They defined the start and end points of wait time as being from the initiation of triage (preliminary patient care prioritization) until the first interaction with an ED resident and/or an attending. The total wait time was referred to as the “Y” or “what you are measuring.”

4 A Patient Care Manager is responsible for both business management as well as clinical care management of nursing and support staff members. The person in this function managed the department’s operational/staffing and capital budgets. The person oversaw daily operational management of the ED and was responsible for maintaining internal and external (regulatory) patient care standards. All ED staff, with the exception of attendings and residents reported to the Patient Care Manager.

5 The ED Registration Manager is responsible for all registration staff and ED patient admission and discharge documentation related to demographics, insurance, and coding information. Registration staff members also perform a variety of clerical support functions in the ED.

6 A Senior Clinician is a senior registered nurse who has several years of experience in the ED, serves as a preceptor/trainer for other nurses, assumes lead positions within the department (i.e., charge nurse), participates on various committees and tasks forces to enhance patient care services, and is viewed as an expert resource person for nurses, support staff, and physician colleagues.

7 A Quality Improvement Coordinator is a senior registered nurse who has the expanded role responsibility of overseeing quality monitoring for the ED based on established internal and external performance expectations. This clinician produces and analyses quality data/reports and contributes to staff and faculty education. The QI Coordinator also participates in and/or leads improvement initiatives.

In the Define phase, the team identified such expected benefits of the ED Wait Time Project as expedited emergency medical care delivery, improved patient satisfaction, reduced patient complaints, increased patient capacity, and improved operational efficiency. Moreover, the potential existed for operating margin gains for the ED via operational efficiencies as well as patient census growth that could be fueled from achieving and sustaining the expected benefits.

The next phase was Measure, in which the team determined the baseline measures and target performance of the ED process, defined the input/output variables of the process, and validated measurement systems. The process flow for patients arriving at the ED as a walk-in or a wheelchair patient is described in Exhibit 1.

The Foundations Team collected the initial baseline study data for a thirty-patient sample while simultaneously conducting a patient-satisfaction survey. The survey determined that most patients were either very satisfied or very dissatisfied, thus confirming that there was definitely a need for improvement. Based on this patient survey, as well as clinician input and external benchmark data, acceptable maximum wait times were established for each of the seven process steps (from beginning of Patient Triage to start of Physician Assessment as identified in Exhibit 1). The team also established an upper specification limit (USL) of 37.1 minutes for the Y—the total wait time for a patient at the ED (the sum of the maximum wait time for each of the seven process steps). Any occurrence of total ED wait time greater than 37.1 minutes was considered a quality defect. Acceptable (upper limit value) Lobby wait time was set at 15 minutes and acceptable (upper limit value) medical doctor (MD) wait time was set at 8 minutes.

The baseline study statistical mean for total wait time incurred in the seven process steps was 62.5 minutes, with a standard deviation of 39.66 minutes. McCrea reported to Dr. Elbridge, “The team counted 210 opportunities for a defect to occur within the seven process steps and found that the ED was operating at a 38.6% ‘defects per opportunity’ performance.” And Elbridge told the Foundations Team: “The data doesn’t lie. This is a no-brainer. We can do better.”

The team analyzed the baseline study data to statistically determine that 98.8% of the variation in wait time was related to two of the seven steps of the process: Lobby Wait Time (time from postregistration to being transferred to an exam room) and MD Wait Time (time from completion of nursing assessment in the examination room to start of physician interaction). This focused the team’s effort solely on the most influential steps of the process.

When McCrea showed the ED attending physicians the results of the baseline study data analyses, reactions were mixed and somewhat uninspiring to the team (see Lobby Wait Time and MD Wait Time: Study 1 in Exhibit 2). The primary concern was the small sample size of thirty patients. As a result, the team decided to make a larger, more detailed data collection effort. This second phase of data collection coincided with the Analyze phase. The Analyze phase used raw study data to establish the key process inputs that affected the process output Y. The Foundations Team used a Fishbone Diagram to help it brainstorm and display critical process factors (referred to as Xs) that were thought to be possible drivers of excessive wait times. Twenty-three X causal factors were selected for further study in the second round of data collection (Exhibit 3). This second baseline survey, with a more satisfactory sample size of 129, validated the original baseline results (see Lobby Wait Time and MD Wait Time Study 2 in Exhibit 2) and helped the team statistically identify the likely root causes of excessive wait time.

