CEO SUMMARY: Since October 2003, the core laboratory of Detroit Medical Center University Laboratories has operated with a home-grown total laboratory automation (TLA) system. The 100-foot automated line currently connects to 11 instruments and the hardware cost only $200,000 to build and install. Not only did it allow 20 medical technologists to shift to other positions within the laboratory, but this automated line has operated 24/7 for almost three years without a major stoppage.
IN AMERICA, THERE IS A GREAT TRADITION of someone who gets the “better idea,” begin to tinker in his garage and eventually emerges with useful new products that find rapid acceptance in the marketplace. The first two statements above aptly describe the activities of pathologist William Neeley, M.D., FCAP, DABCC and his colleagues at the Detroit Medical Center-University Laboratories (DMCUL) in Detroit, Michigan. For the past three years, they have operated a homegrown system for total laboratory automation (TLA) in the DMC core lab.
Now these pathologists are taking the next big step. They have launched a company to sell this automation solution to other laboratories. Under the name Integrated Laboratory Automation Solutions, Inc. (iLAS), based in Troy, Michigan, Neeley and his colleagues are prepared to sell what they describe as an “innovative, reliable, and cost-effective system of laboratory automation.”
Operational Since 2003
The iLAS automation system has operated at DMCUL since October 2003. “In almost three full years of 24/7 operation in a laboratory that performs more than 10 million billable tests per year, there has never been a significant stoppage of the automated line,” observed Neeley. “That speaks to the robust performance of this automated system.
“Simplicity, ease of operation, and reliability were intentionally engineered into our automation,” stated John Crissman, M.D, President and CEO of iLAS and former Chairman and Dean of the Wayne State University School of Medicine. “It is a flexible system. Any laboratory instrument with the capability to interact with automation can be connected to our automated line.”
The DMCUL core lab serves a consolidated laboratory system that includes eight hospitals, as well as physician office laboratories. When Neeley, Crissman, and fellow pathologist David Grignon, M.D, (who is an iLAS officer and current Professor and Chairman of the Department of Pathology at Wayne State University) decided to develop their own automation system, they already had an earlier automation template to draw upon. In the 1990s, Neeley had developed an effective, custom-built automation system at Meris Laboratories, Inc., located in San Jose, California.
With a capital expenditure of around $75,000, Neeley had used a standard commercial conveyor belt and off-the shelf bar code readers. this hardware was married to a custom software code that his lab team created. The combination produced a remarkably effective automation system at Meris. THE DARK REPORT was first in the nation to describe this development to the lab industry. (See TDR, August 2, 1996.)
“Before you ask, yes, I did apply the lessons learned from that automation solution to the design and operation of our current system here at Detroit Medical Center,” noted Neeley. “It is important for other laboratorians to understand what we did here. This system is designed around a few basic objectives.
“First is reliability,” he said. “We wanted an automated line that would perform tirelessly without malfunctions and simple and equipment failures. That’s essential in a laboratory performing 10 million billable tests per year.
“Second is simplicity. Automation must be easy for the lab staff to understand and to operate,” added Neeley. “Laboratory staff should find it intuitive and adaptable to what is needed to support the actual analytical process.
“We addressed simplicity in two ways,” he continued. “All hardware was made from standard, off-the-shelf industrial components. That makes it easy and fast to replace any part. Moreover, because these are standard components, their reliability is incredible. In our three years of operation, the automated line has never broken down and the only component failure that we had was a small sensor that died a natural death after several years of 24/7 operation.
“The other element used to simplify our automation is software,” said Neeley. “We wrote code designed specifically to meet all the needs of a clinical lab. It is software for labs, written by lab people. That means it has all the functions and options that you’ve always wanted in your lab, but could never get from a commercially- developed lab software product.
“Third, our automation is designed to hook up to almost any laboratory instrument,” he observed. “This is truly the open system that laboratories want. If the instrument has the capability to connect to automation, it can operate with our automation solution.
“Fourth, this automated line is modular. The basic module can connect as few as two or three instruments. The laboratory can add automation modules as needed to the line,” Neeley noted. “In our laboratory, we currently use a 100-foot automated line that connects 11 instruments.
“Fifth, we consider software to be our secret to success. As working laboratorians, we designed our automation software to expressly meet all the needs in our laboratory,” he said. “The software is compatible with autoverification software sold by many vendors so that labs can gain maximum efficiency from this automation system.”
“Sixth, our system is intentionally designed to handle all types and sizes of tubes,” continued Neeley. “For us to be competitive in the laboratory outreach market—which is 48% of our volume—we want to accept multiple sizes and types of tubes and be able to run them as primary tubes through our automation.”
The desire to use primary tube sampling reveals some of the inspired genius in Neeley’s thinking about laboratory automation. “In our laboratory, the philosophy is we will accept any tube submitted and run that tube as the primary specimen through our laboratory. That generates huge labor and cost savings even as it often contributes to a better quality specimen,” explained Neeley.
Pathologists Trained In Engineering
TWO OF THE THREE PATHOLOGISTS involved in the founding of Integrated Laboratory Automation Solutions, Inc. of Troy, Michigan completed training in engineering courses.
John Crissman, M.D. completed his undergraduate degree in mechanical engineering. William Neeley, M.D., upon his graduation from medical school, completed coursework in electrical engineering.
Both pathologists say that this engineering background was invaluable in helping them design a total laboratory automation (TLA) system that incorporated off-the-shelf industrial components, met the functional needs of the laboratory, and operated in a simple, straightforward manner.
In another coincidence, both Crissman and Neeley served as residents together. They did their residency at the Institute of Pathology at Case Western University Medical School in Cleveland, Ohio.
Running Short Specimens
“This also touches upon one aspect of lab automation that is seldom dis- cussed,” continued Neeley. “In most automated labs, a short specimen can’t be run on the line. Manual labor is required to correct the situation and that drives up the ongoing cost to run that automated line.
“This is not a small problem,” he added. “Up to 20% of our tubes are short draws. Our solution is pour the specimen in a cup and drop that cup in the primary tube so that instrument probes can sample from that specimen as it moves down the line.”
THE DARK REPORT has done a site visit to the DMCUL core laboratory and watched how this custom- designed automation solution works in actual operation. It has an economical footprint and the flexibility to add or subtract modules without major construction or renovation costs.
One impressive feature about this automation was its very low cost. “To install the track and hook up 11 instruments, we spent about $200,000 on hardware,” stated Crissman. “Software costs were additional. But the total was only about 25% of what a lab must normally pay. However because the system is tailored to our laboratory’s unique needs, we’ve seen substantial gains in productivity and functionality. In fact, we got payback from our automation in about six months!
20 Fewer MTs Needed
“For example, our high volume core lab now operates with 20 fewer medical technologists (MTs),” he explained. “We shifted these MTs to other areas of the laboratory where they do value-added work.
“Our automated line is also designed to accommodate STAT tests and respond to any work flow issue,” Crissman noted. “The software directs STAT tests to specific instruments that have the shortest queue. That means our MTs don’t have to manually handle STAT tubes.”
“Another thing our software does is constantly monitor all the functions along the automated line,” interjected Neeley. “If it determines there is a problem, it will automatically re-route affected tubes. There are two different scenarios for this function.
“The first deals with non-functioning instruments,” continued Neeley. “Because of our large sample volume, we have at least two or more identical instruments on the system. If one instrument fails, the software will automatically route samples to the other instrument and a flashing light will provide a visible alert that something is wrong with the instrument. The instrument may have failed or is about to run out of a reagent.
“The second scenario is when all the same type of instruments fail at the same time,” he explained. “Our software automatically routes affected samples to an aliquot or holding station. Here they are racked and can be easily routed to the instruments once they are returned to service. Otherwise, in many other automated systems, the samples will be placed in storage and a significant amount of labor would be required to locate and retrieve those samples.
“This has proved to be particularly useful when an instrument ceases to function,” stated Neeley. “The software automatically identifies that situation and re-routes affected tubes to the holding station. This feature has radically cut down on misdirected specimens.
“Further, we have a different philosophy about aliquoting,” observed Neeley. “To the maximum extent possible, we use primary tubes. In my view, aliquoting creates multiple ‘dead’ specimens and increases labeling and tube identification problems, as well as the incidence of running out of samples.
Primary Tube Labels
“Since we rely on the label of the primary tube, we avoid the problems of some automation solutions, which may track a tube by the rack number which carries it,” he offered. “Whenever rack and tube become separated without the lab automation recognizing this fact, it creates the opportunity for a lost tube or sample identification errors and the resulting confusion. Our reliance on primary tubes and the single label contributes to the sustained high performance of this automated system.”
During the past four years, Neeley has appeared at several laboratory meetings, including the Executive War College on Laboratory and Pathology Management, to discuss the automation effort underway at Detroit Medical Center University Laboratories. His thinking on automation was prominently featured in CAP Today Magazine in August 2002.
“It was the feedback and positive reaction from other pathologists and labortory directors that motivated us to offer this laboratory automation solution,” commented Neeley. “We are convinced that one reason why TLA has not found wider acceptance is because early generations of this product failed to meet both the needs and the expectations of many laboratories.”
“That is not the case at DMCUL, where this customized laboratory automation system has performed without major interruption in a high-volume core laboratory on a 24/7 basis for almost three years,” said Crissman. “It has generated substantial increases in productivity—at an extremely low cost. Its customized and flexible design supports our operational needs, as well as ongoing work flow redesign.
