April 2008: Volume 34, Number 4
Tools for Benchmarking Performance
How Lean, Six Sigma Improve Lab Efficiency and Quality
By Phil Kibak
The same principles that have allowed Toyota to efficiently and profitably assemble and sell millions of highly regarded automobiles are finding homes in clinical laboratories. The guiding values of the Toyota Production System are to work intelligently and eliminate waste. Incorporated into a set of management principles for maximizing efficiency known as Lean, these techniques are changing the ways clinical labs operate.
Growing numbers of clinical laboratories have applied Lean and another program called Six Sigma, which is a set of principles for controlling variation within a process. By creating continuous process flow and eliminating waste, these laboratories are now starting to reap the benefits. A study of 100 labs conducted by Management Insight, LLC in Ann Arbor, Mich., demonstrates that labs using Lean principles significantly improved their performance, compared with labs that employed traditional management techniques. Sponsored by the Dark Report, an organization that specializes in laboratory management reports and conferences, the study provides details about how Lean labs excelled in improving TAT, maintaining good workflow despite heavy test volume, and reducing excessive TAT, also called outliers. The study was conducted by Thomas Joseph, MBA, MT (ASCP), Managing Partner of Management Insight.
Comparing Labs’ Performance
“Our database evaluated 6 million performance measurements from more than 100 laboratories, 14 of which were early adopters of Lean processes,” said Joseph. “Although there have been a number of case studies illustrating the value of Lean in lab management, this is the first time it has been demonstrated that Lean labs have far exceeded performance levels of the best conventional labs.”
The data show dramatic differences between Lean labs and labs using more conventional management techniques in different areas of performance. For example, 89% of Lean labs had a STAT CBC TAT of 12 minutes or less, while only 16% of conventional labs achieved that level of performance. Typically, Lean labs showed routine CBC TAT of no more than 20 minutes compared with only 30% of conventional labs. As many as 75% of Lean labs had collect-to-receive times of less than 20 minutes for morning draws.
“One of the hallmarks of Lean is the move toward single piece flow, away from the processing of large batches of specimens,” Joseph explained. “By sending patient samples for processing in small batches, even one patient draw at a time, the lab gets the work into the processes much faster, and as a result morning draws get tested and reported much more quickly.”
The study also revealed how reductions in batch size led to dramatic effects on TAT. With a batch size of 15 specimens, overall TAT was approximately 112 minutes. At a batch size of 3, TAT was about 50 minutes. Reducing batch size to 1 improves on that a bit, Joseph said, but the improvement has to be weighed against the additional time the phlebotomist spends walking from the patient site to the pneumatic tube after every draw.
The study also reveals that Lean labs perform significantly better in terms of outliers. When examining TAT beyond 45 minutes from receipt to verification, Lean labs have a lower proportion of outliers (1.0–1.9%) than the better performing conventional labs (2.6%–6.4%).
The Lean approach has a positive effect on every step of the process, according to Joseph. As the volume of tests increases, a bottleneck that raises TAT forms in traditionally managed labs. But volume spikes in Lean labs have little effect on TAT. Joseph also noted that as the lab size increased from 200,000 to 1.5 million billable tests, the TAT rose from about 11 minutes to more than 20 minutes for conventional labs. However, there was no such relationship for Lean labs, he said, adding that the data showed that Lean labs manage workflow efficiently despite increases in annual test volume.
The data also show that Lean labs have, on average, reduced technical staff by 40%.
What the Lean Labs Say
Laboratorians who have experienced the process of making a lab Lean draw dramatically different pictures of their labs before and after the change. In 2004, administrators at the Avera McKennan Hospital and University Health Center (Sioux Falls, S.D.) realized they needed to change the way their crowded and inefficient clinical laboratory operated. After looking at the way Lean had been incorporated into operations at two other medical centers, they contracted with Ortho Clinical Diagnostics’ ValuMetrix Services to implement the process.
“Normally, a Lean project takes about 14–16 weeks to come to fruition, but we took a bit longer because we changed direction, from a redesign of an old lab to building a completely new lab,” said Leo Serrano, FACHE, CLSup (NCA), the institution’s Laboratory Services Director. “We worked on the design of the lab from July to August 2004, and we moved into the new facility in January 2005. We actually reduced lab space by about 1,000 square feet. Lean has totally changed the way we work, and we’re much more efficient and productive. In fact, we were able to add new technology, such as molecular diagnostics, virology, and flow cytometry. And we’ve grown about 30% in volume and now do about 2 million tests per year.”
“The old lab was pretty traditional,” added Serrano, who is a Lean/Six Sigma expert known as a Black Belt. “We had compartmentalized sections with microbiology, blood bank, hematology, and chemistry sections. When a new device was added, we had to scramble to find a place to put it. But that changed with Lean because you design the work area with the processes and flow in mind.”
The lab has a “central core” with nine instruments arranged in an oval pattern. One technician is responsible for loading and releasing the results on all nine instruments and operates within a 6-minute cycle. The lab’s “manual core” handles urinalyses, blood gases, and other tests and is staffed by one or two technicians. In addition, there is a floating technician whose base is the manual core, but who also steps in to assist the central core with calibration, instrument repair, and whatever else may be needed.
The results have been dramatic. “Before we made these changes our morning collection TAT had been taking as much as 70 minutes. Now we’re looking at 35 minutes from collection to verification,” explained Serrano. “We also were able to reduce staffing through attrition. No one lost a job through adoption of Lean, but we were working efficiently enough so we could choose to not replace about 11 FTEs. The lab also has realized significant cost savings, and has had as much as a sixfold return on the investment.”
Communication is the key to the success of Lean, he added. “People tend to be afraid of major change and you really have to communicate what’s going on. We had a lot of staff involvement and kept everyone in the loop at every stage. Lean doesn’t have to be mean.”
A Lean Hospital Experience
The switch to Lean management processes was more of a necessity for the Children’s Hospital and Regional Medical Center in Seattle, Wash. “A few years ago, the state was going to balance its budget by changing Medicaid eligibility, a move that would have presented our hospital with a $10 million shortfall,” said Joanne Simpson, the hospital’s laboratory director. “Although this never actually occurred, it greatly accelerated the Lean movement that already had started here in 1999. As a result our hospital has gone from two Lean projects in 2002 to about 100 in 2007.”
In the lab, Lean implementation has resulted in improved TATs. In May 2004, only 71% of creatinine tests had TATs of 60 minutes or less, but by April 2007 this figure had increased to 97%. Similarly, only 52% of gentamicin tests had TATs of 60 minutes or less in 2004, but by 2007 all such tests had achieved this target. The lab also increased its annual test volume by approximately 20% to more than 900,000. Simpson said that Lean practices have allowed her lab to achieve this higher volume without adding any FTEs. “The standard before was if you raised test volume, you needed to add FTEs to perform the additional testing.”
The Lean transition process took around 6 months, including the design and construction of the facility. “We used to have isolated islands of work with manual and automated testing comingled,” Simpson said. “Now we have automated testing in a cell and manual testing in a separate area. Lean has allowed us to realize a lot of gains despite the fact that 50% of our samples are actually microsamples, and these present unique challenges,” she added. “We often have to move the aliquot from bench to bench to bench, which is very challenging for efficiency. But Lean has helped significantly. By putting processes together in these automated cells, lab personnel reduced the number of steps in the work flow and achieved greater efficiency.”
Staff morale is a huge factor in determining the success of Lean implementation. “I’ve seen labs in our area that have started using Lean techniques but then reverted back to traditional processes almost overnight,” Simpson said. “It takes a tremendous amount of energy to sustain the gains and that’s why staff buy-in is essential.”
Eliminating the Dividing Lines
Gayle Culbertson, Executive Director for Laboratory Services for Iowa Health-Des Moines echoed that sentiment and added that a lot of her lab’s success with Lean techniques comes from perseverance and working closely with lab personnel. “For my long-term employees, it was hard for them to suddenly not be the experts and to be the new students learning new skills. From a management standpoint, you have to be a strong leader and include everyone in the forward path.”
Culbertson did not have an administrative mandate to change the way her lab operated, but was encouraged to investigate Lean by her former mentor. After attending a seminar put on by Ortho Diagnostics she said it was “like a light bulb went on over my head.”
She chose the core lab at Iowa Methodist Medical Center to be the lab system’s first Lean project in 2006. “Our TATs were pretty good, but they were inconsistent,” Culbertson explained. “Our hemoglobin TATs went from an average of 35 minutes to an average of 20 minutes and we can hit that figure 98%–100% of the time. Before we instituted Lean, we were sectionalized. The teams on our day shift did not cross train at all. It was like there was a big dividing line between the hematology and chemistry sections. Now, our automated line in the core lab is mixed and the technologists can run both types of instruments. And that line accounts for about 90% of the work done in the core lab.”
The lab’s efficiency has been enhanced through Lean management, Culbertson added. Before instituting the process the hours per billable test stood at 0.194. In February, that figure had fallen to 0.178. “We’ve experienced tremendous growth over the last 2½ years. Last year, our core lab realized a 12% increase in volume. Overall, our test volume has increased by over 200,000 since we’ve instituted Lean, and we’ve been able to do that without adding personnel.”
Lab Medicine Best Practices
In January, the IOM issued a report recommending that Congress direct the U.S Department of Health and Human Services to establish a program with the authority, expertise, and resources necessary to set priorities for evaluating clinical services and to conduct systematic reviews of the evidence. “This is exactly what we’re attempting to do for laboratory medicine,” said CDC Senior Economist Susan R. Snyder, MBA, PhD, who also is the project officer for “Laboratory Medicine Best Practices: Developing an Evidence-Based Review and Evaluation Process.”
The first phase of the CDC project, completed in December 2007 and soon to be available online, was developed by CDC and Battelle Memorial Institute project teams with the assistance of an external workgroup of multidisciplinary experts. “What we’re doing for the field of laboratory medicine is developing systematic and transparent evidence-based review methods for the purpose of making evidence-based recommendations,” said Snyder, who is with the CDC’s Division of Laboratory Systems. “It comes down to a simple concept—we’re trying to find what works.”
In Phase 1, the teams identified key terms and definitions, inclusion and exclusion criteria for laboratory medicine practices, priority areas for candidate practices, systematic review methods, and an evaluation framework and criteria. The project teams and the expert workgroup completed a proof of concept test of the review methods, and recommended next steps.
What they have also found is that other organizations making evidence-based recommendations have relied primarily on published evidence. The problem facing the field of laboratory medicine, noted Snyder, is that relying only on the limited quantity and quality of available published studies, especially in scientific journals, is generally insufficient for making recommendations. “Our expert workgroup felt strongly and committed to moving forward with this, but without sufficient published evidence we felt we had to try something new and innovative to support this initiative.”
The group recommended tapping the large body of unpublished evidence that exists. “Labs do a lot of internal management and operational evaluations when they implement something like Six Sigma or Lean, and go through continuous quality improvement practices,” explained Snyder. “The expert workgroup decided we needed to figure out how to access this information, and that’s what we’re now doing in Phase 2 of the project.”
Phase 2 will refine and further develop the methods identified in Phase 1. It also will evaluate implementation options for sustaining the process and disseminating information consistent with stakeholder needs, as well as recommend strategies. As part of Phase 2, a pilot test is underway to assess the review and evaluation methods along with querying a newly created network of laboratories for unpublished evidence on practices in two topic areas—patient/specimen identification and critical values communication.
More information on the CDC Laboratory Medicine Best Practices project will be available beginning April 1 at the CDC Website.
The IOM report—Knowing What Works in Health Care: A Roadmap for the Nation—can be viewed at the IOM Website.
Is More Automation the Answer?
Laboratorians might be tempted to think that more automation is the key to improving lab efficiency. But Joseph cautions against this thinking. “The first thing you want to do when you analyze your lab operations is study your process and eliminate the waste,” Joseph explained. “And automation is not the silver bullet that some managers think it is. Most Lean labs have little or no track-based automation. All have developed workcells combining chemistry, hematology, coagulation, and other analyzers that can be staffed efficiently during both low and high volume periods.”
Automation can make a difference in quality, but it’s not the striking change that one generally expects, said C. Martin Hinckley, PhD, President of Assured Quality in Perry, Utah. The firm helps companies improve quality via mistake-proofing and simplifying processes. “There’s an adage among industrial engineers,” explained Hinckley, who has conducted research that has examined management and workflow systems affecting clinical chemistry operations. “You never want to automate a bad process.”
“In theory, Six Sigma will get you error rates of 3 per million,” Hinckley said. “In practice though—and there are data from Motorola, Texas Instruments, and a number of other companies that have applied this for many years—defect rates rarely fall below 1,000 per million. And the reason for this is that Six Sigma does not deal with probability events, which you can’t predict. An example of this, which commonly occurs in the clinical laboratory, is the label—either a label has been left off of a specimen, or the wrong label is affixed. But mistake-proofing addresses this sort of thing and tries to implement interventions that block these potential errors.”
With regard to sample labels, there often are a number of steps in the process and each step increases the risk of error. For instance, in a hospital the phlebotomist often must leave the patient site and walk to the nurses’ station to get the label, and he may pull the wrong label out of the wrong file or folder, or go back to the wrong patient. A better approach, Hinckley offered, is to scan the bar code on the patient’s wristband and immediately print a label at the point of use. But, he added, there are things that stand in the way of achieving a lower defect rate and that usually involves tradition. “We often have procedures in place in hospitals that are very mistake prone and we implement layers of controls to handle them instead of looking for true mistake-proof solutions,” he said.
Although automating a process might make a task easier to perform, it sometimes actually makes the task more difficult. In a study Hinckley published in Accreditation and Quality Assurance, he illustrated this concept with an example involving the Electronic Medication Administration Record (eMAR) system to prevent medication administration errors. With eMAR, nurses summon prescriptions on a computer, scan the patient’s armband barcode, and scan the medication barcode prior to giving the patient the medication in order to create the electronic record. This also makes an automated check of the medication against the correct patient and prescription. But with this system, the nurse must now move the medication cart to the bedside instead of merely carrying the medications into the patient’s room. In addition, the cart cannot be left unlocked and unattended, a practice that can delay an urgent response as the cart must then be moved and secured (Accred Qual Assur 2007: 12; 223–30).
“Effective automation should always reduce the difficulty of the global task,” Hinckley said. “When implemented with what’s referred to as jidoka, which is Japanese for ‘automation with a human touch,’ it helps the worker perform tasks faster and more easily with fewer errors. And achieving the best results with automated equipment requires extensive use of mistake-proofing.”
As a mistake-proofing example, take the case of a centrifuge that shuts down unexpectedly because of an unbalanced pattern of vials, or because of failure to remove a vial from a previous loading. Analysis shows that each of these factors is a mistake. Potential solutions include ways to sense the imbalance by means of a vibration sensor with an alarm; a current sensor, timer, and alarm; and wiring directly into the shutdown switch to activate an alarm if the switch is activated. Adjunctive solutions address the type of alarm—visual, auditory, or both. The concepts are scored on a 1–5 scale with 1 being the poorest and 5 being the best. The selected concept is then implemented and evaluated.
The Bottom Line
But the bottom line, according to Joseph, is that Lean labs perform significantly better and with greater efficiency compared to traditionally managed labs. Instead of being assigned to a specific machine or device, the laboratorian is assigned to a work cell that involves managing and releasing results for several analyzers. “Not only can it boost productivity, but if applied properly this kind of work flow management also can improve cost effectiveness by relying on fewer personnel,” he said.
“At this point, though, it’s still a minority of labs that have adopted these processes,” Joseph noted. “The philosophy of Lean requires a very different and organized kind of work style and to be effective it requires a high degree of discipline and structure. It’s a cultural change that not everyone wants to make, but it’s continuing to take hold. As people see how dramatic the results can be, I think it will be adopted by more lab managers.”
Joseph’s research in laboratory operations has led to the development of a benchmarking service that can assess lab performance in every aspect of lab operations. For more information about the performance benchmarking go to the Management Insight Website.