TY - JOUR
AU - Hawker, Charles, D
AB - Abstract Background In most clinical laboratories, examination quality is considered excellent, whereas pre-/postexamination quality is an area for focused improvement. In our organization, 1 pre-/postexamination quality metric, namely, lost specimens, as tracked continuously for 27 years, has demonstrated steady improvement. During this period, many of our processes transitioned to highly automated effectors. Concurrently, we implemented behavioral controls and reengineered error-prone processes. We believe that this bilateral approach has conclusively lowered our lost specimen rates. Methods Using data spanning 27 years, we plotted the correlation between lost specimens and the implementation dates for 8 major phases of automation, as well as 19 process improvements and engineering controls. Results The lost specimen rate decreased nearly 100-fold. In Six Sigma terms, the 12 month moving average for lost specimens currently hovers at approximately 5.94 sigma, with 11 months at or better than 6 sigma. Although the combination of implementation of process improvements, engineering controls, and automation contributed to the reduction, automation was the most significant contributor. Conclusions The custom automation in use by our laboratory has led to improved pre-/postexamination quality. Although this automation may not be possible for all laboratories, our description of 19 behavior and engineering controls may be useful to others seeking to design high quality pre-/postexamination processes. Six Sigma, automation, process improvement, pre-/postexamination quality, engineering controls, behavioral controls Quality, as a conceptual and strategic goal for medical laboratories, has been an evolving directive since the late 1940s, when the first College of American Pathologists (CAP) proficiency survey was launched.1 However, health care quality in general garnered unfavorable national attention when the Institute of Medicine (National Academy of Medicine) issued a report in 1999, “To Err is Human: Building a Safer Health System”. This report was the first in a series, which estimated the number of patient deaths attributable to preventable errors in United States hospitals to be as high as 98,000 annually.2 The third report in the series was published in 2015, “Improving Diagnosis in Healthcare.” Supported by several decades of research, this report estimates that 10% of patient deaths are associated with diagnostic error.3 These figures are relevant to medical laboratories because the services they provide are an essential element in the development of diagnoses.4 Thus, reducing the incidence of laboratory error is vital to achieving the goal of removing the potential for patient harm. The absence of performance standards for laboratories prompted the formation of the CAP Inspection and Accreditation program in 1961, focusing on examination quality.1 As a result of increased attention, practices such as daily quality control, proficiency testing, and programs for statistical assessment of quality were institutionalized worldwide to address examination precision and accuracy. Also, in the United States, validations performed by laboratories (in accordance with the Clinical Laboratory Improvement Amendments of 1988 [CLIA]) and by manufacturers (through submission to the United States Food and Drug Administration [FDA]) have assured a high level of examination quality and reliability.5 Accordingly, in the modern laboratory, examination mistakes are a minor contributor to diagnostic error; most quality issues now reside in the pre-/postexamination arena.6,7 Pre-/postexamination error—unintentional, nonconforming events during pre- and postexamination stages—is the focus of this report. A lost specimen, occurring before or after testing, is an example of a pre-/postexamination error. From a benchmarking perspective, there has been meager attention given in the literature to lost specimen rates in the laboratory.8 Although these metrics are rarely published, our experience supports that most, if not all, medical laboratories track the number of lost specimens within the scope of their operations. Previously, we reported9 on our 25 year journey to reduce lost specimens within our organization. This most recent report extends that period by 2 years, lists additional improvements, and provides greater detail regarding the effects of our various approaches to error-proofing. Moreover, during the intervening 2 years, the steady improvement in our 12 month moving average has continued; expressed in Six Sigma terms, the moving average has improved from 5.86 sigma in 2016 to 5.94 sigma as of June 2018. The results of our 27 year experiment have been significant. Our outcomes were not the consequence of a single intervention but were multifaceted, iterative, and cumulative achievements that derived from all the methods described in this report. Materials and Methods We have defined lost specimens as those that were permanently misplaced between documented arrival at and documented departure from our facility. The data in this report do not include specimens lost en route by shipping and transport contractors or by referral laboratories. Specimens that are lost and later found fall into the temporarily misplaced category. Although this may be a narrow definition, it is useful for crafting the most appropriate response. When a misplaced specimen is found, the events leading up to the error, along with the location where it was found, are examined and safeguards put in place to prevent recurrence. Also, patient outcomes associated with the misplaced specimen event are assessed and areas of highest risk prioritized. The “misplaced” specimen reports provide a wealth of information for error-proofing the specimen management system, and the associated data sets provide an opportunity for assessing quality. A count or ratio of “lost” specimens, however, provides information about the effectiveness of misplaced specimen detection systems and the efficacy of the response system. The metrics presented in this report establish that we have reduced the rate of specimens that are declared to be lost, signifying progressively tighter process control, which is a proxy measure for quality and patient safety. Since 1991 we have tracked lost specimens using the same fundamental data elements year to year. Recognizing that no industry standards for lost specimen metrics and measurement units have been issued, we have chosen to discuss our results in terms of Six Sigma levels of performance. One of the more commonly used Six Sigma metrics is a ratio. In most cases, the numerator is defects (imperfections) or defective units (unacceptable outputs), and the denominator is process opportunities (product characteristics/production events) or output units (final, billable outputs) expressed in millions. The acronym DPMO generally refers to defects per million opportunities, and the acronym DPMU to defectives per million output units.10,11 Early in our measurement system development, we determined that using DPMO would amplify the denominator and could overinflate our achievement. For this reason, we chose to express our performance as DPMU—defectives (lost specimens) per million output units. In our measurement protocol, output units may be specimens or billed units (a reporting unit used for billing—ie, 1 metabolic panel is 6 performed tests but is 1 billed unit). Regardless of how these terms may be used in other contexts, defining the ratio consistently within our organization was important to establishing measurement reliability and repeatability over time. The Six Sigma method was developed in the manufacturing sector and was not adopted by health care organizations until much later.12 In the Six Sigma community, World Class Quality equates to 3.4 DPMO or less (99.99966% defect free work). Four-sigma performance translates to 6210 defects per million, and 6 sigma performance translates to 3.4 per million,12,13 an 1800-fold improvement in quality. The magnitude of this improvement target may seem overwhelming. However, our results demonstrate that 2 approaches—automation and engineered behavioral