Process Improvement: Real Methods that Work
A single technologist at Columbus Regional Hospital acquires a CT, formerly a two-person job. Scans are now performed on outpatient-designated scanner. RIS screenshot (left) courtesy of Fujifilm.
Process improvement is a staple of high-quality care. With methodologies borrowed from manufacturers, hospitals are honing process improvement to a science, affixing it as a pillar of quality assurance and the fulcrum of improving care. As belts tighten, reforms hang in limbo and public and legal scrutiny reaches new heights, process improvement has become more valuable than ever.

Modern process improvement methods proceed from a common premise: large-scale improvements depend on bettering processes—not individual workers. Under this sweeping insight, poor performance and errors are predictable outcomes of imperfectly designed processes. This view can be traced back to the 1980s, with Lean and Six Sigma manufacturing methods' development in Japan and rapid adoption across industries and continents.

In healthcare, the shift is epitomized by a 1989 New England Journal of Medicine commentary by Donald M. Berwick, MD, MPP, contrasting the 'Theory of Bad Apples' and the 'Theory of Continuous Improvement,' wherein large scale improvements to quality depend not on rooting out bad apples but on eliminating inefficient processes. Consider:
  • To keep up with toughening competition, Columbus Regional Hospital in Indiana managed to increase CT volume by 600 studies per month, cut exam time by 33 percent and spur a 34 percent growth in CT reimbursements—all with fewer scanners and technologists. And they'll valet your car at no charge.
  • Akron Children's Hospital slashed MRI waiting time to get an exam from 28 days to three or fewer—generating a 25 percent increase in weekly exams and a $1.3 million increase in yearly revenue.
  • St. Elizabeth Healthcare in Northern Kentucky now communicates all critical radiology results directly to ordering physicians within 30 minutes of the finding.

Some powerful underlying modi are at work here. One of the more common overlapping methods arising out of Six Sigma is DMAIC—Define, Measure, Analyze, Improve and Control. Meanwhile, Lean prefers a simpler mnemonic (Plan-Do-Study-Act), but expounds a rigorous standard for healthcare: "Anything that occurs in a patient care process that does not provide value to patient care is considered waste," according to Ken Aakre, RT, quality coordinator and CT operations manager for Mayo Clinic's department of radiology in Rochester, Minn., and colleagues, in their February 2010 American Journal of Roentgenology article, "Process Improvement: What Is It, Why Is It Important, and How Is It Done?"

Aakre urges practices to make regular use of quality improvement advisors. Regardless of the size of the practice, a quality improvement expert can justify his or her position on the payroll. Moreover, effective programs demand methodical planning, monitoring and follow-up: time and data analysis should not be spared when paring away inefficiencies.

Down with dose

"In the fall of 2007, back before it was all over the media, our department started discussing the importance of attentive radiation exposure in children. We realized our protocols were not standardized, which was a really big gap and clearly where we needed to start," says Kelly Firestine, RT, CT manager and project leader for Memorial Hermann Healthcare System in Houston's CT radiation dose reduction program.

Firestine dug a bit deeper and found that few of the system's 11 hospitals, 37 CT scanners overall, had any standardized radiation dose protocols for children. Thus, she delineated the project's scope: "To reduce CT head and abdomen radiation dose in kids, without compromising image quality." The objective revealed its challenges immediately; Firestine went straight to the American College of Radiology's (ACR) website to begin to set the network's dose protocols. To her surprise, ACR offered only its recommendation for an abdominal CT scan of a five-year-old.

Enlisting Shannon Lutz, Memorial Hermann's director of performance improvement for retail services, the project got a boost from the Association for Medical Imaging Management (AHRA) and Toshiba, winning a Putting Patients First grant that enabled the radiology department to hold a kaizen, a one-day rapid improvement event aimed at soliciting staff and expert ideas about the project.

By then, Firestine and Lutz had collaborated with University of Texas and departmental physicists, radiologists, hospital executives, administrators, nurses and technologists. "One of the most important moves was recruiting a physician champion and designating several key players to lead the project," Lutz says. The purpose of the physician champion was two-fold: First, the project needed expert dosimetry input to determine protocols and maintain image quality. Second, no clinical radiology project will succeed without widespread support among radiologists, who often hold a group-psyche best penetrated by peer radiologists who understand the advantages of a project and can speak to its worth in disrupting workflow.

The radiologists were instrumental not only at Children's, but across the system. From 2008 to 2009, the mean head CT dose was reduced by 50 percent in children, the average abdominal scan by 60 percent. The mean head CT dose index (CTDI) among infants up to 18 months old was reduced from 56 mGy in 2009 to 25 mGy in 2010. Among patients above 101 pounds, abdominal CTDI between 2009 and 2010 was cut in half, from 16.17 mGy to 8.9 mGy. By 2010, the project had garnered 100 percent compliance. The successes quickly inspired protocols for adult procedures as well. (Visit to learn more about pediatric dose guidelines.)

Lutz and Firestine also report substantial drops in repeat imaging and cost savings. As part of a series of standardizations building on the success of dose reduction, they developed a requirement that for CT scans with excessive motion, children could only be rescanned once, after which sedation or a different test would be necessary. This has increased throughput tremendously for the hospitals, Lutz and Firestine say.

Firestine reflects that few successes would have been possible without meticulous audits, the 'control' phase of Six Sigma process improvement. "Auditing is a big part of process improvement," she says. "Not only does it provide information on compliance, auditing keeps the project fresh in workers' minds, reminding them that they're accountable for their work."

Firestine says her eight hours a month of sampling exam information to test compliance has been instrumental in achieving 100 percent adherence to protocols. All reporting information is shared between hospitals and included as part of technologists' performance evaluations. She indicates that peer pressure from transparent reporting has induced the most motivation to follow the standards and even to scrutinize orders. "Many times, techs will look at an order and check a patient's record, to find that a patient has undergone a recent enough CT scan and consult the radiologist. Technologists are now much more confident in discussing these topics with patients and radiologists."

Patients are customers

Given the nature of the physician-patient relationship, it is often easy to forget that patients are paying customers. "Every project has a customer, and it is important that the team listen to each customer group to understand what is important to them, what their requirements are, and whether they will be satisfied when the goal is reached," Aakre and co-authors write. Columbus Regional Hospital (CRH) did just that.

Following the opening of a local private practice by a group of former CRH radiologists, the hospital began to lose out significantly on CT outpatient visits, with physician offices calling both local imaging providers to more or less bargain for the soonest appointment. Just about this time, CRH's CEO and CFO had rolled out a Lean Six Sigma department. With the CFO seeing radiology as a profit center, the CEO wanting to make CRH a better place to work and Kent Johnson, director of radiology, having compiled a laundry list of ways to improve CT during his 30-year tenure at CRH, radiology secured the bid for Lean Six Sigma's second project.

CRH sent staff to almost every referring physician's office in the community to find out what it would take to get physicians to send their patients to the city's only hospital. This voice-of-the-customer campaign was used as part of the hospital's kaizen brainstorm, which threw together a barrage of ideas from the surveys, techs, nurses and physicians, both inside and outside of radiology. The customer focus was the start of radiology's effort to address all the disadvantages associated with being a large hospital, compared with the conveniences of a walk-in radiology group.

The department began staffing to demand, and shaved 37 minutes off wait time by pre-registering patients to allow them to bypass hospital registration. Staff also prepped patients and inserted IVs before taking them into the exam room. Another problem Johnson noted was that "although we had two techs assigned to each CT," none had specific duties, so they would be shuffling between running the scanner, prepping the next patient, post-processing images, answering the phone and a dozen other tasks.

By designating responsibilities to individual technologists, the department sped average CT exam time from 30 minutes to 20 minutes, enabling CT volume to grow by one-third within 18 months.

