Reducing Laboratory Turnaround Time in Patients With Acute Stroke and the Lack of Impact on Time to Reperfusion Therapy Open Access

This study was conducted by the central laboratory of Siriraj Hospital, Bangkok, Thailand, a tertiary care academic medical center. We included consecutive patients with suspected acute stroke for whom physicians had activated the stroke fast-track system during the 4-month periods before (March 30–July 30, 2018) and after (November 11, 2018–March 11, 2019) the implementation of the quality improvement process. The process occurred during August to early November 2018. Data were extracted from the laboratory information system and electronic medical records. Complete blood cell count (CBC) was analyzed using an XN-1000 hematology analyzer (Sysmex Corporation, Kobe, Japan) and an LH 780 hematology analyzer (Beckman Coulter, Miami, Florida). Coagulation tests were analyzed using CS-2100i and CA-1500 analyzers (Sysmex Corporation, Kobe, Japan). Plasma creatinine was analyzed using a Cobas 8000 analyzer (Roche Diagnostics). The study was approved by the Siriraj Institutional Review Board (Si205/2019).

A new laboratory workflow was implemented to reduce the TAT of laboratory results in acute stroke patients. The intervention consisted of 5 processes. First, we moved the specimen registration to the same location as the specimen transport starting station. Second, personnel at the transport starting station notified laboratory personnel of relevant laboratory disciplines before the arrival of specimens. Third, we applied a color-coding system on blood collection tubes to identify acute stroke patients. Fourth, a timer was set at each process, including centrifugation (if applied) and analysis of hematology, coagulation, and chemistry, as well as the total testing process time after arrival of the specimens. Finally, abnormal CBC result analysis was released prior to the peripheral blood smear review, except for abnormal platelet parameters. Previously, we performed peripheral blood smears in all types of abnormal CBCs in accordance with our smear review criteria¹³  before releasing the test results (Figure, A through C).

We evaluated the new workflow procedure on 2 fronts: clinical and laboratory. On the clinical side, we assessed door-to-needle time in patients who received intravenous thrombolysis. In patients who had mechanical thrombectomy, we assessed door-to-CTA or computed tomography perfusion (CTP), door-to-puncture, and door-to-recanalization times.

We evaluated 5 time periods on the laboratory front. First, the door-to–blood drawn time was from the time of patient arrival at the ED to the time of blood draw. Second, transportation time was the time from drawing blood to the arrival of the specimens at the laboratory. Third, the intralaboratory TAT was defined as the time between the laboratory personnel receiving a sample and the result being reported to the hospital information system (HIS). Fourth, blood drawn–to–laboratory reported time was defined as the time from drawing blood to reporting all results (CBC, coagulation, and creatinine) to the HIS. We also investigated the time from drawing blood to reporting individual test results to the HIS (test-specific total TAT). We set the target time for test-specific total TAT of CBC and coagulation tests at 30 minutes and for plasma creatinine tests at 45 minutes. Finally, the period from ED arrival to reporting all results was called the door-to–laboratory reported time.

Statistical analyses were performed using SPSS version 18 (IBM Corp, Armonk, New York). Categorical variables were described as frequencies and percentages. TAT intervals were presented as the median and interquartile range (IQR). The χ² test or Fisher exact test was used to compare patient characteristics between groups. The preimplementation and postimplementation time intervals were compared using the Mann-Whitney U test. Statistical significance was defined as P < .05.