controls—working together, can yield remarkable results. Throughout this work, we used the Westgard Six Sigma Calculator13 to convert DPMU to sigma scores. Improvement Concepts Godfrey et al14 classified health care error-proofing solutions according to 5 progressively effective principles: elimination, substitution/replacement, facilitation, detection, and mitigation, with elimination being the most effective and mitigation the least effective error-proofing method. Following this construct, the first step toward error-proofing is to remove unneeded and error-prone elements from the design of the work. For example, automation eliminates error-prone tasks, and engineering designs and modifications eliminate risk. Once the work has been cleared of unnecessary tasks and risks, effectors are analyzed and, where possible, machines are used to replace humans. Automation is a reliable method for reducing error; however, it is costly. When automation is infeasible, improved work design and job aids are a viable substitution strategy for human proficiency and memory. The third principle focuses on behavior control through facilitation of human judgment, learning, skill development, and attention. The first 3 principles are error prevention devices; the remaining 2 principles moderate outcomes using error response systems and should be used only after the first 3 have been explored and leveraged. Those 2 moderation principles reduce the potential for patient harm once the error has occurred. Detection systems, such as audits and double checks, catch the error early and improve the probability of full recovery. The final option is mitigation of the effect of the error. Our approach to error management was 2-fold. First, it involved the use of automation to eliminate the risk of error and to execute repetitive and complex tasks using machines instead of humans. Second, it involved the use of process engineering and behavioral controls to facilitate and improve reliability, to detect error, and to mitigate potentially harmful outcomes. Table 1 summarizes our use of these 5 principles to manage lost-specimen potential. Table 1. How We Used the 5 Error-Proofing Principles Task and Risk Elimination  Eliminate or reduce variation using standard work processes and standardized materials and supplies  Eliminate tasks and hand-offs using automation and single-piece flow  Eliminate risk by eliminating hand-offs  Eliminate the risk of rolling tubes using a barrier—table edges, waste-bin lids, and equipment skirting  Eliminate shadows and the potential for error with the judicious use of lighting  Eliminate the risk of separation for blocks, slides, and scrolls by using a submission kit with a clear plastic case.  Eliminate missing parcels using a check out system for nonanalytic shipments received by the Specimen Processing department Substitution and Replacement  Replace human effectors for transporting, sorting, mixing, and storing specimens using automation  Replace human judgment and attention when sorting specimens by including final destinations in bar codes on specimen labels  Replace handwritten accountability logs using bar codes on employee badges  Replace human detection systems for abnormal conditions with automated, enhanced reports, to ensure timely responses Judgment, Knowledge, Skill, and Attention Facilitation  Facilitate memory and judgment when sorting and storing specimens using light-emitting diode (LED) lights on sorting bins (Sort-to-Light)  Facilitate judgment and memory using lost specimen procedures and checklists  Facilitate specimen handling and transport using standard transport tubes  Facilitate problem resolution and new employee integration using teams (pods)  Facilitate and accelerate learning with immediate, blame-neutral feedback systems when errors occur Abnormal Condition Detection  Detect premature storage events, mismatched records, and misdirected specimens using software enhancements and automated reports  Detect misplaced specimens using scheduled visual sweeps, clean line of sight, and adequate lighting  Use pattern analysis and big data sets to detect emerging risk—delayed shipments, delays between status updates in the laboratory information system (LIS), locations where misplaced specimens were found, etc  Confirm the submission of all required elements using a standard kit with a clear plastic case for tissues, slides, and DNA scrolls  Improve search processes by reducing the scope of the search, standardizing materials, and using color coding for waste  Use a double check system to ensure searches are comprehensive and complete Adverse Outcome Mitigation  Mitigate the severity of misplaced-specimen events through event analysis, checklist revision, color coding, and video monitoring Task and Risk Elimination  Eliminate or reduce variation using standard work processes and standardized materials and supplies  Eliminate tasks and hand-offs using automation and single-piece flow  Eliminate risk by eliminating hand-offs  Eliminate the risk of rolling tubes using a barrier—table edges, waste-bin lids, and equipment skirting  Eliminate shadows and the potential for error with the judicious use of lighting  Eliminate the risk of separation for blocks, slides, and scrolls by using a submission kit with a clear plastic case.  Eliminate missing parcels using a check out system for nonanalytic shipments received by the Specimen Processing department Substitution and Replacement  Replace human effectors for transporting, sorting, mixing, and storing specimens using automation  Replace human judgment and attention when sorting specimens by including final destinations in bar codes on specimen labels  Replace handwritten accountability logs using bar codes on employee badges  Replace human detection systems for abnormal conditions with automated, enhanced reports, to ensure timely responses Judgment, Knowledge, Skill, and Attention Facilitation  Facilitate memory and judgment when sorting and storing specimens using light-emitting diode (LED) lights on sorting bins (Sort-to-Light)  Facilitate judgment and memory using lost specimen procedures and checklists  Facilitate specimen handling and transport using standard transport tubes  Facilitate problem resolution and new employee integration using teams (pods)  Facilitate and accelerate learning with immediate, blame-neutral feedback systems when errors occur Abnormal Condition Detection  Detect premature storage events, mismatched records, and misdirected specimens using software enhancements and automated reports  Detect misplaced specimens using scheduled visual sweeps, clean line of sight, and adequate lighting  Use pattern analysis and big data sets to detect emerging risk—delayed shipments, delays between status updates in the laboratory information system (LIS), locations where misplaced specimens were found, etc  Confirm the submission of all required elements using a standard kit with a clear plastic case for tissues, slides, and DNA scrolls  Improve search processes by reducing the scope of the search, standardizing materials, and using color coding for waste  Use a double check system to ensure searches are comprehensive and complete Adverse Outcome Mitigation  Mitigate the severity of misplaced-specimen events through event analysis, checklist revision, color coding, and video monitoring View Large Table 1. How We Used the 5 Error-Proofing Principles Task and Risk Elimination  Eliminate or reduce variation using standard work processes and standardized materials and supplies  Eliminate tasks and hand-offs using automation and single-piece flow  Eliminate risk by eliminating hand-offs  Eliminate the risk of rolling tubes using a barrier—table edges, waste-bin lids, and equipment skirting  Eliminate shadows and the potential for error with the judicious use of lighting  Eliminate the risk of separation for blocks, slides, and scrolls by using a submission kit with a clear plastic case.  Eliminate missing parcels using a check out system for nonanalytic shipments received by the Specimen Processing department Substitution and Replacement  Replace human effectors for transporting, sorting, mixing, and storing specimens using automation  Replace human judgment and attention when sorting specimens by including final destinations in bar codes on specimen labels  Replace handwritten accountability logs using bar codes on employee badges  Replace human detection systems for abnormal conditions with automated, enhanced reports, to ensure timely responses Judgment, Knowledge, Skill, and Attention Facilitation  Facilitate memory and judgment when sorting and storing specimens using light-emitting diode (LED) lights on sorting bins (Sort-to-Light)  Facilitate judgment and memory using lost specimen procedures and checklists  Facilitate specimen handling and transport using standard transport tubes  Facilitate problem resolution and new employee integration using teams (pods)  Facilitate and accelerate learning with immediate, blame-neutral feedback systems when errors occur Abnormal Condition Detection  Detect premature storage events, mismatched records, and misdirected specimens using software enhancements and automated reports  Detect misplaced specimens using scheduled visual sweeps, clean line of sight, and adequate lighting  Use pattern analysis and big data sets to detect emerging risk—delayed shipments, delays between status updates in the laboratory information system (LIS), locations where misplaced specimens were found, etc  Confirm the submission of all required elements using a standard kit with a clear plastic case for tissues, slides, and DNA scrolls  Improve search processes by reducing the scope of the search, standardizing materials, and using color coding for waste  Use a double check system to ensure searches are comprehensive and complete Adverse Outcome Mitigation  Mitigate the severity of misplaced-specimen events through event analysis, checklist revision, color coding, and video monitoring Task and Risk Elimination  Eliminate or reduce variation using standard work processes and standardized materials and supplies  Eliminate tasks and hand-offs using automation and single-piece flow  Eliminate risk by eliminating hand-offs  Eliminate the risk of rolling tubes using a barrier—table edges, waste-bin lids, and equipment skirting  Eliminate shadows and the potential for error with the judicious use of lighting  Eliminate the risk of separation for blocks, slides, and scrolls by using a submission kit with a clear plastic case.  Eliminate missing parcels using a check out system for nonanalytic shipments received by the Specimen Processing department Substitution and Replacement  Replace human effectors for transporting, sorting, mixing, and storing specimens using automation  Replace human judgment and attention when sorting specimens by including final destinations in bar codes on specimen labels  Replace handwritten accountability logs using bar codes on employee badges  Replace human detection systems for abnormal conditions with automated, enhanced reports, to ensure timely responses Judgment, Knowledge, Skill, and Attention Facilitation  Facilitate memory and judgment when sorting and storing specimens using light-emitting diode (LED) lights on sorting bins (Sort-to-Light)  Facilitate judgment and memory using lost specimen procedures and checklists  Facilitate specimen handling and transport using standard transport tubes  Facilitate problem resolution and new employee integration using teams (pods)  Facilitate and accelerate learning with immediate, blame-neutral feedback systems when errors occur Abnormal Condition Detection  Detect premature storage events, mismatched records, and misdirected specimens using software enhancements and automated reports  Detect misplaced specimens using scheduled visual sweeps, clean line of sight, and adequate lighting  Use pattern analysis and big data sets to detect emerging risk—delayed shipments, delays between status updates in the laboratory information system (LIS), locations where misplaced specimens were found, etc  Confirm the submission of all required elements using a standard kit with a clear plastic case for tissues, slides, and DNA scrolls  Improve search processes by reducing the scope of the search, standardizing materials, and using color coding for waste  Use a double check system to ensure searches are comprehensive and complete Adverse Outcome Mitigation  Mitigate the severity of misplaced-specimen events through event analysis, checklist revision, color coding, and video monitoring View Large The chronological list in the next section divides the improvements we implemented into 2 sections—automation, which was executed in 8 stages and replaced human effectors for pre-/postexamination tasks, and 19 process-engineering/behavioral controls, which influence and facilitate human behaviors. Automation Stage A The original automated transport and sorting system, designed and built by MDS AutoLab, Toronto, Canada, began operation on November 17, 1998.15 With a maximum transport capacity of 2000 specimens per hour, it had 30 workstations for specimen accessioning, 3 automatic sorters for delivery to laboratory sections (a total of 90 lanes or sort-groups), and a fourth sorter for storage. Considerable manual handling and sorting was eliminated, along with the associated potential for error. Also, arrival at a sorter updated the laboratory information system (LIS) tracking status, posting the order to the pending report for a section. The immediacy of this process improved tracking and traceability of specimens. The fourth sorter for storing postexamination specimens replaced a manual PC-based process and improved traceability of stored specimens and employee accountability. Stage BA 2 story freezer (−20°C) automated storage and retrieval system (AS/RS), was designed, built, and installed by Daifuku America in December 2003. This system could hold as many as 2.349 million postexamination specimens in 5220 stainless steel trays. A robotic system loaded and unloaded trays from AS/RS storage shelves. A conveyor system moved trays in and out of the freezer through pneumatic access doors. An automated sorter (Motoman Robotics Division, Yaskawa America) retrieved individual tubes from trays when requested by employees. For each specimen, the precise storage location by tray number, row number, and column number was logged, along with the storage date/time and the appropriate discard date/time. The system required employees to scan their employee badge bar code, to ensure better tracking of all specimens entering and leaving the storage system. Stage CIn January 2004, the AutoLab track system was expanded to 72 workstations and 4 delivery sorters (a total of 120 sort groups). This expansion increased hourly throughput from 2000 to 5000. Stage DIn 2004, the original AutoLab storage sorter was replaced with 2 Motoman storage autosorters, each with a capacity of 1000 specimens per hour, that were connected directly to the automated transport and sorting system and automatically placed the specimens into storage trays. Software compared the bar code identification from each tube to the LIS record to check LIS status and to confirm that all ordered testing had been completed. If pending orders were detected, the specimen was redirected back to a delivery sorter. Specimens ready for storage were automatically placed into storage trays or racks on the sorter deck, eliminating 1 source of manual handling error. The movement of full trays from the automated sorter to temperature controlled storage still required manual handling. Stage EFour additional sorters were added in 2006, bringing the total to 8 and doubling the number of destination sort-groups (from 120 to 240). This increase in sorting capacity further decreased manual sorting and contributed to the steady reduction in lost specimens. Stage FSpecimens that cannot be transported and sorted by the main automation system (ie, critical frozen specimens and specimens in containers that do not meet the size and/or dimensional requirements for transport on the track) comprise 15% to 18% of the total volume. Before 2009, technicians manually sorted these specimens by reading an abbreviated destination on the specimen label. As might be expected, specimens were occasionally missorted, leading to higher potential for loss. A Sort-to-Light (S2L) system was designed, built, and installed by our automation and software engineers in 2009 to automate the sorting of these high-risk specimens. Initial implementation was followed by a software upgrade in 2010. The current iteration of the S2L system (Figure 1), implemented in February 2015, consisted of 160 bins contained in freezers, refrigerators, or room temperature (ambient) units. To use the system, a technician selected a color specific bar code reader and signed in by scanning the bar code on their employee badge. As each specimen label was scanned, a corresponding LED array, color coordinated to the linked bar code reader, illuminated around the correct sort bin for that specimen. As many as 9 technicians could sort specimens at the same time because each LED array could illuminate as many as 10 colors. (A video on the S2L system can be viewed at http://www.aruplab.com/testing/automation/videos.) Figure 1 View largeDownload slide The Sort-to-Light system. This system consists of 160 bins in freezers, refrigerators, or room-temperature units. Each bin is outlined by a light-emitting diode (LED) array that can be illuminated in as many as 10 colors. Technicians use a bar code reader, with a handle color matching one of the LED colors, to illuminate the destination bin for each specimen scanned. See also a supplemental video file. Figure 1 View largeDownload slide The Sort-to-Light system. This system consists of 160 bins in freezers, refrigerators, or room-temperature units. Each bin is outlined by a light-emitting diode (LED) array that can be illuminated in as many as 10 colors. Technicians use a bar code reader, with a handle color matching one of the LED colors, to illuminate the destination bin for each specimen scanned. See also a supplemental video file. Stage GA custom, high throughput storage autosorter, which was designed, built, and installed by Automated Tooling Systems (ATS), was implemented in 2010. The new autosorter sorted as many as 4000 specimens per hour into storage trays or racks, based on our specimen storage requirements (time and temperature). Although this system was installed primarily to increase automation capacity, specimens that were previously stored manually were shifted to automated storage, again contributing to the improved lost specimen rate. Stage HA new automated track system was installed in the spring of 2014. The new system employed the MagneMover LITE track (MagneMotion), 10 robotic sorters plus 7 other robotic machines (built by the automation-engineering department of our laboratory), and 3 automated thawing and mixing work cells (built by Motoman). This system increased transport capacity to approximately 5500 specimens per hour and sorting capacity from 240 sort-groups to more than 320. Also, the sort racks could be subdivided into different sort groups by rows within the racks, creating the potential for hundreds more sort groups and further minimizing manual sorting. Process Engineering and Behavioral Controls To reduce the number of false lost specimen notifications, in 1992, a multidisciplinary team developed and implemented a standard procedure detailing how the search for a missing specimen should be conducted, including a section-specific search checklist. The procedure also included criteria for when and how a specimen should be declared lost. The procedure required the labs to notify the Specimen Processing (SP) department when a specimen could not be located. SP staff conducted a duplicate, independent search within 24 hours of receipt of the request. Only after both searches had been completed and the associated checklists signed off would authorization be given to declare the specimen lost. A standardized transport tube with a false bottom and screw cap (Sarstedt #62.612.016) was evaluated, tested, and implemented in 1997 to increase the percentage of specimens that could be handled by the automation system. An added benefit was that processing staff members were more adept at handling standard sized tubes. As our clients began adopting various automation solutions of their own, use of our standard tube modestly diminished. However, the initial improvement and efficiency gains were significant. In 1997, we abandoned an assembly line system for processing specimens in favor of individual work stations. This change effectively eliminated processing hand-offs and lost specimen potential. In 1998, raised edges were installed around all specimen handling workstations, to provide a barrier that would control rolling tubes. Once proven effective, the design was incorporated into all work surfaces in the SP areas. In 1999, a programming change rejected specimens when submitted to storage if all ordered tests had not been completed. This software enhancement was also incorporated into the automation upgrade (stage D) of 2004. In 1999, all trash and biohazard waste receptacles were moved from directly underneath the primary specimen handling areas and were fitted with rounded covers that included a narrow, diagonal opening, to eliminate the potential for unintentional discard. These lids rendered the containers nearly fail-safe—any item placed in the trash bin was placed there intentionally. Training on the purpose and use of the containers ensured that staff members were aware of the potential for error and were coached on the proper method for discarding. In 2000, skirting material was attached around the base of all equipment in the various specimen handling areas, to ensure that specimen tubes could not roll under the equipment and become lost in the small space. This strategy was so successful that we currently require skirting as an integral part of the original engineering design of all new equipment in SP. During a study of causes of labeling error, video cameras were used to record processors labeling specimens. The team conducting the study noticed that the lighting intensities in the videos differed from station to station. The cause of irregular lighting turned out to be a design flaw: the way the lighting fixtures were aligned to the workstations created shadowed areas. The light fixtures were reinstalled in 2003, and changes were made to the type of lighting as well. The elimination of shadowed areas reduced the chance that misplaced specimens would remain undetected. In 2005, a multidisciplinary team with input from front line staff members found that the misplaced specimen search checklists had become obsolete over time as laboratory footprints, personnel, and equipment changed. The procedure was modified to include a periodic “review and revise” requirement. The new checklists were in use by early 2006. Scheduled visual sweeps by assigned SP staff required a documented inspection of floors, the areas under workstations, and the areas behind equipment. This process was incorporated into daily duties and into the codified search criteria for misplaced specimens in 2009. Data from nonconformance reports tracked the areas where misplaced specimens were found, to detect and correct emerging patterns. This information was incorporated into existing search checklists in 2010. As noted in item 9, the checklists were continually updated using the locations where lost specimens were found. To accelerate the integration of new employees and to shorten the time required to achieve efficacy, SP organized processing staff into 4-person teams called pods; the entire SP workforce had been converted to pods by 2010. Each pod had a Team Lead and immediate access to dedicated processing experts, quality staff members, and client support specialists. This approach established a whole team effort for processing specimens. Immediate feedback loops for new employees decreased training time, enabled timely error recovery, and boosted morale. These factors, in turn, contributed to steady improvement in the quality of work. A Paraffin Tissue and Extracted Nucleic Acid Transport submission kit was designed in 2011, to ensure that any combination of paraffin blocks, associated slides, and DNA scrolls stayed together during the processing step. With the kit, all material was submitted in a carton placed in a clear plastic casing, so that processing staff members could view the contents without removing them. The kit itself was larger than the slit in the trash can lids, eliminating the potential that all or part of the submission could be mistakenly discarded. The individual specimens within the kit are only removed once the kit has been received by the performing laboratory. Investigation into a misplaced shipment in 2011 led to changes in the way that incoming shipments from freight delivery services were handled and documented. The point of receipt was moved from the Shipping/Receiving area to the SP area, to eliminate hand-offs and minimize the scope of any missing shipment search. Also, the method for confirming receipt was changed from a batch process to a single-piece handling process. Scans of the bar codes on the shipping box and the badge of the receiving employee linked the 2 and automatically entered the information into an internally developed tracking database. Parcels destined for other, nonprocessing areas had to be “checked out” of this database, improving traceability for all shipments. SP management began to use big data in 2012 to build reports that alerted staff to the potential for missing specimens, containers, and shipments. Big data (large and complex data sets) were used to create reports that tracked shipments and their contents to facilitate reconciliation when the boxes were processed. Automating these “No Track Event” reports, as well as other internal reconciliation reports, has proven effective in detecting issues early in the process. In collaboration with our Security and Compliance departments, multiple video cameras were installed in 2013 in the SP work areas. Health Insurance Portability and Accountability Act of 1996 (HIPAA) and other requirements were considered and addressed with the resolution and placement of the cameras. Full time monitoring of the processing area strategically enhanced our ability to investigate and address nonconforming events. In 2014, a clean line of sight across and under all specimen handling areas was implemented. Electrical and communication wires were organized, tied, lifted, and secured off the floor; standard outlet strips and extension cords were prohibited; and standards for keeping work areas clear and clean were enforced. Along with end of shift reports, visual sweeps (item 10) in areas of concern were incorporated into the end of shift routine. In 2017. the big data reports (item 15) were enhanced to more accurately predict lost specimen potential. SP further developed the “No Track Event” report to send an email alert to the involved processing staff when the tracking status of a received specimen had not been updated for 2 hours. These emails contained instructions and a flowchart on how to perform a proper and thorough search. If the tracking status remained unchanged, specimens identified by this report were escalated to the departmental Team Lead. Further, if no one could ultimately account for specimens identified by this report, a similar alert was sent to the Problem-Resolution Specialist of the department for follow-up and retraining. The overall goal was to judiciously detect the potential for a misplaced specimen while involved staff members were still available for consultation, if needed. This mitigation device resulted in a significant drop in the number of lost specimens in the front end preexamination areas while keeping employees focused on patient care. In 2018, the handling of SP waste containers was modified to facilitate efficient specimen searches, if required. Unique green bags were introduced to differentiate SP waste from all other company waste. The bags were zip tied and labeled with the processing station identification and the date/time of discard. If a search for a missing specimen was required, the number of bags to be searched was limited to the color coded SP bags, resulting in more timely retrieval and in improved patient outcomes. Results The results of our 27-year experience are demonstrated in Figures 2 and 3, with Table 2 containing the key for these figures. We have plotted 2 separate sets of data (Table 3) using billed units as the output unit for the first set and specimens as the output unit for the second set. Figure 2 View largeDownload slide Average annual lost specimen sigma metrics plotted by year showing interventions (as listed in Table 2) with arrows showing the year of implementation. Sigma metrics were calculated13 using total lost specimens per million billed units enterprise-wide. Figure 2 View largeDownload slide Average annual lost specimen sigma metrics plotted by year showing interventions (as listed in Table 2) with arrows showing the year of implementation. Sigma metrics were calculated13 using total lost specimens per million billed units enterprise-wide. Figure 3 View largeDownload slide Monthly lost specimen sigma metrics plotted by year showing interventions (as listed in Table 2), with arrows showing the year of implementation. Sigma metrics were calculated13 using counts of specimens lost per million specimens only within the areas served by the automated systems. Figure 3 View largeDownload slide Monthly lost specimen sigma metrics plotted by year showing interventions (as listed in Table 2), with arrows showing the year of implementation. Sigma metrics were calculated13 using counts of specimens lost per million specimens only within the areas served by the automated systems. Table 2. Key to Automation Systems and Process Improvements or Engineering Controls Shown in Figures 2 and 3 Automation Improvements  A. Original AutoLab track system  B. Freezer automated storage and retrieval system (AS/RS)  C. Major expansion of AutoLab track system  D. Installation of Motoman storage autosorters  E. New sorters added to the track system, bringing the total from 4 to 8  F. Sort-to-Light system  G. New high capacity Automated Tooling Systems (ATS) storage Autosorter  H. New MagneMotion automation system Improvements Based on Process Engineering and Behavioral Controls  1. Standardized procedures for laboratory investigations of lost specimens, including a second search by Specimen Processing personnel and checklists  2. Standardized transport tube  3. Single-piece flow, elimination of hand-offs  4. Raised edges on all Specimen Processing department workstations  5. Not allowing automated storage system to store specimens until all ordered testing was complete  6. Trash bins repositioned out of the work area, slotted covers provided for trash bins  7. Skirts around the base of all laboratory equipment, to prevent specimens from rolling underneath  8. Improved lighting, to eliminate shadowed areas  9. Revised checklists for misplaced specimen searches  10. Visual sweeps in all Specimen Processing and automated preexamination areas at the end of each shift  11. Data-guided search protocols  12. Processing teams (pods) in Specimen Processing  13. Paraffin Tissue and Extracted Nucleic Acid Transport Kit  14. Badge and bar code scanning process to reconcile parcels in shipments from freight delivery services  15. Use of big data (large, complex data sets) with reporting functionality (data mining); using data to preempt misplaced specimen potential  16. Use of video cameras to monitor processing stations  17. Clean line of sight across and under all specimen handling areas  18. Enhanced big data reports for lost-specimen alerts and training  19. Use of tagged and labeled green bags for Specimen Processing waste Automation Improvements  A. Original AutoLab track system  B. Freezer automated storage and retrieval system (AS/RS)  C. Major expansion of AutoLab track system  D. Installation of Motoman storage autosorters  E. New sorters added to the track system, bringing the total from 4 to 8  F. Sort-to-Light system  G. New high capacity Automated Tooling Systems (ATS) storage Autosorter  H. New MagneMotion automation system Improvements Based on Process Engineering and Behavioral Controls  1. Standardized procedures for laboratory investigations of lost specimens, including a second search by Specimen Processing personnel and checklists  2. Standardized transport tube  3. Single-piece flow, elimination of hand-offs  4. Raised edges on all Specimen Processing department workstations  5. Not allowing automated storage system to store specimens until all ordered testing was complete  6. Trash bins repositioned out of the work area, slotted covers provided for trash bins  7. Skirts around the base of all laboratory equipment, to prevent specimens from rolling underneath  8. Improved lighting, to eliminate shadowed areas  9. Revised checklists for misplaced specimen searches  10. Visual sweeps in all Specimen Processing and automated preexamination areas at the end of each shift  11. Data-guided search protocols  12. Processing teams (pods) in Specimen Processing  13. Paraffin Tissue and Extracted Nucleic Acid Transport Kit  14. Badge and bar code scanning process to reconcile parcels in shipments from freight delivery services  15. Use of big data (large, complex data sets) with reporting functionality (data mining); using data to preempt misplaced specimen potential  16. Use of video cameras to monitor processing stations  17. Clean line of sight across and under all specimen handling areas  18. Enhanced big data reports for lost-specimen alerts and training  19. Use of tagged and labeled green bags for Specimen Processing waste View Large Table 2. Key to Automation Systems and Process Improvements or Engineering Controls Shown in Figures 2 and 3 Automation Improvements  A. Original AutoLab track system  B. Freezer automated storage and retrieval system (AS/RS)  C. Major expansion of AutoLab track system  D. Installation of Motoman storage autosorters  E. New sorters added to the track system, bringing the total from 4 to 8  F. Sort-to-Light system  G. New high capacity Automated Tooling Systems (ATS) storage Autosorter  H. New MagneMotion automation system Improvements Based on Process Engineering and Behavioral Controls  1. Standardized procedures for laboratory investigations of lost specimens, including a second search by Specimen Processing personnel and checklists  2. Standardized transport tube  3. Single-piece flow, elimination of hand-offs  4. Raised edges on all Specimen Processing department workstations  5. Not allowing automated storage system to store specimens until all ordered testing was complete  6. Trash bins repositioned out of the work area, slotted covers provided for trash bins  7. Skirts around the base of all laboratory equipment, to prevent specimens from rolling underneath  8. Improved lighting, to eliminate shadowed areas  9. Revised checklists for misplaced specimen searches  10. Visual sweeps in all Specimen Processing and automated preexamination areas at the end of each shift  11. Data-guided search protocols  12. Processing teams (pods) in Specimen Processing  13. Paraffin Tissue and Extracted Nucleic Acid Transport Kit  14. Badge and bar code scanning process to reconcile parcels in shipments from freight delivery services  15. Use of big data (large, complex data sets) with reporting functionality (data mining); using data to preempt misplaced specimen potential  16. Use of video cameras to monitor processing stations  17. Clean line of sight across and under all specimen handling areas  18. Enhanced big data reports for lost-specimen alerts and training  19. Use of tagged and labeled green bags for Specimen Processing waste Automation Improvements  A. Original AutoLab track system  B. Freezer automated storage and retrieval system (AS/RS)  C. Major expansion of AutoLab track system  D. Installation of Motoman storage autosorters  E. New sorters added to the track system, bringing the total from 4 to 8  F. Sort-to-Light system  G. New high capacity Automated Tooling Systems (ATS) storage Autosorter  H. New MagneMotion automation system Improvements Based on Process Engineering and Behavioral Controls  1. Standardized procedures for laboratory investigations of lost specimens, including a second search by Specimen Processing personnel and checklists  2. Standardized transport tube  3. Single-piece flow, elimination of hand-offs  4. Raised edges on all Specimen Processing department workstations  5. Not allowing automated storage system to store specimens until all ordered testing was complete  6. Trash bins repositioned out of the work area, slotted covers provided for trash bins  7. Skirts around the base of all laboratory equipment, to prevent specimens from rolling underneath  8. Improved lighting, to eliminate shadowed areas  9. Revised checklists for misplaced specimen searches  10. Visual sweeps in all Specimen Processing and automated preexamination areas at the end of each shift  11. Data-guided search protocols  12. Processing teams (pods) in Specimen Processing  13. Paraffin Tissue and Extracted Nucleic Acid Transport Kit  14. Badge and bar code scanning process to reconcile parcels in shipments from freight delivery services  15. Use of big data (large, complex data sets) with reporting functionality (data mining); using data to preempt misplaced specimen potential  16. Use of video cameras to monitor processing stations  17. Clean line of sight across and under all specimen handling areas  18. Enhanced big data reports for lost-specimen alerts and training  19. Use of tagged and labeled green bags for Specimen Processing waste View Large Table 3. Differences in Data for Figures 2 and 3 Variable Figure 2 Figure 3 Scope Enterprise-wide Areas served by automation Opportunities Billed units Specimens Interval Annually Monthly Rate Sigma score DPMU Variable Figure 2 Figure 3 Scope Enterprise-wide Areas served by automation Opportunities Billed units Specimens Interval Annually Monthly Rate Sigma score DPMU DPMU, defectives per million output units. View Large Table 3. Differences in Data for Figures 2 and 3 Variable Figure 2 Figure 3 Scope Enterprise-wide Areas served by automation Opportunities Billed units Specimens Interval Annually Monthly Rate Sigma score DPMU Variable Figure 2 Figure 3 Scope Enterprise-wide Areas served by automation Opportunities Billed units Specimens Interval Annually Monthly Rate Sigma score DPMU DPMU, defectives per million output units. View Large The data in Figure 2 represent enterprise-wide performance. Because any specimen leaving the various processing areas across our entire campus might travel to multiple testing locations, opportunities for this data set are counted as billed units. Overall performance was graphed as an annual sigma metric (average DPMU for the 12-month period from July to June). As shown in Figure 2, the 12-month moving average for lost specimens has steadily improved since 1997 and, as of June 2018, had reached a high point of 5.94 sigma. The data in Figure 3 are a subset of the data in Figure 2 and represent performance only in the areas served by automation. Automation handles specimens one at a time; accordingly, in Figure 3, 1 specimen equals 1 output unit. DPMU in Figure 3 was graphed monthly. The annual rate in Figure 2 eliminates much of the variation in Figure 3, yet plotting monthly performance shows a steady decrease in the lost specimen rate and progressively more frequent points falling below the line that denotes the 6 sigma level. The 2 figures together accurately depict the contributions of the 2 approaches to our error-proofing effort. Each of the automation stages was followed by sustained improvement (see stages A–H in Figures 2 and 3). Our findings suggest that reaching the 6 sigma target would not have been possible without automation. The configuration of our operations gave us the opportunity to test this premise. Our campus includes an academic hospital location with an onsite main laboratory and multiple satellite sites, as well as a reference laboratory located in an adjoining business park. The automation described earlier herein has only been implemented at the reference site. Since 2012 lost specimen data have been collected according to the last documented touch—or the last confirmed handling event for the missing specimen. This consistent counting method allowed us to compare metrics between the hospital complex and the reference laboratory. We used 2 independent data sets—reference facility and hospital complex—for the analysis. To ensure comparability, the DPMU numerator is a count of lost specimens and the