The team then began the Improve phase of its project. During this phase of the project, improvements were identified to optimize the outputs and eliminate/reduce defects and variation. The Foundations Team selected critical process variables (Xs) that were controllable and determined the statistical relationships to the Y performance measure.

The ED Wait Time Foundations Team designed recommendations to improve the ED processes, and it conducted a pilot study comprised of 172 patients over a period of thirty days for 10 hours per day during the busiest times of the day and for the specified patient population (Exhibit 4). The three general categories of the ED process changes included:

1. Patient Flow—Changes in how Priority Level II patients were moved through the system, which resulted in most patients by-passing the waiting room and instead proceeding directly to an examination room following registration.
2. Care Team Communication—Changes to patient care team communications through modifications of assignments by ED zones, and utilization of new communication boards.
3. Streamlined Order Entry and Results Retrieval Process—Central clerk support for entering orders and retrieving results of all patient medical testing that was done. This change reduced the amount of time physicians and nurses were occupied on computer terminals and thus unavailable to provide patient care.

Project results from Study 1, Study 2, and the pilot are shown in Exhibit 2 for the Lobby Wait Time and MD Wait Time. Now beginning the Control phase, the team assessed the success of the pby comparing the before and after distributions of performance. The statistical analyses performed by McCrea revealed the pilot Lobby Wait Time mean value came in under the acceptable target of 15 minutes, and the MD Wait Time measure was improving (pilot mean was 8.9 minutes against a Study 1 baseline of 16.1 minutes and a Study 2 baseline of 11.2 minutes). The MD Wait Time acceptable target had been set at 8.0 minutes.

McCrea felt the pilot study had a few statistical limitations that the team needed to discuss as it reviewed the comparison results. She prepared the Mood’s Median statistical significance tests and analyses summary for the team, described in Exhibit 5, which produced the following outcomes:

• Pilot Lobby Wait Times were better than the established 15 minute target, the defect rate dropped, and the 95% confidence interval test on the Study 1 median and Study 2 median versus the pilot median validated statistical significance of the improvement in wait times.
• Results for MD Wait Times were statistically significant in one of two Mood’s Median tests (Study 1 versus pilot). Positive trending was demonstrated in the comparison of Study 2 to the pilot.

It was noted by McCrea that the Hawthorne Effect came into play, particularly in the Study 2 data collection activities. Owing to unreliable automated data-gathering procedures in the ED’s information systems and the need to use intradepartmental manual data surveyors, the presence of bias was recognized. With the pilot study nearing completion, the team needed to decide whether to redo some aspects of its work in the Improve phase.

In the Six Sigma Control phase, the team ensured the new standard operating procedures for moving patients through the ED were maintained. The team next designed metric dashboards and accountability reporting to track and monitor progress. The building blocks for the Control phase were (a) setting guidelines and assigning responsibility for the implementation of such guidelines; (b) instituting data review, accountability and reporting techniques to ensure that the new standard operating procedures are tracked and monitored, and (c) developing methods of communication and recognition for updating the staff and executives on the progress. This was an aspect of improvement initiatives where AMH had been historically weak, and successful execution of the Control phase required much attention and effort from the team.

Change Acceleration Process: Creating Change through Building Relationships

The equation, Q × A = E, was the cornerstone of successful Six Sigma improvement implementation. It inferred that the Quality of process solutions multiplied by the Acceptance level of stakeholders was equal to the Effectiveness of results. The key was that the strength of the A influences the power of Q and vice versa. The Q came from the team’s problem analysis skills, design creativity and the statistical rigor of the Six Sigma tools. But no matter which solutions resulted from the analysis, even if the absolute answer to a problem was identified, the potential for success was limited without the Acceptance of the people affected and involved (those people who need to “own” the changes). For example, if the Q was rated as 10 out of 10, but Acceptance level was only 2, the Effectiveness rating would be 20. But if a less than optimal solution was recommended (Q = 6), but Acceptance level was 6, the Effectiveness rating would be an improved 36. Thus a less effective solution option would have a better opportunity for successful implementation if it was more widely and vigorously accepted by stakeholders then the best solution alternative.

Acceptance was the component that previous quality programs at AMH had lacked and an aspect of Six Sigma that Dr. Elbridge anticipated would have the greatest impact on implementing change and discovering additional improvements. Exerting influence over physicians, who are typically the teachers of residents and the medical community “influencers,” however, was not going to be easy. This approach represented a significant culture change for AMH and other medical centers.

To address the Acceptance factor, Six Sigma tools and techniques utilized were collectively referred to as The Change Acceleration Process or CAP. Rather than one individual process, CAP was a tool kit for preparing the corporate culture to accept change. CAP employs a variety of methods that helps the organization understand what needs to be changed, how, and why. CAP dovetails with Six Sigma’s five statistical phases (DMAIC) throughout the lifespan of a project. Different tools were used at different stages in order to measure, monitor, and influence Acceptance. CAP was always, ultimately, found to be necessary to secure Acceptance of implemented changes. Most importantly, these tools assisted teams in clearly identifying key stakeholders (both in the formal and informal networks); understanding each stakeholder’s level of Acceptance and how they could be influenced; and understanding stakeholder concerns and needs and how these can be incorporated into the implementation plan. If the statistical methodology of Six Sigma problem analysis was the bolt, then CAP was the nut that held it in place throughout the project’s duration.

Change happens not solely because data statistically validates via a pilot what should be done, but because people as a whole commit to making changes. Listening and responding to stakeholders’ needs, fears, desires, and concerns about change improved the Acceptance level and consequently raised the Effectiveness rating.

Several CAP techniques were used throughout the ED Wait Time Project:

• Process Mapping: Helped the team consider all elements of the work flow and information of the process, including activities, decisions and measurement points and setting clear boundaries of where the overall process begins and ends.
• Elevator Speech: A two-minute pitch for a change that links the need for the change with the vision of the future and was used whenever a Foundations Team member had an opportunity to sell the project to a key constituent.
• Extensive and Creative Communications Plans: These were used throughout the project to provide progress reports, status updates, and feedback opportunities for staff.

.

Page 8 UV0339

• The Stakeholder/Resistance Analysis: This tool was used to identify key stakeholders and analyze how they felt about the change initiative. Actions were usually taken to gain strong commitment from key constituents and eliminate or lessen resistance from the more resistant stakeholders.

Managing Doctors Is Not Easy

Getting people to embrace Six Sigma reforms in the ED was going to be an uphill battle. Part of the problem was the nature of the medical profession as a whole. “Influencing the ED physicians will be our biggest challenge,” Dr. Elbridge, the program Champion, remarked in his first Six Sigma meeting with McCrea. He emphasized the challenge presented by the duality of the reporting structure, with doctors holding academic positions in the medical school and ED clinical appointments in the AMH. Along with their academic professional activities, they had regularly scheduled rotational assignments as attending physicians in the ED. Elbridge explained:

There’s no doubt that our ED physicians want to do the right thing, which is to see emergency patients as quickly as possible, but they need to be the source of the solutions. The beauty of Six Sigma is that’s exactly how it works. If we can get them to participate in the activities, they will find the answers within the framework of Six Sigma methodology. I believe this improvement model presents us with the best opportunity for success that we’ve ever had. The fact that physicians are not hospital employees, but independent contractors of a sort from the Medical School, results in a complex reporting relationship where even I, as the AMH Chief of Staff, cannot make them comply with hospital programs. In fact, there is almost a disincentive to participate since their incentives lie with research, education, and patient care specialty. Some will say, “Ah, but this project is patient care.” And the end results can positively contribute not only to improved patient care outcomes, but also smoother workflow for ED physicians and staff once processes are streamlined and made more efficient. Not to mention the impact that satisfied patients and family members can have on your day as a physician. That’s the angle we have to present.

Jane McCrea remained committed and positive about the team’s work and Six Sigma, and she always was inspired after talking to Dr. Elbridge, but she still had to deal with Dr. Hamilton, who was recruited by Dr. Elbridge as the project’s Sponsor. Dr. Hamilton was hesitant to get involved. Even setting up a first informational meeting with Dr. Hamilton became a challenge for McCrea. She recalled, “Dr. Hamilton seems to be looki