The role of total laboratory automation in a consolidated laboratory network.
A decision was made to purchase a total laboratory automation (TLA) system for the Core Laboratory. The rationale to do so was based on the need to decrease total laboratory costs in the network and to improve the turnaround time (TAT). The potential test volume from the original nine hospital sites sent to the Core Laboratory was estimated as 2.4 million CPT-4 tests annually with >1.5 million (5000/day) that could be analyzed on the TLA system. Outreach testing would add another 300 000 tests (CPT-4) to the testing coming from the hospitals. The test menu considered for the TLA would include basic hematology, coagulation testing, routine chemistry, and immunochemistry.
Materials and Methods
We opted to purchase the Clinical Laboratory Automated System (CLAS; see Fig. 1) marketed by the Roche Diagnostics. The CLAS is an open TLA system that brings a sample to a specified point on an interconnected track. The analysis systems that are attached to the CLAS track bring the samples into the processing systems and perform the requested tests. Result distribution is performed through a direct connection from the analysis system back to the host or sent back to the host through the CLAS master controller (MC).
The system consists of three major lines: chemistry, hematology, and coagulation. The procedure for sorting and preparing samples is as follows: (a) The original sample sorter (OSS) receives and sorts patient samples as they arrive in the core laboratory. The OSS can process up to 500 specimens per hour. (b) Within the OSS, the samples are sorted into racks consisting of five separate positions. The sorted tubes within the racks are then sent to one of three major analysis lines. The samples on the chemistry line are centrifuged on the centrifuge unit unless the specimens come to the laboratory already centrifuged at the outlying hospitals. (c) Subsequent to centrifugation, the decapper removes the caps from the sample tubes. (d) The aliquoter reads the rack identification and sample identification and aliquots enough serum/plasma for online analysis (A line), or to prepare the sample (B line) for the other sections of the facility not connected on the CLAS. All of the above functions are dependent on the highly effective integration of bar coding technologies on the specimen tubes. In the case of the NS-LIJ Health System, we opted to standardize our system utilizing Code 128.
[FIGURE 1 OMITTED]
In addition to efficient bar coding of the tubes, the NS-LIJ Health System also standardized the specimen types to ensure compatibility on the automation system and to help reduce iatrogenic anemias present within the system. All microtainer tubes (bullet containers) received in the system are processed locally along with STAT tests in the hospitals.
The CLAS then directs the individual aliquots or parent tubes to the instruments for analysis or sorts them into the correct channels for processing within the laboratory. Once aliquoting is completed and the aliquots are directed to the proper area for testing or processing, the parent tubes are placed into a stockyard for sample storage. Samples for hematology and coagulation are sorted at the OSS and directed to their proper destinations.
Detection systems for minimal volumes, clots, and possible interfering substances are located either at the aliquoter, where the user will receive an error for a failed aliquot, or at the analysis systems, where the user will receive errors for potential hemolysis, icterus, or lipemia. Thus, potential problems are detected, and the users are notified that a potential problem exists.
Specimen tracking is accomplished on the CLAS at the MC. Through an inquiry function, the user is able to track the sample to a specific rack and a specific position within that rack. At different intervals throughout the day, the laboratory assistants take the parent tubes and archive them within the laboratories' specimen storage system, which is locally controlled by a desktop personal computer. Should the clinician call for additional testing, the laboratory assistants retrieve the parent tube and place it on the aliquoter for aliquoting and distribution to the appropriate testing station.
The analyzers attached to the CLAS include three Sysmex 9500SE hematology analyzers, one Sysmex SP-100 slide maker, one Hitachi 917 analyzer, and two Hitachi 747 analyzers. Roche Diagnostics markets all of these instruments. Two MLA-1800 coagulation analyzers, although not attached to the CLAS, are in close proximity to the track. In October 1999, we added two Elecsys 2010 immunochemistry systems to the CLAS.
The CLAS is interfaced to the PathNet Laboratory Information System (LIS; Cerner). Considering the importance of the LIS interaction with the CLAS, the CLAS-LIS interfacing is complex. It consists of one bidirectional interface for the CLAS and three bidirectional interfaces for the chemistry analyzers on the CLAS. An additional unidirectional interface is needed to control the OSS sorting on the CLAS. The reason for these separate interfaces involves redundancy; should a component fail, we have ensured that result uploads and order downloads still occur.
