Analysis of STAT Laboratory Turnaround Times Before and After Conversion of the Hospital Information System

Discussion

The American Recovery and Reinvestment Act of 2009 included provisions that in aggregate comprised the Health Information Technology for Economic and Clinical Health (HITECH) Act.¹ The HITECH Act is focused on improving health-care quality, safety, and efficiency through promotion of health information technology, notably EHR, and through greater electronic exchange of health information.^1,2 The legislation ties payments specifically to the achievement of advances in health-care processes and outcomes.³ Adoption of the modernized EHRS has been extremely slow despite the potential financial incentives in utilizing systems that fulfill the definition of meaningful use as outlined in the HITECH Act.⁴ In a recent survey, Jha et al⁵ found that > 75% of hospitals reported adoption of electronic laboratory and radiologic reporting systems. Our institution adopted an electronic laboratory reporting system in the early 1990s. The transition process to the new application showed a significant increase in TAT and decrease in percent compliance initially; however, laboratory personnel had familiarity with utilizing an LIS before the conversion. This suggests the possibility that the process change to the new EHRS platform initially impacted their ability to process specimens as efficiently as they did before the conversion.

The results indicate that the STAT laboratory was within the defined internal benchmark of average TAT for specimens processed in < 7 min by the 2nd month post-conversion; however, the purpose of this observation was to determine whether and when the TAT returned to the levels noted pre-conversion. There are no national guidelines regarding defined TAT. According to the College of American Pathologists Laboratory General Checklist, “The laboratory has defined turnaround times (ie, the interval between specimen receipt by laboratory personnel and results reporting) for each of its tests,”⁶ but no specific TAT are indicated. These ranges are determined by the laboratory and approved by the STAT laboratory medical director. Also, one would think that when TAT averages returned to < 7 min, compliance would also recover. However, this was not the case. For example, if the STAT laboratory processed 100 samples in 24 h, 95 of the samples would have to have a TAT of < 7 min to attain the 95% compliance target. If 8 samples had a TAT of > 7 min, then compliance would be 92% for that day, even though the overall average TAT could be < 7 min.

The additional step identified during the testing phase (documenting the specimen acquisition time) negatively influenced the TAT and percent compliance to a greater degree than originally anticipated. This step required ordering personnel to log into the system and document the time the specimen was collected. Orders could be placed up to 12 h in advance; however, the collection time had to be entered when the specimen was obtained to complete the order. If this step was not performed, the order could not be retrieved by laboratory personnel to acquire the accession number. To rectify this omission, laboratory personnel had to contact and request ordering personnel to log into the system and document the time the specimen was collected. We speculate that this potentially led to delays in specimen processing. No methodology was devised to separate these results from the overall TAT and percent compliance results to identify the scope of this issue. This additional step proved to be a significant barrier post-conversion, and it is highly recommended that others undergoing a similar transition should focus on any process changes and devise ways to isolate and analyze the impact of any differences noted during the pre-conversion testing and training. It is suspected that the impact of this step was gradually eliminated as clinical staff adapted to the requirements of the new system.

In an attempt to offset anticipated problems (EHRS being taken offline for system resets and software adjustments), procedural steps included a mixture of paper and electronic steps. This method attempted to provide a means for ensuring that required information was available to STAT laboratory personnel in the event the information was missing from the electronic documentation. For example, a paper requisition had to be sent with the specimen to ensure that the required elements for proper documentation were available for entry into the system. These additional requirements may have amplified delays with specimen result processing and negatively impacted TAT and percent compliance during the early post-conversion phase. Again, this is speculative since the data were analyzed in aggregate, and no data were eliminated from the data pool.

Are there other ways to potentially improve the outcomes of a transition to a new EHRS/LIS that will minimize slowdowns and disruptions in the reporting of laboratory results? A recent Institute of Medicine report stated, “Extensive training must be done for the specific product and the specific organizational setting. It is customary for organizations to set expectations for training that require documentation of learning modules and demonstrated competency.”⁷ Before the transition, hospital personnel participated in training processes to prepare for the conversion as described previously. This is one area identified that may not have been utilized to the fullest potential. We were unable to audit the actual time spent by individual users in the test mode. Our experience showed that the training process should have included an audit of time spent by laboratory personnel in the test mode and in processing mock specimens and should have required that the staff document time spent in training. Buntin et al⁸ states that “the ‘human element’ is critical to health IT implementation, and this highlights the importance of strong leadership and staff ‘buy in’ if systems are to successfully manage and see benefit from health information technology.” Despite the pre-conversion training requirements and monitoring by the supervisor and IT staff, the clinical staff and laboratory personnel initially struggled with the changes after the conversion. The adaptation to the new system took several months to return to pre-conversion performance levels.

Another aspect to consider is whether implementation should be a comprehensive versus a gradual conversion. According to the Institute of Medicine, “an organization selects one of two approaches to implementing the technology: either a big bang strategy (ie, the technology is implemented for use throughout the entire organization at the same time) or an incremental approach (ie, the technology is first deployed for use on a small scale within the organization and then, as operating experience is acquired, it is deployed to other parts of the organization in a gradual staged manner). Both approaches can be successful.”⁷ The incremental approach was not an option due to the integration of the new application. This conversion used the big bang strategy, and once the conversion started, it could not be reversed. Use of paper downtime requisitions occurred during several points after the initial conversion as a result of taking the system offline for system resets, software fixes, and upgrades. These negatively impacted TAT and percent compliance. With the benefit of hindsight, we would have set more realistic expectations about the amount of time needed for the STAT laboratory's performance to return to pre-conversion levels. Although there are few reports that delineate a certain length of time for learning a new system (eg, ∼6 months), for better understanding of the system's effects, certain aspects may need to be assessed longitudinally.^9,10 The recovery time for the TAT returning to and staying at baseline took ∼5 months. However, percent compliance took ∼10 months to recover to baseline. These may be realistic expectations when a complex implementation of this magnitude occurs.

The observed increase in TAT and percent compliance had multiple causes. There were delays with specimen processing at various points between users ordering, collecting, documenting, and sending specimens and laboratory personnel receiving, analyzing, result processing, and filing the results. Specific system causes included personnel issues (eg, realistic training to simulate the EHRS in preparation for the conversion, getting acclimated to the differences within the EHRS, and callbacks for missing or incomplete information) and system issues (eg, technical problems, software upgrades, and system downtime) that forced staff to go to downtime processes and revert to paper forms for processing specimens. Although there was an initial increase in TAT and percent compliance, it turned out to be a reasonable trade-off to improve performance and quality in more global ways. The upgraded EHRS enabled us to utilize more advanced functionality that potentially impacts every patient throughout our system.

There are several limitations that should be discussed. First, our results depend on our own internal benchmark, which has been established over time and with multidisciplinary input and review. We are confident that these benchmarks accurately reflect the realities of our setting; however, their generalizability will require applying benchmarks established by others at their own institution. Second, these results are specific to the EHRS/LIS we utilize and are not generalizable to those obtained with other EHRS/LIS systems. Third, this is a dynamic process, and as interventions are made, a review of the applicability of this model should be performed and the necessary modifications made to set reasonable goals that are realistically attainable by the staff.