March 2012: Volume 38, Number 3
Blood, Sweat, and Robots
Who’s Ready for Lab Automation?
By Bill Malone
Lab automation is not just for the largest, cutting-edge reference labs anymore. Even smaller labs are now taking advantage of scalable automation to help them reduce errors, cut costs, and improve patient outcomes. But as more join the push to automate, laboratorians need to evaluate their choices carefully and consider their unique needs in order to achieve the full benefit of automation, experts say.
More choices in automation does not equal less homework for labs preparing for an automation project, according to Denise Uettwiller-Geiger, PhD, whose lab a decade ago became one of the first in New York to embrace automation. “Ten years ago, we were early adopters and there were not a lot of options available on the marketplace for lab automation, especially for a community hospital. Today, the availability of scalable automation solutions for laboratories of all sizes has soared,” she said. “Now laboratories have a lot more options, but as a result, each organization must examine its processes very carefully and identify its specific needs.” Uettwiller-Geiger is director of laboratories and clinical research for the clinical laboratories of John T. Mather Memorial Hospital in Port Jefferson.
Not If, But When
When it comes to automation, labs have lots of choices except one: do nothing. Driven by expanding test volumes, hospital consolidations, and unrelenting pressure to cut costs and reduce errors, labs have been forced to become experts at efficiency. But what happens after you’ve trimmed all the fat? With few options left, automation solutions that used to seem out of reach for the majority of labs have now become not just a possibility, but a necessity, say industry observers.
For many labs, automation appears to be the only answer to one of the biggest problems of all—the labor shortage that seems to have become permanent, according to Peter Manes, vice president of sales for Labotix Automation, a company that builds and markets independent open-architecture automation products. “When I first started in this business, a lot of the so-called automation experts ran around saying that a lab had to be in a hospital with so many beds, or had to cut so much labor to justify it,” he recalled. “Now things have changed completely. I think most people are saying, ‘we’ve got to get this done.’ Especially over the past two-to-three years, it’s not so much a matter of if but when a lab will automate.”
Mike Hoang, a marketing director with Beckman Coulter, agreed. “The mantra has been that small-or mid-sized labs could not afford automation or didn’t need that level of sophistication, but now we’re seeing that these smaller labs are experiencing such a labor shortage that they’re looking into automation simply to keep the lab running properly,” he said. “Of course, the degree of lab automation can vary, so the solutions we offer to a small- or mid-sized lab differ from those we offer large-volume reference labs.” For this reason, laboratorians should understand that not all lab automation is equal, Hoang added. They should look for a solution that fits their unique needs rather than a “cookie cutter, plug-and-play approach.”
Worldwide, the number of labs implementing automation grows by nearly 10% every year, according to industry analyst Greg Stutman, vice president of Boston Biomedical Consultants. “I think we will see a migration of automation from high- to moderate-volume sites as companies offer more flexible solutions,” he said. “In addition, we are in the midst of the first real replacement cycle for automation in the U.S. and parts of Europe, such as Germany, as the first wave of installations from over a decade ago are upgraded.”
The slow economy has made business tough for diagnostics manufacturers, who in response have become more creative in financing their automation products, according to Stutman. “Companies are offering longer finance terms or bundling automation costs together with analyzer costs to arrive at a total cost per test,” he noted.
The cost of many lab automation components, however, has remained about the same over the last decade. Now, the difference is that improvements in the technology enable a lab to achieve a positive return on its investment in a much shorter time, according to Hoang. “It’s difficult to measure the cost of automation, because it varies so much based on the customer’s needs. But from a Beckman Coulter perspective, you can buy the same piece of automation today as we had 10 years ago for about the same price, and because of the constant, incremental improvements that we’ve made, that piece of equipment offers much greater functionality,” he said. “By that measure, it has become less expensive.”
Laboratorians should keep in mind that automation does not always have to mean total automation of all lab operations, and even a small lab can benefit from automating just one or two processes, said Carmen Wiley, PhD. “Even something as simple as having your chemistry and immunoassay analyzers attached and letting the middleware decide which testing goes where first is definitely attainable for almost any size laboratory,” she said. “Small hospital labs also can benefit from a system with automated centrifugation. I think there is a huge benefit just from automating that one process.” Wiley is co-director of chemistry and immunology at Sacred Heart Medical Center and PAML in Spokane, Wash.
Automation can trim turnaround times, cut costs, and boost overall quality of lab services. Success, however, is not automatic. So say experienced laboratorians and industry experts, who encourage laboratorians to study and understand the specifics of their own lab’s workflow before an automation project.