On the CLAS, the download of orders and upload of results occurs at the MC. Usually, the analyzers attached to the system query the MC for the work required. Once the instruments complete these assays, the analyzers upload that information back to the MC, which then sends the results back to the host. The necessity for the three additional bidirectional interfaces is for redundancy on the analyzers. Should the MC fail, the A/B directional switch can be turned, allowing the analyzers to communicate back to the host directly.
Automated hematology has an additional four separate bidirectional interfaces. The interface between the CLAS and the hematology workstation is a "handshake". This means that the CLAS only delivers the samples to the Sysmex HST-402 system. Once the samples arrive at the HST-402, the hematology system communicates directly with the host for order queries and result uploads.
Should a problem occur with the host, the automated system should be considered to be down, and departmental disaster plans will be initiated until the system comes back on-line. The interface maintenance on the system is the responsibility of the vendors, Cemer and Roche Diagnostics. These two parties are responsible for addressing any interface problems between the MC and the LIS system. The operational review of these complex interfaces falls on the LIS team based within the Core Laboratory. When a user reports a problem that is tracked back to a interface problem, the LIS team member initiates a call to the appropriate vendor for resolution.
The overall success of the NS-LIJ Health System Core Laboratory and the RRLs that have been implemented depends on the seamless operation of the LIS system with the various instrument systems and order entry systems in all of the laboratory facilities. Therefore, the cooperation of the instrument vendors with the LIS vendor and the NS-LIJ Health System personnel is paramount for the transparent operations of order entry, specimen analysis, and result delivery. These complex systems, when correctly sized and properly implemented, enable the clinical and anatomical pathology services to realize the goals set forth for the entire system.
Because the coagulation instruments are not directly attached to the CLAS, interfacing is accomplished at the local instruments.
INSTALLATION AND TESTING
The CLAS was designed and customized in the manufacturing plant in Japan. This process took ~3 months. The installation occurred over a 4-week period in November 1997. The system occupies ~2800 square feet in a large open laboratory. Training of key operations took an additional 3 weeks after installation and validation. During the time the CLAS was being manufactured, installed, and tested, we were designing and constructing the Core Laboratory. Core Laboratory design began in September 1997 and was finalized in January 1998.
VOLUME OF TESTING
We began processing clinical specimens on March 23, 1998. The delay from installation to clinical testing was attributable to LIS verification, New York State inspections, and other issues unrelated to the CLAS. On a typical weekday, the CLAS processes ~3000 specimen tubes: 900 hematology, 600 coagulation, and 1500 chemistry and reference testing. It should be noted that this number will increase dramatically after the next group of hospitals comes on-line to send us their laboratory testing volume. Peak activity occurs between 0700 and 0930, during which time [greater than or equal to] 1000 specimens are processed through the CLAS. The system is operational from 0530 until 2330 weekdays and from 0530 to 2300 on weekends.
LABOR REQUIREMENTS FOR THE CLAS
A total of 6.0 full-time equivalents (FTEs) are allocated to operate the CLAS during the day shift. This includes the manager of the general laboratory, 1.0 FTE (laboratory technician) to load samples onto the OSS, and 4.0 FTEs, (medical technologists), who verify results and deal with any complex issues that arise daily. It should be noted that additional FTEs are required to perform the coagulation testing, the hematology slide reading, and transport of samples to the special chemistry area and to the reference testing triage area. With the implementation of the Elecsys systems on-line, the demand for technical help in transporting specimens to special chemistry will be dramatically reduced. During the second shift, 3.0 FTEs are allocated to process specimens through the CLAS: 1 FTE (laboratory technician) to load the samples onto the OSS, and 2.0 FTEs (medical technologists) to verify results and handle any problems. On weekends, a total of 7.5 FTEs are required to process samples through the CLAS: 4.0 FTEs during the day shift, and 3.5 FTEs on the evening shift. The labor requirements necessary to process the samples through the CLAS were configured based on the results of the studies performed that demonstrated specimen arrival times and total sample volumes received throughout the week. Thus, we have maximized our efficiency so that we have the required technical and support personnel present when the samples arrive for processing.
At the NS-LIJ Health System Laboratories, we have been very pleased with the CLAS as a whole. Since going "live", we have experienced only 16 h of unanticipated downtime over 16 months. These incidents have been attributable to either mechanical or electronic problems. The CLAS requires ~10 min of daily downtime to purge the previously day's results from the MC and bleed the air pressure valves on the OSS and aliquot sample sorter.
Quarterly, the system requires ~14 h of routine maintenance. This maintenance is scheduled during off hours and slow periods throughout the day when specimen processing will not be affected. It includes greasing of transport motors, vacuuming dust from components, cleaning of the transport track and the belts within the track, cleaning and greasing of online centrifugation systems, and performance of a system check designed to assure that each component is functioning according to specifications.
Our experience with CLAS demonstrates that the system is easy to operate. This is attributable to the "turnkey" status of the system, i.e., selected preanalytical and analytical components that have been designed and pretested to function on the CLAS line. This turnkey approach, which is favored by Roche, enabled our system vendor to rapidly install the system and begin the system's extensive validation processes required by regulatory agencies. "Open" architecture systems are more complex. These automation systems require many different mechanical and virtual interfaces for foreign equipment and hence increase the amount of time needed to bring these systems on-line.
The preanalytical components are operated and maintained by laboratory assistants. The laboratory assistants are trained for 16 h over 2 days. During this training, they learn to research and fix any system alarm that is generated from a component or system line controller. Analyzer operations and maintenance are areas reserved for the technical staff along with all analysis procedures.
The overall maintenance responsibility for the CLAS is under the direction of the Automated Laboratory Manager, who is responsible for ensuring that system problems that are beyond the scope of the laboratory assistants are resolved in a timely fashion. The Automated Laboratory Manager is also responsible for monitoring each downtime event and escalating the call as needed to maintain the system's overall TAT. By removing the technical staff from routine tasks, we have placed the staff in positions where they can handle any complex tasks relating to the analytical processes. The technical staff also handle any "backups" that may occur, e.g., failed quality control that causes analytical components to need user intervention or calibration.
To effectively service patients in the network hospitals, the NS-LIJ Health System Core Laboratory needed to guarantee TAT thresholds. Core Laboratory TAT is defined from the time the sample is drawn, verified as in-lab, transported via courier to the Core Laboratory, placed on the automation system, and the results reported. The threshold established for the Core Laboratory was <4 h. This goal was established in the early phases of the planning process and was one of the very important basic premises required to justify the automation system and the consolidated network.
The threshold set for RRL TAT is 30 min for the abbreviated menu left in the hospitals (RRL test menu listed in Table 1). The Consolidated Laboratory Network is dependent on the RRLs to efficiently triage the samples for transport to the Core Laboratory. In addition, the LIS, is essential for the overall success of the automation and consolidation. When the LIS fails, we initiate manual backup plans that can impact the TAT needed by our clinicians for the delivery of patient results.
A key concern for TLA is to have it appropriately sized for our volume. With a maximum throughput of 500 tubes per hour and an 18-h uptime, the theoretical limit for output is 9000/day. This value exceeds our present volume of 3000/day and even doubling of that to 6000/day. However, a morning rush of 1000 specimen tubes over a 2-3 h period could stress the system. At this point in time, we feel that the system is adequately sized for our present volume and a possible doubling of our volume.
The NS-LIJ Health System Laboratories has met and exceeded the TAT thresholds set for samples coming from the facilities interfaced. Within the Core Laboratory, our average TAT for all assays on the CLAS is 1.3 h for 1.5 million tests. With the addition of the Roche/Hitachi Elecsys 2010 system to the CLAS, the same TAT should be realized for immunochemistry testing, e.g., thyroid studies, fertilities, and anemia panels.
Financially the TLA purchase and implementation requires a carefully well-thought out evaluation (1). Six concerns can impact projections:
* Financial savings depend on what the baseline was for the system, i.e., the less efficient the laboratory was before the consolidation and TLA implementation, the greater the savings.
* The savings are very dependent on testing volume. In the case of the NS-LIJ Health System Laboratories, our automation was only cost justified after the introduction of the total testing volume for the system. Any additional increases in volume once the TLA system has been installed can usually be absorbed without significant cost increases.
* Savings for the hospitals within the system are dependent on the current salaries and benefits for the personnel who will be replaced.
* Within the NS-LIJ Health System, once the decision was made to move toward TLA with a centralized Core Laboratory and surrounding RRLs, the number of FTEs within the system was reduced through attrition and re-engineering of people into other positions required by the hospitals or the Core Laboratory. Since the Core Laboratory was conceived and subsequently implemented, no employees within any of the hospitals have lost their positions. This policy has allowed the continued professional growth of the clinical and anatomical laboratory personnel and enabled the system to start realizing expense reductions.
