Medical Device Link.

AUTOMATION FOR DIAGNOSTICS

The AutoMate 800 sample processes by Beckman Coulter Inc. (Photo courtesy Beckman Coulter Inc.)

A recent landmark study conducted by the U.S. Institute of Medicine confirmed this in part. The study report states that medical errors, such as errors or delays in diagnosis, failure to employ indicated tests, use of outdated tests, prescription of the wrong medication, and mistaken identity, contribute to more than 1 million injuries and around 100,000 deaths in American hospitals annually.¹ The Centers for Disease Control and Prevention seconded that report’s placing of some blame on testing laboratories. CDC identified the elimination of laboratory errors leading to adverse patient outcomes as one of its Seven Healthcare Safety Challenges.

In an effort to reduce sample-mishandling errors by eliminating manual processes, many laboratories have turned to automation. Automation has proven helpful to them generally by increasing testing accuracy and specifically by reducing the number of preanalytical processes performed manually that often generate errors.

This article describes recently developed preanalytical process automation technology that can significantly improve the laboratory work flow while minimizing the occurrence of testing errors attributable to sample handling. But first it examines the market circumstances that make automating clinical laboratories so important.

Market Pressures on Laboratories

Figure 1a. (click to enlarge) Laboratory sample-preparation process steps before automation. This map of potential process errors and outcomes is based on a Power Processor automation system (Beckman Coulter Inc.; Fullerton, CA).

Many laboratory mistakes are associated with manual processes performed by technologists, particularly centrifugation, aliquotting, and specimen retrieval, which have a high potential for human error (see Figure 1a). Manually processing samples not only too often results in errors, it also can entail long turnaround times. Having to wait for test results leads to delays in patient diagnosis and treatment and can prolong the hospital length of stay (LOS).^2,3 In addition to feeling pressure to reduce processing errors and help shorten patient LOS, laboratories are challenged by growing test volumes and a shrinking pool of technologists.

Figure 1b. (click to enlarge) Laboratory sample-preparation process steps after automation. This map of potential process errors and outcomes is based on a Power Processor automation system (Beckman Coulter Inc.; Fullerton, CA).

Patient Length of Stay. In today’s cost-driven environment, hospitals face intense pressure to reduce LOS. When a hospital is operating at or near capacity, reducing the time each patient occupies a bed frees space for new admissions and hospital transfers, thus increasing revenue opportunities. Many hospitals therefore put pressure on laboratories in turn. They demand that testing processes be completed quickly and efficiently, as well as accurately, in order to improve their own LOS metrics.

In a nonautomated laboratory, test turnaround time is subject to variations in individual work habits, speed, and error propensity, and suffers from insufficient staffing due to the current labor shortage affecting laboratories. A laboratory outfitted with integrated automated systems has fewer of these limitations. It can process tests more quickly, more reliably, and virtually without error.

Shortage of Technologists. A large number of accredited programs in clinical laboratory science and medical technology have closed in recent years owing to reduced enrollment, and decreasing numbers of students are graduating from each remaining program. This trend has resulted in an ongoing shortfall of qualified laboratorians that will continue well into the future. In the United States, the current 14% vacancy rate in medical technologist positions that is critical now will only worsen after 2008, when sizable numbers of professionals working in the field are expected to begin to retire. With laboratory technologists in short supply, testing facilities have to find a way to make sure their skilled workers use their time efficiently—by dedicating it to tasks requiring skill.

Increasing Test Volume. Many U.S. testing laboratories are facing increasing volumes or work even as they struggle with understaffing. This further drives the need for faster, more-efficient processing techniques. Laboratories’ larger workloads are due partly to the aging of the population. However, the situation is exacerbated for some laboratories by their own efforts to expand their workloads with outreach business. As a result, laboratories are seeking new ways to process their routine tests quickly and efficiently. Those that take on additional specialty work have effectively fewer resources available to handle routine testing and so need some form of assistance to process even their everyday workloads.

Efficiency through Automation

As laboratories strive to address these market challenges, they are often forced to do more testing with fewer resources. Many are also realizing that the greatest potential for process improvement lies in their pre- and postanalytical testing processes—and these laboratories are turning to automation for help.

Indeed, laboratory automation has much to contribute to the improvement of medical care. Improvements often begin with relatively simple steps. In the case of medical laboratories, that would be automating sample handling in the preanalytical process, which is the most labor-intensive process in the laboratory and the stage of analysis during which most errors occur. Comprehensive—also called total—automation systems encompass automation of the analytical and postanalytical stages of processing, as well.

Automated bar coding and specimen handling make positive patient identification more certain and can reduce the incidence of lost or improperly labeled test tubes. Automated analyzers can systematize front-end test preparations and improve the quality of critical test reruns. They employ software that conforms with the requirements of the regulatory agencies that oversee issues involving test accuracy and patient safety. The software even reports test results to those agencies.

Other benefits of laboratory automation include faster testing turnaround—allowing more-rapid treatment of critically ill patients—and more-precise, consistently reliable test results. Laboratories that have automated most of their testing have documented reductions in the kinds of processing errors that historically have endangered patient health.

Case Histories

One laboratory that has responded successfully to the challenging trends enumerated above by implementing automation is the Oklahoma University Medical Center (OUMC) in Oklahoma City. After establishing automation in its chemistry operation, the OUMC facility was able to eliminate 10 per diem employees—the equivalent of five full-time employees (FTEs)—representing an estimated annual compensation of $50,000–$55,000 each, including benefits. This dropped the staffing level in the core laboratory from 65 people to 60.

Also, automating systems enabled the facility to close its two short-term assessment and treatment (STAT) laboratories and transfer the technologists working in them to the core laboratory. Besides saving roughly $1.5 million per year, this move freed up technologists’ time, which enabled the core laboratory to bring in 22 tests that were previously sent out to a reference laboratory. All of these gains were realized without the need to hire additional staff.

Despite the fact that a significant number of FTE-equivalents had been eliminated, OUMC’s laboratory experienced a 35% increase in overall testing volume, seeing the number of annual tests rise from 3.5 million to 5 million. Before automation, such a jump would theoretically have required a 35% increase in technologists on hand, or approximately 18 additional FTEs.

OUMC has had similar success in reducing turnaround time, thus improving the center’s ability to provide more-timely diagnosis and shorten patient stay. After automation, its laboratory began getting trustworthy, reliable immunoassay results back in 40–50 minutes instead of 3–4 hours. Representative time savings were about 80%. Likewise, the facility received electrolytes after a wait of only 30–40 minutes rather than 2 hours, a time savings of about 75%. These accelerations carry the potential to reduce patient LOS.^2,3

Figure 2. (click to enlarge) Comparison of annual increases in revenue at John T. Mather Memorial Hospital (Port Jefferson, NY) before (2001 to 2002) and after (2004 to 2005) the hospital instituted automation and the use of decision-making hospitalists. Source: Denise L. Uettwiller-Geiger and Wayne Shattes, “Lab Automation Speeds Information to Clinicians for Faster Clinical Decisionmaking, Reducing Patient Length-of-Stay,” a poster presented at the American College of Healthcare Executives, 2007.

Another hospital to draw a correlation between automation and shorter LOS is John T. Mather Memorial (Port Jefferson, NY), a 248-bed community hospital. Mather Memorial recently addressed its LOS problem by hiring staff members, called hospitalists, who could make clinical decisions more efficiently than primary care physicians.⁴ This action, combined with the institution’s prior investment in laboratory automation, resulted in significantly reduced LOS. Between 2004 and 2005, admissions to the hospital increased by 315, from 11,775 to 12,090, while patient care days dropped by 2372, from 79,808 in 2004 to 77,436 in 2005. The increased admissions boosted revenues by nearly $1.6 million (see Figure 2).

Another benefit of automation for Mather Memorial has been the hospital’s ability to absorb a 100% increase in emergency department visits since 1996. Furthermore, the healthcare facility has exhibited better compliance with regulatory and accreditation standards regarding patient safety.

Refinements in Automation Technology

Taking advantage of developments in automation technology, laboratories can implement new strategies of preanalytical sample handling that will preserve the integrity of the samples and also speed up the postanalytical, as well as the preanalytical, phases of diagnostic testing. Improving the speed and consistency of these specimen-handling processes to reduce turnaround time helps to minimize bottlenecks in work flow.

Figure 3. The AutoMate 800 sample-processing system from Beckman Coulter Inc. (Fullerton, CA).

One new technology has focused on effectively and flexibly managing the sample tube from the time it enters the laboratory until the time it is thrown away. This is embodied in the AutoMate 800 sample processor from Beckman Coulter Inc. (Fullerton, CA), a compact system that represents the capabilities of today’s automated systems in helping laboratories maintain sample integrity, increase efficiency, reduce manual processing errors, accommodate higher testing volumes, and avoid expanding staff (see Figure 3). Providing examples of engineering innovation, it offers a number of original technologies designed to maximize the ability of automation to optimize the process flow.

On-Demand Centrifugation. One drawback of most automation systems is that they lack the flexibility to handle the wide variety of sample tubes with which a laboratory deals. Some tubes—those used for complete blood counts, for example—require no centrifugation, while others might have been spun before reaching the laboratory. Each tube has its own centrifugation and testing requirements. But because laboratories often do not want to spend time managing each centrifuge load individually, technologists can have trouble keeping up. However, with the new Beckman Coulter processor, tubes can be programmed by test type, with some getting no centrifugation and others being subjected to one- or two-cycle spins in the built-in refrigerated, self-balancing centrifuge.

The system’s standard centrifuge protocol can be used for chemistry, immunochemistry, and routine coagulation factor tests. If a higher-quality sample is needed for the coag test, for example, the system will automatically run the tube through the centrifuge once using the standard protocol, then read it again as it comes out and recognize that it needs a second spin cycle in order to provide the sample quality desired. Or, it can take a complete blood count sample and, without any special handling, direct it to bypass the centrifuge altogether.

The instrument loads and runs the centrifuge on an as-needed basis, providing on-demand centrifugation for chemistry, immunoassay, and coagulation testing. Following user-defined protocols, it automatically loads the centrifuge, balances the buckets, runs the centrifuge, and unloads the tubes when completed. The system accommodates a variety of tube sizes.

Automated Fluid-Level Sensing. When a blood specimen is centrifuged in the sample processor, it separates into two principal components: a compacted portion of red blood cells, and a portion of serum or plasma. The system’s sensing device reads both levels simultaneously to save time. It measures the fluid externally, through as many as three layers of labels. In addition, it detects the presence or absence of a cap on the tube, and stores that information for future reference.

After testing, when the tubes are sorted for final storage, the system measures how much sample remains and notes that information for future recall. Then, if a physician wants to order a test to be rerun, the laboratory can immediately find out how much of the original sample is left and confirm whether a rerun from the original tube is possible.

STAT Sample Override. Laboratories commonly must deal with STAT samples, prioritizing them above routine samples for rapid turnaround on tests whose results are critical. The latest Beckman Coulter instrument features a dedicated STAT input area. STAT samples placed there enter the queue ahead of already-input routine samples.

As soon as the drawer closes, the automated system loads and processes the STAT samples, going back to its routine work when it is finished with the urgent processing.

Intelligent Aliquotting. Whether the laboratory needs to aliquot just enough sample fluid for one test or enough for many, the system’s capacious 4-ml aliquot tip automatically draws and transfers the fluid volume necessary for all aliquot samples. It creates up to eight aliquots per primary tube, and uses computerized prioritization to ensure that the most critical samples are created first should the sample volume be minimal. This optimizes performance and accelerates throughput. Plus, automated tube labeling helps to eliminate manual sample-preparation errors in this task area, and ensures faster, more-accurate secondary-tube preparation.

Dynamic Programming. Laboratories dealing with tubes arriving from outreach laboratories have found it a nuisance that their automated systems could not tell them whether the tubes had already been centrifuged and did not have to be run again. However, new systems give them an efficient way to determine whether the centrifuge can be bypassed. They offer flexible system programming to combat this problem, processing each tube as necessary.

Most automated systems have a fixed definition and position for every sample-tube tray which technologists cannot change without shutting down the entire system and conducting a reconfiguration. Systems like the AutoMate 800, however, have a unique bar code at the bottom of each rack that instructs the system regarding the tubes’ identity and how they are to be treated. All of the tubes that come into the facility from outreach laboratories go onto such coded racks. Likewise, laboratories can identify a rack for pediatric samples; the bar code would alert the system not to decap or aliquot those sample tubes.

This capability additionally enables the system to shift automatically with the laboratory’s daily work flow—accommodating more chemistry samples in the morning, for example, and more immunochemistry samples in the afternoon. These racks can be added and removed at any time, and as frequently as necessary, in order to keep up with workload fluctuations.

Smart Decapping. On the newer automated systems, tube caps are selectively removed for open- or closed-tube sampling, depending on the needs of the analyzer that will run the tube. Removed caps are automatically dispensed into the biohazard bag, which optimizes laboratory safety by minimizing employee exposure to biohazardous aerosols.

Postanalytical Sample Storage and Tracking. The AutoMate 800 system remaps samples for storage and catalogs tube location and residual sample volume. Or, if the testing is completely done, the system will sort the tube to a storage rack to be placed into a refrigerator, freezer, or room-temperature storage facility, as appropriate for the test. The routing of the sample is determined by the duration of storage, the temperature conditions, or the user-defined destination. At the end of the tube’s useful life, the system automatically prompts the laboratory to dispose of it.

Conclusion

Automation is a crucial resource in today’s clinical laboratory environment, helping to offset a number of healthcare industry trends—specifically, laboratory errors caused by manual mishandling of samples, a worsening labor shortage in laboratories, and the drive to shorten hospital stays for patients. An automated solution provides faster, more-efficient, and more-accurate testing. This leads, in turn, to faster diagnosis or treatment, fewer errors in critical test results, shorter turnaround times, higher physician satisfaction with laboratory performance, and a better work environment for laboratorians.

Any one of these improvements would contribute to patient healthcare, but the latest laboratory automation systems represent a significant step forward in test quality from the perspective of laboratory technologists, doctors, and patients alike.

Ron Berman is vice president of the Automation and Information Systems Business Center at Beckman Coulter Inc. (Fullerton, CA). He can be reached at rberman@ beckman.com.	Paul Ashton is a senior staff systems engineer for automation at Beckman Coulter Inc. He can be reached at pjashton@ beckman.com.	Jeff Quint is worldwide product manager for automation at Beckman Coulter Inc. He can be reached at jfquint@ beckman.com.