Paul M. Silverman, MD, left, and Steve Venable, MBA, MD Anderson Cancer Center, Houston.

A management technique known as Six Sigma has been generating truly incredible productivity results for those Fortune 500 firms that have embraced its data-driven business improvement methodology.

Six Sigma is based on the deceptively simple idea that being the best at whatever you do can have a near miraculous impact on your business and the bottom line. In fact, its promised gains sound almost too good to be true. But adherents who have jumped in with both feet insist they are not overstated.

The stated benefits of striving for Six Sigma perfection include a 20% increase in profit margins within 1 year, a 4% annual gain in market share, a reduction in the cost of producing goods and services to less than 10% of revenues, and a 10,000-fold quality improvement from an average of 35,000 defects per million operations to an all but nonexistent 3.4 defects per million activities.

While many in the business community look at Six Sigma as a road map for survival, health care providers grappling with the realities of managed care and escalating costs have been reluctant to get involved in this rigorous and seemingly arcane improvement initiative. But that is going to change as major medical centers realize the potential gains are too great to be ignored.

One of the first to rise to the challenge is the University of Texas MD Anderson Cancer Center (MDACC) in Houston, whose extensive treatment and diagnostic services have resulted in its being ranked the No. 1 United States hospital for cancer care in the recent survey of America’s Best Hospitals by US News & World Report.

Figure 1. Initially, the entire CT cycle was mapped in a box plot.

The radiology department is championing the business process at MDACC. It has employed Six Sigma to successfully eliminate bottlenecks in its CT operations, gaining through the process a 28% increase in patient throughput. While it is still tweaking the project as well as monitoring it to make sure that the changes it has implemented are sustained, it is now pursuing a second Six Sigma project to reengineer its film library for faster image retrieval and heightened efficiency.

“I think Six Sigma is going to be the wave of the future in diagnostic imaging,” believes Steve Venable, MBA, MDACC’s special projects coordinator for diagnostic imaging. “The rigor of this methodology is probably not appropriate for all issues, but it is just the ticket where you are having basic, fundamental problems and need to understand what is going on. Our CT project has generated efficiency improvements that have given us free capacity equivalent to two additional scanners. If we had had to purchase them, they would have cost us $3 million.”

Paul M. Silverman, MD, left, and Steve Venable review some of the numerous data generated by the Six Sigma process improvement project in CT.

The ability to schedule more CT patients also has had a direct effect on the bottom line, says Paul M. Silverman, MD, radiology’s chief of body imaging. “CT is the highest demand area in the hospital, and body imaging accounts for approximately 51% of diagnostic radiology revenues. Since our per procedure CT costs have remained flat, Six Sigma has made a large contribution to the institution’s financial bottom line.”

A Pressing Need

The need for greater CT capacity had by mid-1998 become critical for MDACC, a state-supported institution where the mission to cure cancer placed a considerable burden on the faculty radiologists to provide clinicians with precise diagnostic information about their oncology patients, many of whom look to MDACC as their last chance for a cure. About 80% of the patient population is outpatient. “Even with six scanners, we could not meet the institutional demands,” says Venable, who joined MDACC as a radiology administrator in 1995. As a result, MDACC had a continuing backlog and some patients had to go on a waiting list. Others were lost to other hospitals and imaging centers.

“The assumption was that the bottleneck was a hardware limitation,” Venable recalls. “The traditional approach to such a problem would have been to buy more hardware. But we were smart enough to realize that was probably not the best way to solve the problem.” An alternative approach-employing Six Sigma methodology-was proposed by William Murphy, Jr, MD, then head of diagnostic imaging, and Christine Capitan, the department administrator. Venable was tapped for the project, which got under way in August 1998. Their first action was to retain a consultant who trained a cross-section of CT personnel-technologists, receptionists, nurses, film librarians, patient transporters-in Six Sigma methodology. The consultant then served as a resource during the various phases of the project. When Silverman joined MDACC in June as chief of body imaging, he enthusiastically joined the effort to bring Six Sigma from industry to medicine. In September, he subsequently was named chairman of the Six Sigma steering committee.

The project had four main goals:

  1. Increase capacity by 10%.
  2. Improve technologist morale and create a more pleasant work environment.
  3. Reduce patients’ wait time, which was averaging 60 minutes from when they arrived until they began being processed.
  4. Eliminate repeat examinations due to technical errors, such as performing the wrong procedure, or losing the films.

These goals all have been met or exceeded. When the project started, the CT section, working two 8-hour shifts, was on average performing 207 CT examinations per day on some 115 patients. Since the Six Sigma changes were put in place on January 1, 2000, the CT section with the same six scanners (two of which had been upgraded during 1999, as previously planned, to multislice capability) has been performing around 300 examinations per day on some 160 patients-a 28% increase in patient throughput and a 45% increase in the number of examinations.

Even with these productivity gains, MDACC still faces a seemingly insatiable demand for CT studies. To meet it, two new multislice scanners were delivered in September 2000, and upgrades have also begun on two older helical scanners. In addition, a nonhelical scanner that was previously used only for angio/interventional procedures is being upgraded to helical status. This will give MDACC nine CT body scanners-five of them multislice systems.

The next challenge for Silverman and Venable was a proposed increase in the number of patients that can be scheduled during 2001 by 53%. The process of ratcheting up from 160 patients per day to a new goal of approximately 240 patients per day began on September 6, 2000, when the first increment of 20 additional patients was budgeted.

“But even with the projected increase, we expect to be saturated in the not too distant future,” Silverman says.

Achieving the budgeted level of patient throughput is dependent on more than just the new and upgraded scanners, notes Venable. Rather, it is contingent on MDACC being able to scale up the entire CT section and its supporting infrastructure.

“The MDACC experience is that you usually have bottlenecks in areas other than the scanners,” he says. “We cannot increase any volumes until we change our printing technology, and ensure that we have enough nursing staff to prep the patients, film librarians to match and hang studies, radiologists to do the interpretations, and transcriptionists to type the reports. In addition to more staff, scheduling more patients also means, among other things, that you need more film printers, more interpretation areas, additional bathrooms, and more chairs in an expanded waiting room.

“Six Sigma looked at and addressed all of these areas,” Venable says. “The throughput increases we have achieved come from scaling the entire operation, from patient induction through examination interpretation. We never looked at gaining capacity by only adding scanners.”

The Five Basic Processes

Figure 2. Front-end (patient) cycle box plot enabled MDACC to identify those steps with the largest variation as the opportunities for the greatest improvement.

Six Sigma methodology employs five basic processes-defining, measuring, analyzing, improving, and controlling. “The first 6 or 7 months were spent defining and measuring-asking questions about how the CT unit really operated, and getting quantifiable answers,” Venable says. “Because Six Sigma is data driven, you need hard facts. You cannot make decisions based on perceptions or innuendo. Defining and measuring proved to be a very complex and difficult task for us, but it enabled us to determine what areas we should tackle first, and how to allocate our resources.”

One of the main reasons data gathering proved so troublesome was that input from technologists was critical to understanding work flow, and MDACC was experiencing a 50% turnover rate for technologists. It was determined that the MDACC pay structure was not competitive in the Houston market. Although it was only about 5% below those of other institutions, this salary gap necessitated ongoing recruiting and required the use of expensive temporary help. This made the overall cost of filling these jobs far higher than what other institutions were spending. To stabilize the technologist work force and attract the best available personnel, MDACC made the decision to henceforth be a leader in compensation.

Other moves to make technologists happier involved standardizing procedures to assure consistency, and promoting teamwork. “To get them working together, we had to break up the cliques that had developed over time and were a hindrance to cooperation,” Venable says. “We were able to implement some changes and make those cliques go away.” Technologist education has been a key priority of Silverman, working in conjunction with a corporate liason.

In the analysis phase of Six Sigma, it was determined that patient handling and not scanning speed was the primary problem. “We looked at the entire process, from the moment a patient arrived and checked in until we had the films read and the radiologist’s report transcribed,” Venable says. “New multislice scanners provide the ability to scan a patient from the top of the head to the tip of the toes in 90 seconds. But it can take 10 minutes to get the patient on and off the table. So the number of patients we could handle, not the number of examinations we could perform, became the limiting factor.”

“Every CT scan is individually prescribed by a radiologist,” Silverman notes. “So every patient’s record and chart are reviewed by a radiologist, who is also assigned as a teacher for residents and fellows, and additionally consults with referring clinicians. The teacher then protocols the way the study will be performed. We found that if we had all the information that radiologists needed to make decisions at hand when the cases were presented to them, it cut down the amount of time needed to get those orders written, which reduced the patient wait time. One teacher comes in at 7 am and a second comes in at 12:30 PM, which allows for the protocoling of up to 300 CT studies we have been doing on a typical day. Patients who used to wait for an hour before their paperwork was initiated now wait only an average of four and a half minutes.”

The next process improvement for Six Sigma methodology at MDACC is the reengineering of the film library.

Another change in the way CT examinations are done at MDACC has been to utilize nursing assistants to help technologists move and position patients. This enables them to do other patient-related technological activities. “When we started this project, the average examination time-from the beginning to the end-was 23 minutes,” Silverman says. “Today it is about 17 minutes.”

Silverman’s most worrisome problem for the immediate future is recruiting the necessary faculty to read all of the images nine systems can produce. They must be specialists in cancer diagnosis, and they must also be willing to accept salaries well below private practice remuneration, in return for which they will have the opportunity to spend up to 30% of their time on research.

Last June he recruited two additional radiologists for the body imaging section-which reads MRI and ultrasound scans as well as CT-and he hopes to sign up four or five more by year end. This will give him 22 faculty members to start 2001. “Recruiting has become extremely difficult because of all of the opportunities in private practice,” he notes. “The pool of available people willing to sacrifice salary has been shrinking. Nevertheless, we have been able to attract the higher-quality faculty because of the defined mission, as well as the exceptional standard of practice, the unique case load, and the incredible work environment under the leadership of our president, John Mendelsohn, MD.”

Professional changes

Silverman reports that Six Sigma also has resulted in changes designed to maximize the time of the professionals. “We want our radiologists to read films, not be transcriptionists or film hangers,” Silverman says. New technology has enabled radiologists with computer consoles at their workstations to access prior radiologic reports of related studies as soon as the current study is bar coded. Also there is immediate accessibility to the patient’s clinical record and pathology results. “Future technological add-ons are soon to come,” Silverman says.

Figure 4. The benefit of Six Sigma is that it facilitates the ability to common size each process step. Six Sigma is the Z value calculation that equals the standard deviation of 6.

One time-saving innovation has been to utilize facilitators-one in the morning and one in the afternoon-at the CT desk. These people free up the radiologists by handling incoming phone calls and returning calls to clinicians who need answers.

Another time-saving innovation has been to relieve radiologists of the need to hang films by assigning additional people to the film library staff who match just-completed studies with the most recent previous study. Then they hang the films on multi-panel alternators so the two studies can be compared. When the radiologist has completed his or her interpretation, the alternator is then rehung with a new case.

Figure 5. CT patient wait time is plotted in minutes. UCL is the upper control limit, S bar is the sample standard deviation, and LCL is the lower control line.

One reason for hard copy interpretations is that MDACC gets a large number of outside CT films brought in by referral patients. “MDACC is on the threshold of a unique intranet solution to providing digital CT images to clinicians in the primary medical center building and in a separate office building across the street so they can be reviewed by physicians office on standard personal computers. The radiology department is also planning for a conversion to PACS but is proceeding carefully since the large number of cases require that the system be extremely robust,” Silverman explains.

Transcriptionists, who have become an endangered species in many hospital imaging departments, still have a role at MDACC. The standard time to have a typed report ready is 24 hours, although some are done in less than 6 hours. But referring physicians can access the verbal report as soon as it has been dictated, or see the transcribed report before it has been signed. “Voice recognition technology is not yet to the point that we can utilize it,” Silverman says. “We are such a high volume center that voice recognition will not be implemented until it is adequate to service the cancer center’s needs.”

Figure 6. Average daily CT procedures were plotted by week, illustrating that a radiology department schedule is a dynamic, flexible process and providing important staffing info.

As adapted from industry, Six Sigma methodology has had such a positive impact on CT processes that the radiology department plans to deploy it in other areas with expectations of similar success.

Richard B. Elsberry is a contributing writer for Decisions in Axis Imaging News.