* Autoverification algorithms designed to enhance post-analytical testing can impact the automation system and consolidated network by enabling more efficient use of technologist time. Additional to autoverification algorithms, the NS-LIJ Health System Laboratories have implemented reflexive testing on a limited basis, only after testing approval has been received from each of the clinical pathologists for each of the RRLs. Examples of these algorithms are high total creatine kinase (CK) reflexing to perform a CK-MB, or abnormal urinalysis reflexing to perform a microscopic examination. It is important to note that within the NS-LIJ Health System Laboratories, all reflexing rules are documented as part of the policies and procedures for both the Core Laboratory and the RRLs. Policies and procedures within the NS-LIJ Health System Laboratories stipulate that analyte retesting is performed when the technologist working in a specific workstation deems it necessary.
* System hospital savings are dependent on the timeframe in which the hospitals are implemented onto the LIS system. Once a hospital is implemented, its personnel can start accessioning locally and send the routine and esoteric samples to the highly efficient Core Laboratory for processing. This approach also allows further reductions to be made in the overall TAT for the STAT samples that remain within the hospitals. This overall reduction in TAT may lead to enhancements in patient care and a possible reduction in overall patient stays by providing critical information to clinicians in a timely fashion.
* The attachment of additional workstations onto the TLA can also represent additional savings that were not originally projected, e.g., for coagulation and immunochemistry.
During our exploratory phase, reduction in labor expenses had been forecasted. The vendor guaranteed that we would not need more than 17.7 FTEs to run samples on our system for the contracted volume of 3775 specimens over 24 h. To date, we require 9 FTEs to process 2000 specimens. Please note that we are still not fully implemented. When our consolidated system is fully implemented, we have projected labor savings of $2.7 million for the system.
We have noted both a reduction of expenses and system-wide standardization because of consolidation efforts and the integration of TLA within our Core Laboratory. Both the Core Laboratory and the RRLs have enjoyed a reduction in expenses. This is attributable to the renegotiation of reagent and equipment contracts based on the volume of testing being performed by both the Core Laboratory and the RRLs. In addition, because our consolidated network has moved to performing testing on a cost per test basis, we are able to make decisions regarding our expenses and efficiently budget month-to-month operations.
Our standardization efforts have enabled further reductions in expenses, both fixed and qualitative. We are able to negotiate terms for quality-control material, general laboratory supplies, and other critical needs of our system such as courier costs, gasoline, and insurance. Because of standardization, we need only write one procedure manual that can be applied to all facilities and have one set of reference and critical values. This has enabled the member hospitals to be a valuable part of all decisions and activities.
Some issues that can cloud the overall financial picture involve the political atmosphere in which the automated laboratory and consolidated network operates. Hospitals tend to "downsize" in anticipation of TLA; thus, the baseline choice for comparison can make huge differences against projections. The best overall financial scenario is as follows. (a) The consolidated system has inefficient laboratories already in place. (b) Testing volume for the intended automated laboratory is sufficient to warrant the inclusion of automation. (c) Salaries and staffing for the technical areas of the laboratory are high before automation is implemented. (d) The system is willing to move toward autoverification algorithms. (e) The system will incorporate one LIS system for all laboratory functions. (fl The system agrees to send as much work to the automated laboratory as possible.
We have optimized the use of clinical laboratory space. In the Core Laboratory, our planning efforts yielded the greatest impact on workflow because of our careful consideration to facility planning and detail. Within the RRLs, the space vacated by the full-service laboratories has been utilized by other departments that can bring valuable services to the community.
We conclude that the TLA has been successful within the NS-LIJ Health System Laboratory network. This platform is not the only instrument platform that will yield cost reductions and patient care enhancements (2). Although TLA was our only option in 1995 when we decided to automate and consolidate, today two other types of automation platforms are available, "modular automation" and "task-targeted automation".
Modular automation provides for processing presorted samples on modules that are defined by the end user to automate preanalytical processing, analytical processing, or both within a single attached piece of equipment. The advantages of modular systems are its flexibility, relative ease of implementation (interfacing, validation, and facility planning), and ability to automate components of the testing process (3). Therefore, implementation costs are reduced and the integration time frame is markedly accelerated, allowing for rapid return on investment. Modular automation enables the user to choose the right processing modules that will provide maximum efficiency and cost benefits.
Workstation consolidation is the key to the success of this automation platform. Modular automation may reduce the number of instruments in the laboratory and assist with the elimination of the need to split samples for different workstations. It also may simplify operations, reduce TAT, and enhance efficiency while enabling capacity expansion.
Task-targeted automation is designed to automate a certain process, e.g., primary tube sorting, aliquoting and labeling, and secondary sorting. Primarily, these systems have been utilized most effectively within the preanalytical areas of the laboratory (4). Automating the preanalytical sections can represent savings of up to 60-70% of the total overall laboratory costs. Task-targeted automation can also be used to assist the laboratory with archiving functions.
As with any automated system, careful attention needs to be placed on the objectives of the user and how to achieve these goals. Factors that need to be considered carefully are the types of tests to be performed, volume of testing for each module, sample arrival times, current workstation landscape, and future needs. When these questions are solved, the user can make the right decisions relating to module configuration and preanalytical applications.
In conclusion, our system has benefited from the consolidation efforts and the implementation of the TLA system. The actual dollar savings are predicated on the remaining hospitals coming live and their willingness or ability to make the necessary staff reductions. Although our system remains in the growth phase, we have realized our efficiencies in TAT for those hospitals brought live.
(1.) Statland BE, Stallone R0, Seaberg RS. Building a case for a consolidated laboratory network. Clin Lab Prod 1999;28(June): 35-7.
(2.) Seaberg RS, Statland BE, Stallone R0. Planning and implementing total laboratory automation at the North Shore-Long Island Jewish Health System Laboratories. Med Lab Obs 1999;31(June):46-54.
(3.) Fiedler G, Gabrek L. Modular[TM] system at work. J Assoc Lab Automation 1999;4(March):56-8.
(4.) Demiris CH, Ciment PR. Task targeted automation a cost effective and easily justified approach to automation for improving the efficiency of the clinical laboratory. J Assoc Lab Automation 1998; 3(November):64-6.
RICHARD S. SEABERG,  ROBERT O. STALLONE,  * and BERNARD E. STATLAND 
 North Shore-Long Island Jewish Health System, 10 Nevada Dr., Lake Success, NY 11042.
 University of Minnesota Law School, 203 Bank St. SE, Minneapolis, MN 55414.
 Nonstandard abbreviations: NS-LIJ, North Shore-Long Island Jewish; RRL, Rapid Response Laboratory; TLA, total laboratory automation; TAT, turnaround time; CLAS, Clinical Laboratory Automated System; MC, master controller; OSS, original sample sorter; LIS, laboratory information system; FTE, full-time equivalent; and CK, creatine kinase.
* Author for correspondence. Fax 516-719-1062; e-mail Roberts@nshs.edu.
Received December 7, 1999; accepted February 11, 2000.
Table 1. RRL test menu. Chemistry Acetaminophen (a) Alcohol (a) Ammonia Amylase BUN Calcium Carbamazepine (a) CK-MB estimate CK-MB mass assay [Cl.sup.-] C[O.sub.2] CPK Creatinine CSF glucose CSF protein DBILI Digoxin Glucose [beta]-HCG quantification Ionized calcium (a) [K.sup.+] Lactate (a) Lithium (a) Magnesium [Na.sup.+] 0smolality Phenobarbital Phenytoin P[O.sub.4.sup.2] Primidone (a) Procainamide (a) Quinidine (a) Salicylate (a) TBILI Theophylline Valproic acid (a) Hematology Bleeding time CBC (b) with differential D-Dimer ESR Fibrinogen Fluid counts, body & CSF Hematocrit Platelet counts PT/aPTT Thrombin time Urinalysis Urine microscopics Other Blood gas analysis Crystals Culture set-ups (emergency) Gram stains (a) May vary by hospital. (b) CBC, complete blood count; ESR, erythrocyte sedimentation rate; BUN, blood urea nitrogen; CSF, cerebrospinal fluid; PT, prothrombin time; aPTT, activated partial thromboplastin time; CPK, creatine kinase; DBILI, direct bilirubin; HCG, human chorionic gonadotropin; TBILI, total bilirubin.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Clinical Chemistry Forum|
|Author:||Seaberg, Richard S.; Stallone, Robert O.; Statland, Bernard E.|
|Date:||May 1, 2000|
|Previous Article:||Development of standards for laboratory automation.|
|Next Article:||Developing an automation concept that is right for your laboratory.|