denominator is billed units for both data sets. Two questions are relevant to the comparison: during the 72-month period represented by this data set, were the performance means for the 2 sites the same and was the rate of improvement equivalent between sites? To answer the first question (are the performance means the same?), we initially used a paired 2-sample t test for means. The results show that the DPMU means for the 2 data sets are not equal (t71 = 6.053; P ≤.05). For the first data set (the reference site), the mean (SD) was 6.39 (3.469) DPMU; for the second set (the hospital setting), that value was 12.46 (7.832). The volume of specimens handled at the reference site was significantly higher than that handled at the hospital complex. The reference facility has a sufficiently high denominator to accurately calculate DPMU each month; however, for the hospital group, the denominator for any given month was well below the per million units mark. A probable count of lost specimens per million units can be extrapolated from the actual ratio, but to do so would suggest an unwarranted level of confidence in our ability to predict outcomes. To address this issue, the sigma score to answer question 1 was calculated twice: once using the extrapolated ratio (noted in the previous paragraph) and once cumulatively each time the denominator reached the million units mark. The DPMU for the reference facility and the hospital site were, again, compared using a 2-sample t test for means. The second test confirmed that the means are not the same (t14 = 3.228; P ≤.05). For the reference site, the mean (SD) was 6.39 (3.469) DPMU; for the hospital group, it was 11.07 (5.011). For the second question (is the rate of improvement equivalent?), we plotted the DPMU rate over time for both sites (Figure 4). The rate of improvement at the reference site during the 72-month period was approximately 1 DPMU every 18 months (y = –0.0672x + 8.8503; P ≤.05). The rate of change for the hospital site, by contrast, remained relatively flat (y = 0.0069x + 12.206; P≥ .05). The poor fit of this linear regression line supports the conclusion that any change in the number of lost specimens was not a function of time—that change over time was not occurring predictably. Nevertheless, we have seen 7 defect free months in the past 6 years, confirming that engineering and behavioral controls have error-proofing value. Figure 4 View largeDownload slide Comparison of the rate of defectives per million output units (DPMU) change over time between the reference site and the hospital complex. Figure 4 View largeDownload slide Comparison of the rate of defectives per million output units (DPMU) change over time between the reference site and the hospital complex. Although there were marked differences between the work environment of the hospital laboratories and that of the reference laboratory, the aforementioned engineering and process improvements have been shared or were similar across all processing sites. If we assume all other influences equal, comparable outcomes could be expected. As demonstrated earlier herein, they were not. A difference of the aforementioned magnitude suggests a strong influence from another source. The strongest contenders for that source were automation and the academic hospital environment—a less structured, highly complex, and more chaotic environment. We have no evidence to support the hypothesis that processes performed in the hospital environment were poorly controlled. In fact, quality indicators monitoring the path of workflow at the hospital sites—ie, labeling error, turnaround time, corrected reports, and other indicators—show that performance was equivalent to or better than the enterprise-wide standard. Thus, we believe that automation was the differentiator. Following this line of reasoning, the onsite main laboratory at the hospital location is now moving toward total laboratory automation. The total of 8 automation stages were interleaved with 19 behavior controls and process reengineering projects. During the 27-year period, many of these projects were carried out simultaneously, making it difficult to assess the specific impact of any single engineering control or behavioral intervention. This happened, first, because the overall influence of these activities is compounding, each building on the success of the last, and second, because behavioral change requires time to establish equilibrium. Positive changes in performance metrics for these initiatives may not be immediately evident after the implementation, given that behavioral changes are more organic and interdependent. In fact, performance measures downstream of nonautomation initiatives have demonstrated progressive and incremental improvement over time, although they are less dramatic and less closely correlated with the point of change than those of automation initiatives. The following example illustrates the compounding effect of ongoing process improvement. One engineering control, namely, slotted lids on trash receptacles in SP (step 6), failed—a specimen fell into a trash bin. The video monitoring system (step 16), which was in place to enhance detection of nonconforming events such as this one and to inform the response when failures occur, showed the specimen falling into the bin through the protective lid. A subsequent search of trash bags was conducted. The search was time consuming and labor intensive but successful. The specimen was found and was viable for testing. Although a failure of this particular engineering control is extremely rare, this event demonstrated that such a failure is possible, and a mitigation response was developed. SP implemented the use of color coding in the form of green trash bags to distinguish SP trash bags from those generated elsewhere in the organization (step 19). These green bags were also tagged with workstation identification numbers and the date/time of discard to ensure a speedy recovery, should a similar event occur. Not only have each of the aforementioned interventions built the foundation for the next iterative improvement but they have also improved performance overall and enhanced workplace conditions in general. For instance, transitioning to single piece flow reduced hand-offs and improved throughput efficiency. Improved lighting reduced labeling errors, and implementing standardized transport tubes reduced inventory—the single tube replaced a variety of submission tubes. Cleaning up and standardizing work areas had the added benefit of improving workplace safety. The many big data reports developed to monitor the potential for lost specimens were used to improve outcomes for many other examinations, and pre-/postexamination processes. Finally, the automated systems we have described addressed pre-/postexamination quality issues on many levels, which resulted in improved efficiency, efficacy, turnaround time, and safety. We recognize that data collection protocols during the 27-year span have changed at times, as could be expected in an evolving measurement system. We have considered the implications of this limitation and have determined that the discrepancies were statistically insignificant. Describing our output quality in Six Sigma terms allowed us to normalize data during the 27-year period covered by this report. Discussion In 2000, Nevalainen et al16 collected and analyzed quality-indicator data from 3 clinical laboratories representing a cross section of medical laboratories in the United States and compared those data with error rates reported in 12 selected CAP Q-Probes (short-term, external, peer-comparison, and benchmarking studies17). The error rates for preexamination processes, expressed as a Six Sigma metric, ranged from 2.20 sigma (therapeutic drug monitoring timing) to 4.25 sigma (chemistry specimen acceptability). For examination processes, the sigma values ranged between 3.45 sigma (Pap smear rescreening for false negatives) and 3.85 sigma (proficiency testing). Finally, the sigma value for reporting error, a postanalytical process, was reported as 4.80 sigma.16 A decade later, Meier et al18 reported on a 12-year compilation of data from 7 selected CAP Q-Tracks (long-term