Lean sigma. Six sigma. Lean six sigma. They are buzzwords in the pharmaceutical and contract research organization (CRO) industries, with executives tossing around “black belt” and “green belt” references at conference cocktail parties like so many half-eaten hors d'oeurvres.
But what is it exactly? How do you do it? And does something from 1980s, from the manufacturing sector, have any relevance in the bureaucratic, service-oriented world of industrial science?
At the Drug Information Association's (DIA) 45th annual meeting in San Diego in 2009, Tim Strauss of Covance explained the terminology more concretely than anyone we've ever heard. We called him to learn a bit more.
Strauss, executive director of business process improvement for Covance, explained that the company began incorporating six sigma concepts about seven years ago, gingerly starting with parts of the business that were finite, easy to compartmentalize, and lacking in variation. An example, he said, is a lab-based area like toxicology, which “tends to be easier to measure, and highly repetitive.”
By the way, six sigma is focused on reducing variation within processes, while lean sigma is about improving the way processes flow.
Strauss explained that the broad ideas involved any time six sigma is brought to bear include: identifying the problems that are making outcomes in a particular area less than stellar; creating elaborate maps about how processes are currently carried out; measuring exactly how serious each of the problems are; analyzing precisely how the process creates a problem; making a goal for how much of the variation is tolerable; identifying ways to tighten and improve the process; removing variation. Then the process starts all over again, with a new, elaborate map to illustrate how the revised processes will work.
But how did all that apply to process problems on the clinical side of Covance? With case report forms (CRFs). Employees were spending far too much time trying to find missing CRF pages.
“It was happening at various times across a study,” Strauss remembers. “We didn't have the data on the amount of time spent on it, but we knew it was worth taking a look at. It was high volume, a source of pain, and it was measurable.” And though the missing CRF pages problem seemed unwieldy, it was actually, in theory, a manageable project.
“Think about this process,” says Strauss. “One patient going into one site for one study at one location and producing one CRF. Now think about that across multiple studies all over the world, and you're talking about thousands and thousands of CRFs that we need to manage and make sure they are properly identified. Huge, yes. But really just a matter of realigning a repeatable process.”
Covance selected a team to address the issue. “It's important to pick people who are experiencing the pain, who want to do this, and want to be trained," says Strauss. "Then you get a real champion who will say, 'Here is the process.'”
The champion in this case was a green belt, a person highly skilled—certified, even—in the ways of six sigma. He and the trainees set about creating a detailed chart listing when exactly it tended to become clear to project managers and clinical research assistants (CRAs) that CRF pages might be missing, and then listing all of the times during a study when they might spend time looking for those pages.
Then the team asked all employees who came in contact with CRFs to clock themselves when looking for missing pages. Specifically, they were asked: How much time per every 1,000 CRF pages is spent looking for missing ones? Next, they analyzed what was going wrong.
8,000 Missing Pages
“You do a baseline metric on where we are today, then do analysis of all the processes,” Strauss says. “What are all the things that cause us to have to spend this much time and what are the variables?”
In one study, 8,000 CRF pages were missing and had to be tracked down. All were eventually found, but “we realized we had a process out of control. We had nothing operating in the background always making sure where the pages were,” he says.
The standard deviation, explains Strauss, represents the amount of variation in the process. The goal was to take all the variation out of the process, and make it completely regimented. Next, the team set a goal—a key part of the process. In this case, the team sought to reduce the amount of time spent looking for missing CRF pages by 85 percent.
Once the team understood where processes were breaking down, they pondered the details of each stumbling block along the way. The answers, once they stopped to look at them, were straightforward. The solution: software to track each CRF, allowing all parties involved to see its status. “It was a process change, with technology supporting it,” Strauss explains.
Covance created the system in-house using Excel and SAS, which took six months. It's called CRF Tracker, and can be accessed only by Covance personnel. It is yet another system for employees to use as they work on trials. Strauss says that "while fewer, more integrated systems is almost always ideal, our employees appreciate the quick and accurate results of this complementary system."
The six sigma team developed new training for the technology, and in the end, achieved a 98 percent reduction in hours spent looking for missing CRF pages, as measured in a small number of studies. Covance was able to alleviate the big, end-of-study crunch that happens at each site regarding the whereabouts of pages. “A lot of people like their job a lot more now,” he laughs.
A final process map was created after the technology was in place. “It gives you something to monitor the project going forward,” he says. “It can, on a monthly basis, allow us to say, 'Here's where we are on this.' ”
Strauss is the first to say that six sigma principles are not in and of themselves revolutionary. “If you pull together all kinds of statistical tools around process, you'll see that they've existed in many forms for many years,” he says. “What is new and different is pulling the tools together in a rigorous process and creating a system whereby you are forced to do each one before you get to the next.”
Six sigma concepts have been catching on fast in the pharmaceutical and CRO industries, but they're not ubiquitous. Strauss sounds wistful in thinking about how things might be if six sigma were more widely used, imagining a tight, streamlined, glitch-free process. The industry might transform research into a fast-producing machine of beauty. He hopes it's just a matter of time.
“There is really no hiding place where these concepts can't apply,” says Strauss. “You just find the variation and shift the mean.”
—by Suz Redfearnd9A2t49mkex