CRH now uses one of its two scanners for outpatient imaging only, saving the other for the unpredictable fluxes of emergent cases. With the help of an expanding ED and the background growth of CT seen across the country, the Lean Sigma project increased monthly volume approximately 60 percent, from just over 1,000 exams to 1,600, which has upped average monthly CT revenue about 40 percent from $687,000 to $960,000. Outpatient imaging is now available same-day, and a host of conveniences like free valet parking has led to a jump in patient satisfaction.

Johnson credits project leaders and staff. "Having the staff work on these projects instead of the managers is really important. They deal with the processes every day, they know the bottlenecks, so they're the right people to identify and fix these problems." In his opinion, "the manager's job is to knock down the barriers to change, to enable the workers who see the problems to fix them on their own."

Expeditious MR

Another throughput project at Akron Children's Hospital has made Lean Six Sigma projects a staple of the hospital's radiology department. Akron's executives as well as the chairman and the director of radiology take part in frequent kaizens and projects under the rigorous DMAIC operation, while also holding daily huddles to reinforce current improvement projects. And because the kaizens consistently involve key stakeholders, such as hospital executives and Michael Rubin, Akron Children's chairman of radiology, "the events get rid of a tremendous amount of red tape and speed up projects that would normally take months to get passed."

"Plain and simple, I received multiple calls weekly from physicians complaining because they could not get MRIs in less than a month," Rubin says. Without adding more machines or staff, the department increased the average number of scans per week (between 2009 and 2010) from 86 to 107, resulting in an additional $1.3 million in yearly MRI reimbursements.

Lean Six Sigma enables such catalyzing shocks to throughput largely because of the DMAIC focus on good data, by which teams can pinpoint issues and bottlenecks. The methodology also comes with a slew of statistical analysis methods, such as "Pareto charts, which illustrate the frequency with which a certain problem recurs, projecting clear priorities that need to be tackled first," Aakre notes.

Managing critical results

A recent spurt of studies has centered on improving communication of critical radiology results, a source of serious risk to patient safety and the second-most common class of malpractice suits in the U.S.

Technology is playing a crucial role in reversing this trend. St. Elizabeth Healthcare, a system made up of five hospitals and four outpatient imaging centers in Northern Kentucky, was plagued by trivial obstacles to the effective communication of results, including empty printer trays, results lost among other printed pages, failed initial communications with referring physicians which never drew follow-up and a lack of independent audits.

The hospital has made use of pre-installed software used primarily for billing to overcome these challenges. All reports with critical results are electronically documented with short details and instructions to ordering physicians. If checked as critical, the standardized document is printed from PACS to the main hospital's Clerical Office of Digital Media and Communication (created solely for communicating critical results). A digital media staff member or radiology tech, in the office 24/7, calls the ordering physician and reads the radiologist's instructions, with all calls recorded and archived on the office computer.

According to Digital Media Supervisor Teresa Hatfield, the system has amassed greater than 95 percent compliance. Hatfield performs daily audits of critical results reporting, bolstered by the St. Elizabeth quality assurance department's formal monthly reports. She says one-on-one training accounts for the near-perfect compliance and notes that the recorded conversations "have saved us over and over" from physician offices who claimed that they were never informed of critical results.

Often radiology processes are rife with inefficiencies and unrealized fruitions, but there are solutions to truly improve efficiency and quality. Crucial elements of process improvement are:
  • Follow Lean or Six Sigma.
  • Pick physician champions who are enthusiastic about the project and carry the department's respect.
  • Include a healthy number of experts and executives to guarantee support, implementation and sustainability.
  • Don't underestimate staff's ability to identify common problems and buy into solutions.
  • Perform root cause analysis, including kaizens and voice-of-the-customer surveys. Aakre also recommends asking why a problem occurs five times to get at its true root. "Each 'why' will provide more and more detail about the root cause. Often times with asking 'why' only once you do not touch on the true root cause."
  • Splurge on data, both to diagnose the problem and to ensure compliance and accountability; the 'Control' phase of Six Sigma programs lasts two years.