Three hundred fifty-eight patients were enrolled in the acute stroke fast-track program during the study period. There were no significant differences in baseline characteristics between preimplementation and postimplementation groups (Table 1). In the preimplementation group (n = 187), the median (IQR) age was 66 (56–78) years and 103 patients (55.1%) were male. In the postimplementation group (n = 171), the median (IQR) age was 64 (52–75) years and 90 patients (52.6%) were male. The most common comorbidity was hypertension (118 [63.1%] preimplementation versus 96 [56.1%] postimplementation), followed by dyslipidemia (60 [32.1%] preimplementation versus 52 [30.4%] postimplementation) and diabetes mellitus (59 [31.6%] preimplementation versus 50 [29.2%] postimplementation). The most common presentation was hemiparesis in both groups (119 [63.6%] preimplementation versus 108 [63.2%] postimplementation). The most common final diagnosis was ischemic stroke (105 [56.1%] preimplementation versus 104 [60.8%] postimplementation), followed by hemorrhagic stroke (33 [17.6%] preimplementation versus 28 [16.4%] postimplementation) and transient ischemic attack (14 [7.5%] preimplementation versus 11 [6.4%] postimplementation). Other neurologic diseases included seizure, transient global amnesia, brain metastasis, peripheral nerve diseases, brain tumor, dementia, posterior cord syndrome, myasthenia gravis, idiopathic Parkinson disease, and brain arteriovenous malformation.

The specimen transportation time was significantly decreased, from 11 (IQR, 8.4–16.4) to 9 (IQR, 6.3–12.8) minutes after the implementation (P < .001) (Table 2). The intralaboratory and test-specific total TATs of CBC, coagulation tests, and plasma creatinine were significantly decreased in the postimplementation group (P < .001 for all). The blood drawn–to–laboratory reported time was significantly decreased, from 43 (IQR, 36.0–51.5) to 33 minutes (IQR, 29.2–35.8; P < .001), and door-to–laboratory reported time decreased from 49 (IQR, 40.3–58.0) to 37 minutes (IQR, 33.6–43.4; P < .001).

The numbers of CBC and coagulation test results reported that exceeded the target time were significantly reduced after the implementation of the new workflow (40 [21.4%] preimplementation versus 1 [0.6%] postimplementation, P < .001 for CBC; 23 [12.3%] preimplementation versus 1 [0.6%] postimplementation, P < .001 for coagulation tests). The number of plasma creatinine results reported after 45 minutes was not different after the implementation (1 [0.5%] preimplementation versus 0 [0%] postimplementation, P = .34) (Table 3).

The proportion of patients who received intravenous thrombolysis or mechanical thrombectomy was not different between preimplementation and postimplementation (18 [9.6%] preimplementation versus 19 [11.1%] postimplementation for intravenous thrombolysis, P = .65; 14 [7.5%] preimplementation versus 10 [5.8%] postimplementation for mechanical thrombectomy, P = .54). The blood drawn–to–laboratory reported time of the laboratory tests improved after the implementation of the new workflow in patients who received thrombolysis or mechanical thrombectomy (44 minutes preimplementation versus 34 minutes postimplementation for intravenous thrombolysis, P = .02; 41 minutes preimplementation versus 31 minutes postimplementation for mechanical thrombectomy, P = .03). However, door-to–laboratory reported time did not improve in either patient subgroup (46 minutes preimplementation versus 39 minutes postimplementation for intravenous thrombolysis, P = .07; 43 minutes preimplementation versus 39 minutes postimplementation for mechanical thrombectomy, P = .29). Median door-to-needle time for patients who received thrombolysis (28 minutes preimplementation versus 35 minutes postimplementation, P = .11) and door-to-puncture time (68 minutes preimplementation versus 97 minutes postimplementation, P = .69) and door-to-recanalization time (97 minutes preimplementation versus 110 minutes postimplementation, P = .50) in the mechanical thrombectomy group were not different (Table 4).

