CME/SAM. Determination of Turnaround Time in the Clinical Laboratory

Transcription

1 Clinical Chemistry / Turnaround Time in a Clinical Laboratory Determination of Turnaround Time in the Clinical Laboratory Accessioning-to-Result Time Does Not Always Accurately Reflect Laboratory Performance Brie A. Stotler, MD, MPH, and Alexander Kratz, MD, PhD Key Words: Turnaround time; Accessioning time; Quality assurance; Quality improvement; Clinical laboratory CME/SAM Upon completion of this activity you will be able to: define total laboratory turnaround time and describe 1 common method for monitoring this metric. discuss limitations of the commonly used accessioning-to-result time as a metric for the laboratory turnaround time. list preanalytical steps that may be excluded by the accessioning-toresult turnaround time metric. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p 753. Exam is located at Abstract Laboratories often focus on computer-generated components of turnaround time (TAT) reports that capture time from sample accessioning to reporting of results, and use this indicator to determine performance. This study assesses limitations of accessioning-to-results times and details ways in which to evaluate laboratory-controlled TAT. Samples were sent via pneumatic tube and times from arrival to accessioning were determined. Staffing was increased and the delay between sample arrival and accessioning was measured again. Significant delays were seen between specimen arrival and accessioning, which were not captured with computer-generated TAT reports. When TAT was calculated to include these delays, the TAT goal was not achieved. Increasing the number of employees significantly decreased delays. Laboratories must ensure that TAT reports encompass all laboratory-controlled parts of the testing processes. Analysis of causes for discrepancies between computer reports and clinician perceptions, combined with targeted measurements and well-designed interventions, can decrease TAT and improve service. Clinical laboratories take pride in being extremely data driven. Many quality indicators are continuously monitored, analyzed, and used to allocate resources and improve service. These quality indicators include the turnaround times (TATs) necessary to report laboratory results to clinical staff. The total TAT for laboratory assays includes the entire interval from ordering of the test to the clinician s awareness of the result (ie, brain-to-brain ). It consists of the intervals from order placement to specimen collection, as well as the time necessary for transport to the laboratory, accessioning in the laboratory, centrifugation, aliquoting, additional preanalytic steps if necessary, transport times within and between laboratories, analysis time, the time after completion of analysis until result verification, and the time it takes for the clinical team to be informed of the result. 1 Lack of full control over phlebotomy and specimen transport make it difficult for the laboratory to address delays caused in the preanalytic stages that take place outside the laboratory. 2 In addition, ordering and collection times are not always fully documented for all samples, and it is currently not possible to determine when clinicians become aware of most laboratory results. Thus, it is impossible for laboratories to use the total TAT as a quality assurance measure. Accessioning by the laboratory is defined as ordering the assay in the laboratory information system (LIS) or as scanning the barcoded samples as received. Result time is defined as the release of finalized results into the hospital s electronic medical record. In contrast to ordering and collection times, the accessioning and result times are always well documented for all samples in the LIS; therefore, it is easy to obtain detailed reports on the accessioning-to-result 724 Am J Clin Pathol 2012;138: Downloaded 724 from

2 Clinical Chemistry / Original Article component of the total TAT. Thus, most laboratories focus on the accessioning-to-result time in their TAT reports. Although this approach does not encompass the total TAT, it is generally accepted as the best representation of those elements of the TAT that the laboratory controls. The current report documents an important limitation of the accessioning-to-result verification time, describes a way to address this limitation, and shows how careful attention to feedback from clinical staff and proactive data collection can be used to improve clinical laboratory performance. Specifically, this study quantifies the interval between sample arrival in the laboratory to sample accessioning in the LIS, and discusses how this delay, often not captured with