studies for intrainstitutional comparison of quality performance measures17). The median results reported in 2011 demonstrated steady improvement in outcome measures during the past decade.18 However, overall, achieving a world class level of quality has been elusive in medical laboratories. As a result, in recent decades, increased attention has been directed to reducing the error potential for laboratory processes in general and for pre-/postexamination processes in particular. Pre- and postexamination processes are largely carried out using a series of manually executed steps that rely on judgment, attention, and memory, all of which are notoriously difficult to anticipate or control. Using only behavioral interventions to produce an exceptional level of quality may be beyond what is possible for many laboratory processes to accomplish. Standard process redesign tools and engineered control devices may enable a higher level of quality, but any process that relies on human effectors will exhibit a higher degree of variability and, thus, a lower level of reliability and quality. Eliminating human error requires automation, robotics, and software, accompanied by advanced process design.19,20 Although much of our automation has focused on tasks performed in the Specimen Processing area at the reference facility, viewed as a whole, the activities described earlier herein addressed aspects of every pre-/postexamination process enterprise-wide. Cooperation and collaboration between all work areas has been fundamental to holding the gains we realized with each progressive improvement. We recognize the importance of a whole organization, multidimensional approach to solving complex problems. This report provides impetus and encouragement for other laboratories seeking to improve pre-/postexamination quality. The custom automation implemented in our laboratory was a major contributor to our improved lost specimen metric. However, the experience of other medical laboratories suggests that the standard, easily obtainable automation available to most laboratories will lead to a higher level of quality than could be achieved using only behavioral and engineering controls.21 Our experience advocates for relentless, iterative effort using a wide range of methods and techniques to achieve world class levels of quality and to ensure the safety of patients. Support This work was fully supported by ARUP Laboratories, Inc. Personal and Professional Conflicts of Interest Bonny Messinger and Charles Hawker are retired from ARUP Laboratories, Inc. David Rogers is an employee of ARUP Laboratories, Inc. Acknowledgements We thank Jonathan Genzen, MD, PhD, Christopher M. Lehman, MD, and Anne Daley, MS, for their reviews of the manuscript and helpful suggestions; Olivia Carril, PhD, for assistance with statistical analyses; and D’Arcy Monforte, BA for assistance with the graphics in Figures 2, 3, and 4. Abbreviations CAP College of American Pathologists CLIA Clinical Laboratory Improvement Amendments of 1988 FDA United States Food and Drug Administration DPMO defects per million opportunities DPMU defectives per million output units LIS laboratory information system AS/RS automated storage and retrieval system S2L Sort-to-Light ATS Automated Tooling Systems SP Specimen Processing HIPAA Health Insurance Portability and Accountability Act of 1996 LED light-emitting diode References 1. Dorsey DB . Evolving concepts of quality in laboratory practice. A historical overview of quality assurance in clinical laboratories . Arch Pathol Lab Med. 1989 ; 113 ( 12 ): 1329 – 1334 . Google Scholar PubMed 2. Institute of Medicine Committee on Quality of Health Care in America . Kohn LT , Corrigan JM , Donaldson MS , eds. To Err is Human: Building a Safer Health System . Washington, DC : National Academy Press ; 2000 . 3. Committee on Diagnostic Error in Health Care, Board on Health Care Services, Institute of Medicine, The National Academies of Sciences, Engineering, and Medicine . Balogh EP , Miller BT , Ball JR , eds. Improving Diagnosis in Health Care . Washington, DC : National Academies Press ; 2015 . 4. Hallworth MJ , Epner PL , Ebert C , et al. ; IFCC Task Force on the Impact of Laboratory Medicine on Clinical Management and Outcomes . Current evidence and future perspectives on the effective practice of patient-centered laboratory medicine . Clin Chem. 2015 ; 61 ( 4 ): 589 – 599 . Google Scholar Crossref Search ADS PubMed 5. Bonini P , Plebani M , Ceriotti F , Rubboli F . Errors in laboratory medicine . Clin Chem. 2002 ; 48 ( 5 ): 691 – 698 . Google Scholar PubMed 6. Howanitz PJ . Errors in laboratory medicine: practical lessons to improve patient safety . Arch Pathol Lab Med. 2005 ; 129 ( 10 ): 1252 – 1261 . Google Scholar PubMed 7. Plebani M . Errors in clinical laboratories or errors in laboratory medicine ? Clin Chem Lab Med. 2006 ; 44 ( 6 ): 750 – 759 . Google Scholar PubMed 8. Shahangian S , Snyder SR . Laboratory medicine quality indicators: a review of the literature . Am J Clin Pathol. 2009 ; 131 ( 3 ): 418 – 431 . Google Scholar Crossref Search ADS PubMed 9. Messinger BL , Rogers DN , Hawker CD . Use of automation and process improvement to achieve a Six Sigma Level of Nonanalytic Quality . J Appl Lab Med. 2017 ; 2 ( 1 ): 86 – 91 . Google Scholar Crossref Search ADS 10. Kubiak TM , Benbow DW . The Certified Six Sigma Black Belt Handbook, Second Edition . Milwaukee, WI : Quality Press , 2009 . 11. iSixSigma Dictionary . https://www.isixsigma.com/dictionary/. Accessed December 10, 2018 . 12. DelliFraine JL , Wang Z , McCaughey D , Langabeer JR 2nd , Erwin CO . The use of Six Sigma in health care management: are we using it to its full potential ? Qual Manag Health Care. 2013 ; 22 ( 3 ): 210 – 223 . Google Scholar Crossref Search ADS PubMed 13. Westgard J . The Six Sigma calculators . https://www.westgard.com/six-sigma-calculators.htm. Accessed December 10, 2018 . 14. Godfrey AB , Clapp TG , Nakajo T , et al. Application of healthcare-focused error proofing: principles and solutions directions for reducing human errors . Paper delivered at the ASQ World Conference on Quality and Improvement Proceedings ; Seattle, WA ; May 16–18, 2005 . 15. Hawker CD , Roberts WL , Garr SB , et al. Automated transport and sorting system in a large reference laboratory: part 2. Implementation of the system and performance measures over three years . Clin Chem. 2002 ; 48 ( 10 ): 1761 – 1767 . Google Scholar PubMed 16. Nevalainen D , Berte L , Kraft C , Leigh E , Picaso L , Morgan T . Evaluating laboratory performance on quality indicators with the six sigma scale . Arch Pathol Lab Med. 2000 ; 124 ( 4 ): 516 – 519 . Google Scholar PubMed 17. College of American Pathologists . Quality management programs . https://www.cap.org/laboratory-improvement/quality-management-programs. Accessed December 10, 2018 . 18. Meier FA , Souers RJ , Howanitz PJ , et al. Seven Q-Tracks monitors of laboratory quality drive general performance improvement: experience from the College of American Pathologists Q-Tracks program 1999–2011 . Arch Pathol Lab Med. 2015 ; 139 ( 6 ): 762 – 775 . Google Scholar Crossref Search ADS PubMed 19. Astion ML . Getting realistic about reliability: an interview with Dr. Roger Resar [online newsletter] . Lab Errors Patient Safety. 2007 ; 3 ( 4 ): 1 – 6 . 20. Resar RK . Making noncatastrophic health care processes reliable: Learning to walk before running in creating high-reliability organizations . Health Serv Res. 2006 ; 41 ( 4 Pt 2 ): 1677 – 1689 . Google Scholar Crossref Search ADS PubMed 21. Yu H-YE , Lanzoni H , Steffen T , et al. Improving laboratory processes with total laboratory automation [epub ahead of print] . Lab Med. 2018 ; doi: https://doi.org/10.1093/labmed/lmy031 . © American Society for Clinical Pathology 2019. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Automation and Process Re-engineering Work Together to Achieve Six Sigma Quality: A 27-Year History of Continuous Improvement
JF - Laboratory Medicine
DO - 10.1093/labmed/lmy081
DA - 2019-04-08
UR - https://www.deepdyve.com/lp/oxford-university-press/automation-and-process-re-engineering-work-together-to-achieve-six-E6TzRSl1Lu
SP - e23
VL - 50
IS - 2
DP - DeepDyve
ER -