Raising Lab Performance
“These are all reasons why we believe that other laboratories will be interested in learning how they can use this lab automation system to achieve similar benefits,” added Crissman. “We are working pathologists and have created a TLA system that helps other pathologists and lab directors raise the performance of their laboratory operations.”
Crissman, Neeley, and Grignon are the newest entrepreneurs in the pathol- ogy profession. Although they are enthusiastic about the prospects for selling their automation solution to other laboratories, they also recognize the challenges. “We understand our place in the market,” said Neeley. “Our contribution to the laboratory is to simply provide tube transport and to not compete with instrument vendors.
“To achieve this, our automation system can readily connect to a host of analyzers,” he continued. “It is modular, handles a variety of tube sizes, and both the hardware and software can be quickly customized to our laboratory customer’s unique needs.”
“Best of all,” added Crissman,” is the fact that almost any laboratory can acquire this system at a rock-bottom cost. We estimate that a smaller laboratory can automate and connect between one and three analyzers for as low as $100,000. To do a system comparable to DMCUL, with up to 13 instruments, the total cost would run about $500,000.”
Crissman, Neeley, and Grignon have no misconceptions about the business challenge that awaits them. Pathologists and laboratorians are skeptical about new technology and new companies, and that is certain to be true about a young company offering a new automation system.
“With the three-year track record of this system in the Detroit Medical Center University Laboratory, we know we have a credible product,” predicted Crissman. “Because we can sell it at a price that is substantially below other automation options, we are hopeful that the combination will encourage interested laboratories to come by and ‘kick our tires.’ Moreover, it is likely that our first customer will get the deal of a lifetime, since we’d like to see this system in operation at other laboratories around the country.”
10 Principles of Laboratory Automation As Developed by William Neeley, M.D.
“CLINICAL PATHOLOGY IS MY PASSION. From the time I left medical school, I have spent my time in the clinical laboratory, working to improve the quality of test results and streamline work processes in the laboratory,” stated William Neeley, Medical Director of Labs at Detroit Medical Center University Laboratories in Detroit, Michigan.
In visits to laboratories under Neeley’s direction over the past 10 years, THE DARK REPORT has developed a list of 10 critical parameters that Neeley uses when he develops an automated work flow solution for his laboratory. In some ways, Neeley’s lab automation imperatives fly in the face of popular wisdom. But too often, it is this willingness to buck the popular wisdom that allows innovators to achieve a higher level of performance.
The following list of items is titled “William Neeley, M.D.’s 10 Precepts of Laboratory Automation” and represents our attempt to articulate the specific problem- solving strategies Neeley uses to create a reliable, effective, and low-cost total lab automation solution.
William Neeley, M.D.’s 10 Precepts of Lab Automation
1. SIMPLIFY ALWAYS, AT LOWEST COST—Automation, in its best configuration, should serve the primary needs of the laboratory. Avoid solutions which require work-arounds or alterations of “common sense” work flow. Avoid complexity.
2. EMPHASIZE RANDOM ACCESS AND RANDOM SPECIMEN—Establish a work flow that incorporates analyzers capable of random access and different specimen types. Make this flexibility serve improved work flow through the laboratory.
3. RUN TESTS IN REAL TIME, MINIMIZE BATCH TESTING—This improves quality, speeds time to result and evens out work load on the instruments (which reduces the need for analyzers with a higher throughput).
4. AVOID SECONDARY IDENTIFICATION LABELS ON THE TUBE—Maximize how the laboratory uses the primary label on the tube, from pre-analytical to analytical and post-analytical.
5. EMPHASIZE PRIMARY TUBE TESTING AND ALIQUOT AFTER CHEMISTRY AND IMMUNOCHEMISTRY TESTING—Minimize the need to create aliquots by running primary tubes through as many analyzers as possible before creating aliquots of the specimen.
6. AUTOVERIFICATION IS ABSOLUTELY ESSENTIAL—This greatly decreases labor and allows medical technologists (MTs) to focus on problems and value-added tasks, while avoiding countless MT hours spent looking at thousands of unremarkable results.
7. BE CAPABLE OF HANDLING MULTIPLE TUBE TYPES AND SIZES—Saves ongoing labor needed to handle the specimens. Particularly important for laboratories in the outreach business that want to compete on both service and lower cost.
8. STRIVE ALWAYS FOR THE LOWEST COST—And remember that few vendors will offer the most cost-effective answer to the need.
9. REDUCE COMPLEXITY—Simplify at every opportunity. Try to use “off the shelf” solutions in software and equipment whenever possible.
10.USE MIDDLEWARE TO GUIDE WORKFLOW—Be creative and use middleware whenever it can simplify processes. Use it to direct automation and to support lab staff.