standard TAT reports, affects laboratory service. Materials and Methods The study was conducted on the Columbia University Medical Center (CUMC) campus of the New York-Presbyterian Hospital (NYPH; New York, NY) academic medical center. The center has more than 1,000 inpatient beds, separate adult and pediatric emergency departments, multiple specialized intensive care units, and very active surgical, transplant, and cardiac services. A core laboratory provides chemistry, immunochemistry, and hematology laboratory results to the entire medical center and to off-site outpatient clinics and physicians offices. To provide appropriate TATs for timecritical specimens, the core laboratory has separate receiving and preanalytic areas for urgent (stat) and routine samples. The stat preanalytic area receives approximately 1,200 samples per day. Most samples are delivered via pneumatic tube; a small percentage of samples are hand-delivered by messenger. During weekday day shifts, 3 employees are responsible for opening pneumatic tube carriers, sorting bags containing samples, opening the bags, accessioning the samples in the computer system, centrifuging the samples, and delivering the samples to the appropriate analytic areas. The laboratory s policy is to separate bags received in the stat receiving area into 3 priorities. Highest priority samples are received from the emergency department in red bags. Second priority samples are those with electronic stat orders from other hospital units received in transparent bags. Lowest priority samples are stat samples received with a paper requisition. Study Design To assess the interval between sample arrival in the laboratory and accessioning in the LIS, a laboratory supervisor sent patient samples via the pneumatic tube system from the routine preanalytic area of the laboratory to the stat preanalytic area. Samples were sent every hour on weekdays, and the samples identifiers and the time of dispatch were recorded. The actual transport time from the routine area to the stat area through this method was established to be less than 1 minute. Each pneumatic tube carrier contained 3 plastic bags; each plastic bag represented one of the 3 processing priorities described before. Three bags per carrier were sent to assess whether the laboratory s prioritization scheme yielded different types of delay. Technical and clerical staff members were not informed that a study was in progress. Data Collection and Analysis The times of accessioning and release of finalized results were retrieved from the LIS (Mysis, Tucson, AZ). The time of dispatch of the carrier was used as the receipt time. Accessioning was defined as either scanning the samples into the LIS (for samples received with a barcode label) or as ordering of the assays by the laboratory clerk in the LIS (for samples received with a paper requisition). The accessioning time was used as a substitute for the time the bag containing the samples was opened. The interval between receipt and accessioning was assumed to be because of delays in the stat preanalytic area. The time the final results were released into the electronic medical record ( result times ) for complete blood counts (CBCs) and basic metabolic panels (BMPs) was recorded. Intervention To determine if increased staffing would significantly improve TAT, the staffing of the stat preanalytic area was increased from 3 to 5 employees on weekdays during the day shift. The 2 additional employees were specifically instructed to concentrate on opening specimen bags and accessioning specimens. Results Discrepancy Between the Laboratory s TAT Data and Perceptions of the Clinical Staff The NYPH-CUMC core laboratory s goal is to report the results of at least 90% of all stat CBCs and BMPs within 1 hour of receipt in the laboratory. Because the accessioning of specimens, but not their arrival in the laboratory, is documented in the LIS, computer reports of the accessioningto-result time are used as a substitute for the receipt-to-result time to track TAT. TAT data collected by the laboratory using the accessioning-to-result time indicated that our quality goals were generally met. However, ordering physicians felt that it often took the laboratory significantly longer than 1 hour to complete stat tests, even in cases in which the clinical staff personally sent the sample to the laboratory via the pneumatic tube system or hand-delivered the sample. We decided to investigate the root cause for this service quality discrepancy between our data and the assessment of the clinical staff. We suspected that delays between arrival of the specimens in the Downloaded from Am J Clin Pathol 2012;138:

3 Stotler and Kratz / Turnaround Time in a Clinical Laboratory Table 1 Interval Between Stat Specimen Arrival and Accessioning by Order Priority * Priority 1 Priority 2 Priority 3 All Bags Original Increased Original Increased Original Increased Original Increased Staffing Staffing Staffing Staffing Staffing Staffing Staffing Staffing No. of Samples Time, min Mean Range Median SD P value <.01 <.01 <.01 <.01 SD, standard deviation. * Priority 1 samples are those with the highest priority, received from the emergency department in red bags. Priority 2 includes stat samples with electronic orders. Priority 3 samples are received with a paper requisition. Percent of Samples % 75% 39% 88% 48% 94% 10 min 15 min 20 min Original staffing Increased staffing 66% 99% 30 min Arrival Time to Accessioning Time 84% 100% Figure 1 Percentage of samples accessioned by specific times after arrival. 1 h laboratory and accessioning could cause the delays reported by the clinical staff; such delays would not be reflected in the existing TAT reports because they measured the time from sample accessioning, not from sample arrival. Interval Between Specimen Arrival in the Stat Area and Sample Accessioning To determine the interval between sample arrival and accessioning, samples were sent from the routine preanalytic area of the laboratory to the stat receiving area of the laboratory via the pneumatic tube system. Because transport time in the pneumatic tube system from the routine area to the stat area is less than 1 minute, the arrival time in the stat area was known. There were significant delays between sample arrival in the stat receiving area and accessioning Table 1 and Figure 1 and Figure 2. Mean delays were 32 minutes for all 3:50 3:21 Send to Accession Time 2:52 2:24 1:55 1:26 0:57 0:28 0:00 8:07 8:15 8:40 10:04 10:04 10: Am J Clin Pathol 2012;138: Downloaded 726 from 10:20 11:05 11:10 11:40 11:46 12:00 12:02 12:40 12:47 12:58 Time of Day 13:10 13:24 13:29 14:37 14:39 15:00 15:05 15:19 15:30 15:31 15:55 16:09 16:21 16:38 17:19 Figure 2 Delays in accessioning individual stat samples by time of day. The x-axis is the time of day in military time; the y-axis is the time (in hours:minutes) between specimen arrival and accessioning.

4 Clinical Chemistry / Original Article specimen groups; half of all specimens were delayed for more than 22 minutes. One-third of all specimens were delayed for more than 30 minutes. On some days, mean delays were more than 46 minutes. There was no variation in the length of the delays by time of day (Figure 2). The staff followed the laboratory s standard operating procedures with regard to prioritizing samples: red bags (ie, emergency department samples) had higher priority than electronically ordered samples, which had higher priority than samples with a paper requisition in a plain bag (mean delays were 20, 35, and 41 minutes, respectively [Table 1]). Effects of Accessioning Delays on Total Laboratory TAT To determine the effects of the delays in accessioning on total laboratory TAT, we measured the time to result for samples requiring a BMP and a CBC. These tests were chosen to represent assays that do and do not require centrifugation, respectively. Mean intervals between accessioning and results were 26 minutes for CBCs and 42 minutes for BMPs Table 2. Thus, the delays in accessioning were of the same order of magnitude as the total time needed in the laboratory to perform all the remaining preanalytic, analytic, and postanalytic steps after accessioning: transport to the centrifuge, centrifugation, transport to the instrument bench, technical analysis, review of results, and release of results into the electronic medical record. Nonetheless, using our usual quality metrics, the TATs for stat samples as routinely reported (ie, accessioning-to-result ) generally met the laboratory s targets (ie, 90% of samples reported within 1 hour). However, when TATs are measured from receipt in the stat area (ie, receive to result ), the targets were not met (ie, results of only 72% of CBCs and 29% of BMPs were sent in less than 1 hour). Possible Solutions to the Delays in Accessioning: Laboratory leadership held multiple meetings involving ordering physicians, managers, supervisors, and staff working in the stat receiving area to identify possible solutions to the accessioning delays. Analysis of the present workflow showed that the delay occurred in opening the bags and accessioning the samples. Once this occurred, additional preanalytic steps, such as aliquoting and centrifugation, flowed quickly. Having identified the source of the delay, we considered 4 possible approaches to eliminate it: redesigning the workflow in the preanalytic area, requiring electronic physician