For example, Wiley recommends a technique she calls “a day in the life of a tube.” This approach uses a tube of blood to create a narrative that records how samples move through the lab. In the process of writing the blood tube’s story, staff get a new perspective on their lab’s workflow. Wiley undertook her first major automation project in 2003 while at the Marshfield Clinic in Wisc. “At that time, automation still felt very new to us, and as the lab director I didn’t feel that we understood well enough how samples actually moved through our lab,” she said. “I worried that we were so focused on trying to maximize what automation could do for us, we couldn’t see the forest for the trees.” To get the big picture view without losing the details, Wiley met with her technical supervisors and asked them to take her on a tour of a day in the life of a tube—not just to talk her through it, but literally walk through the lab following a tube and documenting it. Having been through this exercise, she recommends that labs looking at automation do the same, with the goal of conceptualizing what a day in the life of a tube would look like after automation (See Box, below).
Before You Automate
Questions for Colleagues and Vendors
As more labs prepare to automate, laboratorians who have been through large automation projects recommend that laboratorians take time to research how a prospective vendor’s products work in real life. Carmen Wiley, PhD, presented the following questions at the December 2011 AACC conference in Las Vegas, “Optimizing Your Lab Automation: Lessons from the Front Line.” Wiley suggested that the team planning lab automation set up conference calls with other labs already using a prospective vendor’s products and consider the following questions.
- How did the lab function prior to automation?
- What were the goals for lab automation?
- What expectations were met?
- What expectations were not met?
- What is the lab’s up time?
- How does the lab handle planned and unplanned down time?
- What maintenance is required for the automation line and the instruments?
- Which are the strong and weak analytical assays?
- What was the learning curve for staff?
- What are the laboratory information system requirements?
- Did the lab have to remodel?
- What was the worst part of the process of automating?
- What was the best part of the process of automating?
- What were some unexpected gains?
- What were some unexpected losses?
Likewise, even if a lab has the resources to move toward total lab automation, the management team should not assume that each and every process must be automated, noted Wiley. While at first this may seem counter-intuitive, some tasks may be better left as manual processes. For example, a hospital lab will receive samples from pediatric, elderly, and extremely ill patients that may arrive in unusual tube sizes or with very small volumes. Similarly, a reference lab can receive non-standard tubes from some clients. If these samples only make up a fraction of test volume, it might be easer to handle them manually rather than spend a lot of effort trying to force a workaround within the automated system. “Some of these unusual samples don’t make sense to automate if you’re going to give them special care anyway,” Wiley commented.
By the same token, labs should make sure they don’t automate a bad process, noted Manes. “A significant part of what we do is helping customers prepare for automation,” he said. “A lot of labs have been through Six Sigma and Lean projects, and if they haven’t done that, we highly recommend they consider it. As much as it pains me, there have been many times where I’ve had to tell folks you’re just not ready for automation yet. They need to do that leaning process beforehand to make sure they’re automating good processes.”
Labs should also take a good look at their current laboratory information system (LIS). In fact, the LIS should be seen as the core of lab automation, said Lisa-Jean Clifford, CEO of lab software company Psyche Systems. “If you’re using an LIS the way it’s intended to be used, then it’s definitely the central hub of your laboratory,” she said. “So if you’re looking to embark on any major task in your lab, it would behoove you to understand your current LIS’s full features and capabilities and whether you need something more. Sometimes people assume that all LISs have the same functionality. And the answer is no, not all LISs are equal.”
For a lab to benefit from automation, everything in it has to be bi-directional, Clifford emphasized. The LIS pushes information out—to instruments, to other applications, to outreach solutions, to other facilities—and at the same time captures information and pulls it back into the LIS. “This way, data lives not only with the patient case information that’s accessible for the final reporting, but also in the database of the LIS so that the laboratory can do true data management, statistical reports, and data churning,” Clifford said.
Planning for the Long Term
Getting ready for automation requires plenty of introspection. At the same time, those interviewed for this article stressed that it is equally important to look at the lab from the outside—how will the needs of patients, physicians, hospitals, and others change over the course of the product’s lifecycle? “Flexibility and scalability are essential,” said Hoang. “A lab may consider itself small- or mid-sized today, but in 10 years, after growth in patient population or organizational consolidations, it might find it’s a large lab. The lab’s automation solution needs to grow with it.” Automation pieces like tracks and specimen sorters frequently outlive other lab instruments’ 5- to 7-year replacement cycle, he noted.
How a lab might partner with other hospitals, increase its outreach business, and bring tests in-house are just a few of the areas that require thought. Uettwiller-Geiger encourages laboratorians to look at the lab from the hospital’s perspective. “Today we’re taking a more global view, aiming for improved bed management in the hospital,” she said. “Most hospitals are not getting any more beds, so it’s really important, particularly in the emergency room, to see how the lab’s turnaround time helps clinicians make appropriate decisions in terms of triage, diagnosis, and treatment. Now we’re not talking about shaving 5 or 10 minutes off. We’re asking whether we can shave a minute off. We’re always trying to beat the clock.”