In patients who received intravenous thrombolysis, the door-to–laboratory reported time was longer than the door-to-needle time in 15 of 18 preimplementation patients (83.3%) and 13 of 19 postimplementation patients (68.4%) (P = .45). No patients in the preimplementation group and one in the postimplementation group had a history of anticoagulant medication. In patients who underwent mechanical thrombectomy, the door-to–laboratory reported time was greater than the door-to-CTA/CTP time in 11 of 12 (91.7%) during the preimplementation period and 7 of 10 (70.0%) during the postimplementation period (P = .29). The door-to–laboratory reported time was longer than the door-to-puncture time in 2 of 13 (15.4%) during the preimplementation period and 1 of 10 (10.0%) during the postimplementation period (P > .99). During the preimplementation period, CTA/CTP, puncture, and recanalization times were missing in 2, 1, and 2 patients, respectively. One patient lacked documentation of CTA time. One patient with distal blood clot migration lacked documentation of CTA, puncture, and recanalization time. In another patient with no definitive large vessel occlusion, recanalization was not performed.

Acute ischemic stroke is a major public health problem in adults worldwide. Time is crucial for the initiation of acute therapy, including intravenous thrombolytic therapy and mechanical thrombectomy. We reorganized the laboratory workflow to decrease TAT for acute stroke patients and evaluated the new laboratory workflow. We found that time intervals between blood drawn and specimen received decreased from 11 to 9 minutes (P < .001), and blood drawn–to–laboratory reported time decreased from 43 to 33 minutes (P < .001). However, door-to-needle time in the intravenous thrombolytic group and door-to-puncture time in the mechanical thrombectomy group were not improved after the implementation. Door-to-needle, door-to-CTA/CTP, and door-to-recanalization time seemed to increase after the implementation, which may derive from factors apart from the laboratory aspect.

This study improved the workflow process starting from blood draw; thus, door-to–blood drawn time was comparable between preimplementation and postimplementation, but door-to–laboratory reported time significantly improved. Door-to–laboratory reported time in patients who received thrombolysis or mechanical thrombectomy decreased after the implementation but was not statistically significant, most likely because of the small sample size in this subgroup. The median door-to–laboratory reported time in this study was improved compared with a previous study¹²  conducted in the same institution in 2012 (49 minutes preimplementation period in this study versus 73.5 minutes in the previous study). The improvement of door-to–laboratory reported time could be attributed to the implementation of a transport pipeline system instead of a human courier in 2013, which eliminated the batching and manual packaging steps. Additionally, newer models of automated analyzers for hematology, coagulation, and clinical chemistry with increased throughput were introduced during the period from 2012 to 2018.

In this study, 83.3% of patients received thrombolysis prior to the completion of laboratory results during the preimplementation period, indicating that abnormal laboratory test results were not suspected in the majority of patients and thrombolytic therapy was given without delay while results were pending. After the implementation of the new laboratory workflow, the proportion of patients who received thrombolysis prior to the completion of laboratory results was reduced to 47.4%. Similarly, the majority of patients receiving mechanical thrombectomy underwent CTA/CTP prior to receiving laboratory results (91.7% preimplementation and 70.0% postimplementation), indicating that physicians did not wait for plasma creatinine before performing CTA/CTP unless there was a suspicion of renal impairment. Groin puncture occurred after the completion of laboratory results in most patients because this process required transferring patients to a different location, contacting various team members, and involved numerous steps.

This study assessed 2 outcomes: time to clinical intervention and laboratory TAT. There was no correlation between these outcomes, which was understandable given that physicians generally followed the guidelines^10,11  not to wait for laboratory results unless they observed certain warning signs, such as a history of anticoagulant use, clinical findings associated with abnormal bleeding, or a history of renal impairment.

One of the most common causes of ED test delays is related to collecting and transporting specimens.¹⁴  To reduce transportation time, we moved the specimen registration to the same location as the specimen transport starting station. Using point-of-care testing (POCT) is a solution to reduce laboratory TAT because POCT eliminates the transportation step. Previous studies^15–18  showed that POCT coagulation devices provide rapid and reliable results compared with central laboratory testing. However, another study¹⁹  demonstrated that 4.6% of cases showed differences between POCT and central laboratory testing at the critical thrombolysis cutoff (international normalized ratio 1.7). A recent review of POCT for ED patients found that POCT improved laboratory TAT in patients with nonviral conditions, but there was inconsistent improvement in the length of stay at the ED.²⁰ 

In the analytical phase, intralaboratory communication delays, technical delays, specimen delays, personnel delays, and laboratory accidents are important.²¹  We improved intralaboratory communication by notification of specimen arrival in advance, and applied color coding on the tubes to make them easily recognizable. We reduced technical delays by setting a timer at each process and reduced specimen delays by reviewing the peripheral blood smear only in abnormal platelet specimens before releasing the preliminary results. Laboratory TAT can be reduced by various methods, including the use of lean methodologies,^22,23  monitoring, root cause analysis, and corrective actions,²⁴  reorganizing the location as well as staff, and using automated technologies.²⁵ 

A meta-analysis²⁶  of interventions to reduce door-to-needle times in acute ischemic stroke patients found that implementing combination strategies resulted in the greatest reductions. No studies examined rapid laboratory testing as a single reduction strategy among the 96 included in a meta-analysis.²⁶  A previous study²⁷  demonstrated that implementing nursing quality improvement reduced median door-to-needle time from 73 to 49 minutes. Another study showed that introducing a stroke nurse reduced median door-to-needle time from 36 to 25 minutes.²⁸ 

Our study demonstrated that improving laboratory TAT in isolation did not result in a reduction in time to clinical intervention. Improvement in time to clinical intervention is more important than improving laboratory TAT alone. Quality improvement requires interdisciplinary collaboration and is not solely the domain of one discipline. Intralaboratory improvement efforts are worthwhile because they can result in more timely laboratory results, which provides physicians with greater confidence when making the therapeutic decision. Multidisciplinary team coordination is required to demonstrate clinical impact.

We recommend that the team include emergency medical services providers, nurses, emergency medicine physicians, neurologists, diagnostic radiologists, radiology technicians, laboratory personnel, pharmacists, anesthesiologists, neurointerventionists, neurosurgeons, and hospital administrators.²⁹  Then, stakeholders can apply lean methodology by creating a value stream map that details the process steps in a workflow and identifies non–value-added activities from start to finish.^30,31  All team members need a thorough understanding of their respective roles and responsibilities, as well as the interrelationships and dependencies among the various hospital services involved in the timely diagnosis and treatment of an acute ischemic stroke.³²  Effective and rapid communication among stroke team members is critical for optimizing efficiency and, consequently, for early treatment.^33,34  Additionally, continuous improvement requires regular monitoring of key performance indicators and acknowledgment by team members.^35–37 

This study has some potential limitations. First, this was a single-center study and so the findings might not be generalizable to other hospitals with different stroke fast-track protocols and different laboratory process workflows. Second, this was a retrospective, before-after study, and the observed findings may have been due to other unrecognized contributing factors such as differences in individual productivity or subtle differences across the patients we studied. However, we were unable to identify any other significant systems or operations changes in either clinical or laboratory processes during the study period. Finally, we did not evaluate the effects of reduced laboratory TAT on hospital or ED length of stay or other patient outcomes.

This laboratory improvement process significantly decreased transportation time, TAT of individual tests, and blood drawn–to–laboratory reported time. Despite the improvement of laboratory TAT, the time to treatment of acute ischemic stroke patients was not different before and after the implementation of the new laboratory workflow. Generally, physicians did not wait for laboratory results unless they observed certain warning signs such as a history of anticoagulant treatment, clinical abnormalities consistent with abnormal bleeding, or a history of renal dysfunction. Individuals with a suspicion of an abnormal test had to wait for the results of a laboratory test before starting intravenous thrombolysis or mechanical thrombectomy; otherwise, the treatment was initiated without delay while laboratory results were awaited.

The authors would like to thank Noppadol Arechep, BSc; Jiraporn Bhucharoen, BSc (MedTech); Bongkot Wesarachkitti, MSc; Apisit Pan-ame, BSc (MedTech); and Rungkarn Budda, BSc (MedTech), for providing relevant information.