order entry for all specimens, introducing preanalytic automation, and adding more full-time equivalent (FTE) employees to the preanalytic area. The task of sample receipt, including opening the sample bag, electronic order entry, and accessioning samples, is the responsibility of 2 full-time stat area clerks. Once accessioned, the samples proceed to the preanalytic work bench, staffed with 2 technologists, where centrifugation and aliquoting occur. These technologists open pneumatic tube canisters for the clerks during busy hours, but do not accession samples. Redesigning workflow to include sample accessioning as part of the technologists responsibilities would likely only push the delay to their step in the process rather than alleviate it, because these technologists, like the clerks, are already operating at full capacity. We therefore did not pursue a workflow redesign. Most samples received in our stat area arrive with a barcode label and need to be scanned to be received in the laboratory. However, a significant percentage of samples come with a handwritten requisition and require the clerical staff to place the necessary orders in the LIS. The time spent accessioning these samples causes delays in the accessioning of all other samples. We therefore considered making electronic order entry and printing of barcode labels obligatory. This would relieve clerks from this task and allow them to focus on opening bags and receiving specimens by scanning the barcodes. Implementing this approach would have required significant expenditures for label printers and computers, as well as full cooperation of clinical staff. It would also be difficult to enforce such a policy, because many of the samples received with a requisition come from the operating rooms, where electronic order entry may not be practical. For these reasons, we decided not to implement this solution. Table 2 Effects of Including Preaccessioning Delays on Total Stat TAT Before and After Increasing Staffing CBC BMP TAT Defined as TAT Defined as TAT Defined as TAT Defined as Accessioning-to-Result Receipt-to-Result Accessioning-to-Result Receipt-to-Result Before increasing staffing Mean (range) TAT, min 26 (8-100) 53 (21-150) 42 (22-86) 78 (49-227) Samples with results within 1 hour, % After increasing staffing Mean (range) TAT, min 26 (10-123) 35 (13-125) 43 (22-121) 51 (25-127) Samples with results within 1 hour, % CBC, complete blood count; BMP, basic metabolic panels; TAT, turnaround time. Downloaded from Am J Clin Pathol 2012;138:

5 Stotler and Kratz / Turnaround Time in a Clinical Laboratory We also considered the purchase of an automated preanalytic system. This would free the staff from the preanalytic tasks of centrifugation and aliquoting, allowing them to open bags and accession samples instead. The reallocation of staff would likely reduce the processing delay; however, the cost of implementing an automated system would be significantly higher than the cost of additional clerical FTEs. Therefore, we opted to first test the least expensive solution, of increasing staffing in the preanalytic stat area to significantly reduce delays. Study to Determine the Effect of Increased Staffing on Delays in Accessioning Two additional employees were assigned to the preanalytic stat area. Delays in accessioning were measured in the same way as in the preintervention observational study. Staff members were aware that an intervention trial was in progress, but they were blinded to the nature of the trial; that is, they were not told that pneumatic tube carriers were sent to the stat area and followed by supervisors. The Student t test was performed to assess the significance of the mean difference in delay times between the 2 groups, with the significance set at P <.05. Effects of Increased Staffing on Accessioning Delays The effects of increased staffing on accessioning delays are summarized in Table 1 and Figure 1. The mean delay in accessioning was reduced from 32 minutes to 8 minutes; the median delay was reduced from 22 minutes to 7 minutes. Eighty-eight percent of all samples were accessioned within 15 minutes (compared with 39% before the intervention). The difference in mean delays between the 2 groups (original staffing compared with postintervention staffing) was statistically significant (P <.01), as was the difference between the delay for each priority level (P <.01; Table 1). Effects of Increased Staffing in the Preanalytic Stat Area on Total Laboratory TATs The effects of the intervention on total laboratory TATs are summarized in Table 2. The mean times from accessioning-to-result for CBCs and BMPs were basically unchanged. However, because the interval between receipt