A significant investment in automation also brings with it a long-term relationship with one or more vendors. On this point, labs are moving in several directions. Many rely on a single vendor for instruments and automation. Others opt for a mix of vendors that will work together. Stutman of Boston Biomedical Consultants noted that vendors often look at automation as a means to secure reagent and instrument contracts.
Further complicating these choices, some diagnostic manufacturers that also sell automation solutions have systems that will work with other vendors’ products, while others don’t. Systems that will not easily admit or communicate other vendors’ instruments are considered closed.
Manes has seen Labotix’s business increase every year, a trend he attributes to laboratorians’ desire to acquire best-of-breed instruments from several vendors and still connect them through an independent automation solution. “After making a large investment in automation, a lab does not want to have to throw that all away if one day it choses another instrument,” he said. Labotix systems are open to any vendor’s instruments, and Manes noted that increasingly, other vendors are giving up on the closed system model. “In the old days, independent automation vendors such as Labotix and the manufacturers that had automation systems were bitter rivals,” he said. “Now, that has dramatically changed with most vendors.”
While his company’s automation solutions will connect with other vendor’s instruments, Beckman Coulter’s Hoang emphasized the benefits of consolidating with a single vendor. “I see customers wanting to streamline their operations and simplify the number of customer-vendor interactions. They want to achieve a holistic solution for their lab,” he said. “However, we’ve always been able to connect third-party analyzers to meet the needs of our customers.”
How various instruments and automation pieces work together could become even more important soon. One of the biggest trends in automation, Hoang and others noted, is consolidating core lab disciplines. In this kind of consolidation, formerly discrete lab areas such as chemistry, immunoassay, hematology, and coagulation become one super-system on a single automation track. So far, few labs have successfully consolidated this way, but according to Hoang, many of his customers would like to do so.
Tracing a Blood Tube’s Path
The typical path of a sample in the laboratory can take many twists and turns. Before embarking on an automation project, labs can get a better handle on their workflow by tracing the steps of a blood tube through the lab. The same concept applies to describing exactly what a lab needs from an automation system. Carmen Wiley, PhD, offered the following example for a hematology analyzer at a recent AACC conference.
- The 50-tube rack gets placed on the inlet. A robotic arm picks up the tubes individually and places them into individual pucks, one tube per puck, on the automation transport line.
- The tubes automatically proceed to the first bar code reader to electronically document arrival into the laboratory computer system. The tubes spin into position automatically so the bar code reader can scan the label.
- Non-hematology, lavender-top tubes with HbA1c, ESR, and RBC-folate orders only get routed to the outlet, the designated area for other testing, for manual transport to the appropriate testing area.
- Lavender-top tubes with hematology orders, with or without other tests, proceed down the automation line to be loaded into hematology racks by a robotic arm and then automatically proceed to the analyzer for testing.
- If a slide is required, the slidemaker will make a slide and the slidestainer will stain the slide.
- The tube then automatically proceed either to the ambient stockyard for storage or to the outlet if additional off-line testing is required.
The Technologist’s Changing Role
One of the hurdles for labs in the early days of lab automation was the fear of job elimination among seasoned technologists. But for labs that are often chronically short-staffed, automation can actually enhance jobs instead of eliminating them. Not surprisingly, overworked technologists often welcome the opportunity to perform other work in the lab once automation takes over more mundane, repetitive tasks.
Often, the technologist time that automation frees up can be used to bring more tests in-house, Wiley said. “After automation is put in, I’ve often been able to move those staff to other areas of the lab where they’re actually making better use of their skills,” Wiley said. “For example, all of a sudden you have these really smart, highly trained medical technologists whom you can use to bring in testing that you used to send to an esoteric laboratory.”
Uettwiller-Geiger said she considers automation in her lab to be a recruitment and retention tool. “I believe that an associated benefit for laboratory automation is that it does create a high-tech environment that allows us to use people’s talents to the best of their abilities,” she said. “With automation, we have been able to bring in other programs with more complexity, such as molecular methods, which I think expands the role for medical technologists and makes the job more attractive.”
Share Your Story
Considering all the planning and hard work labs put in over the course of implementing automation, Wiley encourages others to share their experiences with colleagues. “Each lab has a unique experience gathering information, creating a plan, and finding creative solutions on their path to automation,” she said. “And the best way we can improve this process—both labs and vendors—is to share our experiences.”