and accessioning was reduced dramatically, the laboratory now met the 1-hour TAT goal for 96% of the CBCs (compared with a baseline of 72%) and 77% of the BMPs (compared with a baseline of 29%). Effects of Increased Staffing in the Preanalytic Stat Area on Other Sections of the Core Laboratory Because of the increased staffing of the stat area, staff members in client services and supervisors reported a noticeable decrease in phone calls and pages by ordering physicians inquiring about delayed or missing stat specimens. In addition, because stat specimens were less likely to be delayed, the technical staff had fewer problems in handling the pending logs. Discussion Like many laboratories, we defined the laboratorycontrolled TAT for the most time-sensitive stat specimens from the time of accessioning in the laboratory to the release of the results into the electronic medical record. Using this criterion, our data indicated that we were generally meeting our TAT goals. Recurring complaints from clinicians that our TATs were actually longer than we reported led us to investigate the validity of the accessioning-to-result time as a complete measure of laboratory-controlled TAT. We identified significant delays between sample arrival in the laboratory and subsequent accessioning. Our hypothesis was that the delays occurred because of a disproportion between the patient sample workload and the number of employees available in the stat preanalytic area to handle this workload. After considering several possible interventions, we performed an interventional study by adding FTEs to the preanalytic area. The study demonstrated that the addition of 2 clerical FTEs would significantly improve TAT in our stat area during day shifts on weekdays. The average annual salary of a clerical FTE at our institution is approximately $35,000; including benefits, the cost of adding 2 FTEs is therefore approximately $100,000. This cost is balanced by faster disposition of patients in the emergency departments, allowing for higher throughputs, increased patient satisfaction, and therefore increased revenues. In addition, shorter TAT of laboratory results for patients in operating rooms will lead to more efficient use of surgical space and personnel, also contributing to increased hospital revenues. Quantification of these additional sources of revenue is beyond the scope of this study. Howanitz and Howanitz 3 noted that improving TAT requires education of a wide variety of individuals, longterm planning, and completion of innumerable tasks. Small investments in the laboratory may improve TAT and greatly improve clinicians efficiency. Our findings testify to the truth of this statement. The tasks performed in the preanalytic area, such as opening specimen bags and accessioning samples, are mainly performed by nontechnical staff, not by trained, licensed, laboratory technologists who are more difficult to recruit and require higher salaries. Therefore, it should be relatively simple to increase the staffing in this area Modern LIS systems allow laboratories to obtain extensive reports on various quality assurance data. These reports can be extremely useful for the day-to-day management of 728 Am J Clin Pathol 2012;138: Downloaded 728 from

6 Clinical Chemistry / Original Article the laboratory, allocation of resources, and long-term planning. However, our experience reinforces the continued importance of old-fashioned monitoring tools: talking to customers, paying attention to their opinions and complaints, and questioning the data in the electronic reports. In summary, laboratories continually need to reassess the validity of their quality data in light of customer feedback. The described approach may also be useful when justifying the need for additional employees during budget discussions with hospital administration. From the Department of Pathology and Cell Biology, Columbia University College of Physicians and Surgeons, New York, NY, and Clinical Laboratory Services, NewYork-Presbyterian Hospital, New York, NY. Address reprint requests to Dr Kratz: NewYork-Presbyterian Hospital, Core Laboratory, 622 West 168th Street, PH3-363, New York, NY 10032; ak2651@columbia.edu. Acknowledgment: The authors thank Steven Spitalnik, MD, for critical reading of the manuscript. References 1. Kilgore ML, Steindel SJ, Smith JA. Evaluating stat testing options in an academic health center: therapeutic turnaround time and staff satisfaction. Clin Chem. 1998;44: Hawkins RC. Laboratory turnaround time. Clin Biochem Rev. 2007;28: Howanitz JH, Howanitz PJ. Laboratory results: timeliness as a quality attribute and strategy. Am J Clin Pathol. 2001;116: Downloaded from Am J Clin Pathol 2012;138: