Reduced Workflow Times for Reperfusion Therapy After Acute Ischemic Stroke Using a Visual Task Management Application

Abstract

Background

Reperfusion therapy for acute ischemic stroke efficacy is highly time dependent; therefore, stroke centers are required to further reduce the delays from hospital arrival to treatment efficiently. We developed a visual task management application, Task Calculation Stroke (Task Calc. Stroke: TCS), to facilitate hospital acute ischemic stroke treatment by supporting parallel staff task completion. We evaluated TCS for the reduction of reperfusion therapy delays and improvement of clinical outcomes.

Methods

In this multicenter cohort study, patients were directly admitted to 4 comprehensive stroke centers in Japan and given intravenous tissue plasminogen activator and/or mechanical thrombectomy from June 2018 to December 2020. The research team visited each facility and instructed the staff on TCS use for acute ischemic stroke (training stage), after which the staff used TCS independently (TCS stage). We then compared door‐to‐needle time for intravenous tissue plasminogen activator, door‐to‐puncture time for mechanical thrombectomy, and clinical outcomes at discharge according to the modified Rankin Scale among patients treated before training (original stage), during the training stage, or the TCS stage.

Results

During the study period, 316 patients with acute ischemic stroke received reperfusion therapy; of these, 246 received intravenous tissue plasminogen activator and 162 mechanical thrombectomy (including 92 receiving both the treatments). The mean door‐to‐needle time was significantly reduced from 58.0 minutes in the original stage to 54.6 minutes in the training stage (P=0.049) and 47.8 minutes in the TCS stage (P<0.001). The door‐to‐puncture time did not change during the training stage; however, in the TCS stage, it significantly reduced from 93.8 minutes in the original stage to 88.5 minutes (P=0.004). The distribution of modified Rankin Scale scores at discharge significantly shifted favorably at the TCS stage (P=0.003).

Conclusion

In this study, TCS application could reduce workflow time for reperfusion therapy and might have led to improved clinical outcomes.

Nonstandard Abbreviations and Acronyms

AIS

acute ischemic stroke

D2I

door‐to‐brain image

D2N

door‐to‐ needle

D2P

door‐to‐puncture

D2R

door‐to‐recanalization

I2N

image‐to‐needle

I2P

image‐to‐puncture

ICT

information and communication technology

IV‐tPA

intravenous tissue plasminogen activator

mRS

modified Rankin Scale

MT

mechanical thrombectomy

QM

quality management

TCS

Task Calc. (Calculation) Stroke

TICI

Thrombolysis in Cerebral Infarction

Clinical Perspective

What Is New?

• This multicenter cohort study, conducted at 4 comprehensive stroke centers in Japan, found that a visual task management application could significantly reduce workflow time for treating acute ischemic stroke and improve clinical outcomes for patients at discharge.

What Are the Clinical Implications?

• This study focused on improving clinical multitask progress management for acute stroke care at comprehensive stroke centers using information and communication technology.

• Clinical multitask progress management was improved via 4 functions of our application: alert, synchronous timer, visual task management, and history analysis.

Ischemic stroke is one of the leading causes of premature death and long‐term disability worldwide. ¹ In Japan, 290 000 new or recurrent strokes occur each year, and 66% are ischemic strokes. ² Acute reperfusion therapy, including intravenous tissue plasminogen activator (IV‐tPA), mechanical thrombectomy (MT), and combined IV‐tPA/MT therapies, can rescue neural tissue and preserve its function in patients with acute ischemic stroke (AIS). ³ , ⁴

The benefit of reperfusion therapy is highly time dependent, and a longer delay from stroke onset generally results in poorer functional prognosis. ⁵ , ⁶ Guidelines recommend a time interval from patient arrival to IV‐tPA administration, known as the door‐to‐needle (D2N) time, of ≤30 to 60 minutes, and from patient arrival to the start of MT, known as door‐to‐puncture (D2P) time, of ≤60 to 90 minutes. ⁷ , ⁸

The Target: Stroke project that involved leading stroke centers succeeded in reducing D2N time by organizational quality management (QM) activities using 10 key strategies, including emergency medical service prenotification; activation of the stroke team with a single call; rapid acquisition and interpretation of brain imaging; and use of specific protocols and tools, premixed tissue plasminogen activator, a team‐based approach, and rapid data feedback. ⁹ , ¹⁰ However, many stroke centers still find it challenging to achieve these target values. ¹¹ , ¹² Critical to faster diagnosis and treatment is parallel rather than serial workflow. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ However, achieving parallel task initiation and completion requires a strong commitment to preparation, organization, and communication among all stroke team members. ¹⁷ Furthermore, routine parallel workflow is difficult without established QM protocols and optimal organization of treatment sites and staff availability in the hospital.

To assist hospitals in initiating QM protocols for timely AIS treatment, we developed a visual task management application named “Task Calculation Stroke (Task Calc. Stroke: TCS)” using information and communication technology (ICT) to support communication and coordination among the team, thereby enabling parallel task completion at the stroke center. ¹⁸ , ¹⁹ This multicenter, observational study assessed the value of TCS‐based AIS care for reducing the delay from arrival to reperfusion therapy and improving clinical outcome.

Methods

Data that are not provided in the article and additional information about materials and methods are available from the corresponding author on reasonable request.

This was a multicenter, observational study of patients with AIS admitted directly to 4 comprehensive stroke centers in Japan who received IV‐tPA and/or MT between June 2018 and December 2020 (30 months). Adults (aged ≥18 years) treated for AIS with IV‐tPA, MT, or both were enrolled. The study period was divided into 3 stages: an original stage for observation before TCS training, a training stage to introduce TCS, and a practice or TCS stage of AIS care using TCS. At the beginning of the training stage, a research team composed of QM specialists, stroke specialists, and ICT specialists visited each facility and delivered a 3‐step instructional program on AIS care using TCS for 2 or 3 sessions (Table 1).

Table 1.

The 3‐Step TCS Training Program

Step 1. Build a team and share the purpose of TCS‐based AIS care

The research team visits each stroke center; explains TCS‐based AIS care to the stroke team, including ED nurses, ED physicians, pharmacists, radiology technicians, laboratory technicians, SCU nurses, and stroke specialists; and obtains their consent to participate in the study.

Step 2. Analyze AIS care workflows and organize critical tasks into TCS

The research team analyzes the AIS care workflow for each stroke center and, if necessary, suggests improvements. In addition, critical tasks that could be rate‐limiting steps in AIS care, such as imaging and blood tests for IV‐tPA, as well as securing treatment rooms and staff for MT are implemented in TCS so that the team can conduct quality management.

Step 3. Conduct simulations before introducing TCS‐based AIS care

The research team conducts a simulation of AIS care using TCS with the participation of stroke team at each stroke center.

AIS indicates acute ischemic stroke; ED, emergency department, IV‐tPA, intravenous tissue plasminogen activator; MT, mechanical thrombectomy; SCU, stroke care unit; and TCS, Task Calculation Stroke.

After the 3‐step instructional program, smartphones or tablets with the TCS application installed were distributed to staff in the emergency department (ED), computed tomography room, magnetic resonance imaging room, angiography room, hematology laboratory, and stroke care unit and to individual stroke specialists. Each facility then began to practice AIS care using TCS (TCS stage).

Important Features of TCS

TCS is a mobile application for stroke care management that can be used on commercially available smartphones or tablets via the internet. The 4 important features of TCS ¹⁴ , ¹⁵ are summarized in the following list.

TCS has an alert function to temporarily focus the stroke team's efforts on stroke emergencies without a phone call. When ED physicians receive a prearrival notification from emergency medical services about a patient with suspected stroke, they must contact various stroke team members from several divisions of the hospital by phone. However, sometimes the phone lines are so busy that they must call back repeatedly. It is like a message game between multiple departments (Figure 1A). In such cases, ED physicians, using TCS, can quickly notify all stroke team members with the artificial voice message “Code stroke initiated” with just 1 click of the device. Moreover, when the patient arrives at the hospital, the ED physician can notify the entire team with the artificial voice message “The target patient has arrived.” Furthermore, when the code stroke tasks are completed, the ED physician or stroke specialist can notify all stroke team members with the artificial voice message “Code stroke completed” with just 1 click of the device (Figure 1B).

Figure 1.

Alert function of Task Calculation Stroke.

A, When emergency department physicians receive a prearrival notification from emergency medical services about a patient with suspected stroke, they must contact various team members from several divisions of the hospital by phone. However, sometimes the phone lines are so busy that they must call back repeatedly. It is like a message game between multiple departments. B, In such cases, emergency department physicians, using Task Calculation Stroke (Task Calc. Stroke: TCS), can quickly notify all stroke team members with the artificial voice message “Code stroke initiated” with just 1 click of the device. Moreover, when the patient arrives at the hospital, the emergency department physician can notify the entire team with the artificial voice message “The target patient has arrived.” Furthermore, when the code stroke tasks are completed, the emergency department physician or stroke specialist can notify all stroke team members with the artificial voice message “Code stroke completed” with just 1 click of the device. CT indicates computed tomography; ER, emergency room; and MR, magnetic resonance.

Synchronous timer functions enable all team members to estimate the start time of their tasks for the target patient. On the TCS start dashboard, the input of information, including the probable time (in minutes) before the patient's arrival at the hospital, is followed by clicking the start button (Figure 2A[a]); the screen then changes, and synchronized timers appear at the center of the dashboard (Figure 2A[b]). From prearrival notification to patient arrival, TCS displays the time remaining until the estimated arrival time on a yellow background (Figure 2A[b and c]). After arrival, the time elapsed from arrival is displayed on a white background (Figure 2A[d]). The synchronous timer function allows the entire team to estimate the start time of their task for the target patient so that they can minimize the waiting time for starting the task by adjusting their work.

Figure 2.

The synchronous timer, visual task management, and performance history analysis functions of TCS.

A, Synchronized timer function: on the TCS start dashboard, the input of information, including the probable time (in minutes) before the patient's arrival at the hospital, is followed by clicking the start button; the screen then changes, and synchronized timers appear at the center of the dashboard (a). From prearrival notification to patient arrival, the TCS displays the time remaining until the estimated arrival time on a yellow background (b and c). After arrival, the time elapsed from arrival is displayed on a white background (d). Visual task management function: critical tasks of stroke care assigned to medical staff in each department (nurses, doctors, radiology technicians, and laboratory technicians) are displayed as colored hexagons in order of priority on the TCS dashboard (b through d). Performance history analysis function: after the completion of all tasks, the main processing time is immediately visualized and can be compared between patients (e). B, The colors indicate the following: clicking on the “tasks to do” icon shows the status and progress of each task by a change in color: unconfirmed patient arrival notification (white), confirmed patient arrival notification (yellow), completed patient task preparation (olive green), processing patient task (blue), completed patient task (green), occurrence of an error (red), and task that no longer needs to be performed (gray). CBC indicates complete blood cell count; CT, computed tomography; DTN, door to needle; ER, emergency room; EVT, endovascular therapy; IV‐tPA, intravenous tissue plasminogen activator; MT, mechanical thrombectomy; PT‐INR, prothrombin time–international normalized ratio; RN, registered nurse; SCU, stroke care unit; and TCS, Task Calc. (Calculation) Stroke; POC, point of care; NCU, neuro intensive care unit; PM&R, physical medicine and rehabilitation; TASK, task.

There are visual task management functions to detect the locations of problems with delay or occurrence, similar to a bird's‐eye view. Critical tasks of stroke care assigned to medical staff in each department (nurses, doctors, radiology technicians, and laboratory technicians) are displayed as colored hexagons in order of priority on the TCS dashboard (Figure 2A[b through d]). Clicking on the “tasks to do” icon shows the status and progress of each task by a change in color: unconfirmed patient arrival notification (white), confirmed patient arrival notification (yellow), completed patient task preparation (olive green), processing patient task (blue), completed patient task (green), occurrence of an error (red), and task that no longer needs to be performed (gray) (Figure 2B). This feature of TCS enables the physician who is overseeing the practice to have a single view of the overall processing status of the stroke care task like a bird's‐eye view. Thus, if any department is experiencing delays or problems, you can change the display color of the task and enter the problem in the chat box in the TCS to immediately share the problem with your team, preventing unexpected delays in the process.

There is a performance history analysis function to make real‐time feedback easy. TCS automatically records the processing time data for every case. Key processing times can be visualized and compared among patients, which makes it easy for real‐time feedback of time metrics to the stroke team members (Figure 2A[e]).

Outcome Measures

The primary outcomes were the D2N and D2P times. The secondary outcomes were the door‐to‐brain‐image (D2I) time, the door‐to‐recanalization (D2R) time, the proportion of patients with a modified Rankin Scale (mRS) score of 0 to 2 (the scale ranges from 0 [no symptoms] to 6 [death]) at discharge, and the in‐hospital mortality rate. For reperfusion status after endovascular therapy, the Thrombolysis in Cerebral Infarction (TICI) grade ²⁰ was used. Successful recanalization was defined as TICI grades 2b to 3, and the cases that had attained this status were used for D2R time analysis.

Standard Protocol Approvals, Registrations, and Patient Consent

The Ethics Committee of Fujita Health University approved this study. Opt‐out consent was obtained instead of written informed consent. We provided all potential patients with information explaining the proposed research plan (purpose, required individual data, and study duration) via the website of each stroke center and gave them the opportunity to opt out at any point during the study. All procedures were in accordance with the Helsinki Declaration of 2000 and the Declaration of Istanbul of 2008.

Sample Size

The sample size was determined a priori to ensure the feasibility of the study and avoid overfitting the model used in the primary analysis. Because we planned to include 14 covariates and a group explanatory variable consisting of 3 categories in the model, at least 240 patients were required (16 free parameters multiplied by 15) to avoid overfitting. ²¹ The final sample size (n=316) greatly exceeded the minimum required.

Statistical Analysis

Categorical variables are expressed as numbers and percentages, and continuous variables are expressed as medians with interquartile ranges. Multiple linear regression models were used to assess independent associations between TCS use and key task processing times (D2N, D2P, D2I, D2R) with adjustments for age, sex, facility, initial National Institutes of Health Stroke Scale score, prestroke mRS score, large vessel occlusion, atrial fibrillation, hypertension, diabetes, dyslipidemia, smoking, alcohol consumption, and time of arrival (during or outside business hours) as covariates. Because the distributions of all key task processing times were skewed to the left (shorter), the outcomes were natural log‐transformed to normalize the residuals. The coefficients obtained from the regression model were then back‐transformed to display the results. The association between mRS score at discharge as an ordinal category and use of TCS was assessed by ordinal logistic regression analysis adjusted for the same covariates as in the primary analysis. We used multiple imputations to assign values to missing covariates. Group differences in clinical outcomes were analyzed by the Kruskal–Wallis rank sum test or Pearson χ² test as indicated among patients with a prestroke mRS score of 0. A 2‐sided P value<0.05 was considered to indicate statistical significance for all tests. All statistical analyses were performed using R version 4.1.0 (http://www.rproject.org).

Results

Patient Characteristics

In this 30‐month multicenter cohort study, 316 patients with AIS were eligible for analysis. Of these, 246 received IV‐tPA and 162 received MT (these totals include 92 patients who received both IV‐tPA and MT). All 316 patients were included in the D2I time analysis (154 in the original stage, 78 in the training stage, and 84 in the TCS stage). The 246 patients who received IV‐tPA were included in the D2N time analysis (117 in the original stage, 63 in the training stage, and 66 in the TCS stage), and the 162 patients who underwent MT were included in the D2P time analysis (81 in the original stage, 35 in the training stage, and 46 in the TCS stage). Successful reperfusion was not achieved in 22 patients undergoing MT, and therefore 140 patients treated with MT were included in the D2R time analysis (68 in the original stage, 31 in the training stage, and 41 in the TCS stage). Patient selection and stratification are shown diagrammatically in Figure 3. The baseline characteristics of the patients treated in the 3 stages of the study are summarized in Table 2. There were no significant differences in baseline demographics, stroke severity, and hours from last known well time to arrival. Statistically significant differences were observed in the frequencies of hypertension, dyslipidemia, and large vessel occlusion.

Figure 3.

Flowchart of patient selection and stratification.

Shown are the total numbers of patients treated during all study stages, the numbers treated during each stage (original, training, and TCS) and the specific treatments (IV‐tPA, MT, or both). Except for 7 cases with insufficient time information, we included 246 patients for analysis of door‐to‐needle time and 162 patients for analysis of door‐to‐puncture time. However, recanalization failed in 22 patients undergoing MT; thus, only 140 patients were included in the door‐to‐recanalization time analysis. IV‐tPA indicates intravenous tissue plasminogen activator; MT, mechanical thrombectomy; and TCS, Task Calc. (Calculation) Stroke.

Table 2.

Characteristics of Patients Treated During Each Study Stage

Variable	No.	Overall, N=316	Original stage, n=154	Training stage, n=78	TCS stage, n=84	P value*
Age, y	316	78 (70, 86)	78 (70, 86)	78 (68, 84)	80 (70, 84)	0.635
Sex	316					0.881
Female sex		148 (46.8)	72 (46.8)	35 (44.9)	41 (48.8)
Male sex		168 (53.2)	82 (53.2)	43 (55.1)	43 (51.2)
Hypertension	316	163 (51.6)	93 (60.4)	32 (41.0)	38 (45.2)	0.008
Diabetes	316	75 (23.7)	37 (24.0)	19 (24.4)	19 (22.6)	0.960
Atrial fibrillation	316	102 (32.3)	57 (37.0)	20 (25.6)	25 (29.8)	0.183
Dyslipidemia	314	80 (25.5)	47 (30.9)	12 (15.4)	21 (25.0)	0.037
Pre‐mRS score	307	0.0 (0.0, 1.0)	0.0 (0.0, 1.0)	0.0 (0.0, 2.0)	0.0 (0.0, 0.0)	0.213
LKW arrival	314	70.0 (44.0, 162.5)	80.5 (44.2, 179.8)	70.0 (48.0, 140.0)	57.0 (42.0, 145.5)	0.358
Normal business h	316	171 (54.1)	85 (55.2)	43 (55.1)	43 (51.2)	0.821
NIHSS score at admission	306	14.0 (7.0, 22.0)	15.0 (8.0, 23.0)	12.5 (6.0, 19.5)	13.5 (6.2, 20.5)	0.126
Large vessel occlusion	315	208 (66.0)	109 (70.8)	42 (54.5)	57 (67.9)	0.045
Method of therapy	316					0.674
IV‐tPA alone		154 (48.7)	73 (47.4)	43 (55.1)	38 (45.2)
MT alone		70 (22.2)	37 (24.0)	15 (19.2)	18 (21.4)
Combined IV‐tPA and MT		92 (29.1)	44 (28.6)	20 (25.6)	28 (33.3)

Data are shown as number, median (interquartile range), or number (percentage). IV‐tPA indicates intravenous tissue plasminogen activator; LKW, last known well; mRS, modified Rankin Scale; MT, mechanical thrombectomy; NIHSS, National Institutes of Health Stroke Scale; and TCS, Task Calc. (Calculation) Stroke.

^* Kruskal–Wallis rank sum test; Pearson χ² test.

Primary Outcomes

The geometric mean D2N time was significantly reduced during the training stage compared with the original stage and further during the TCS stage (58.0, 54.6, and 47.8 minutes, respectively). Multivariable analyses with adjustments for age, sex, facility, initial National Institutes of Health Stroke Scale score, prestroke mRS score, large vessel occlusion, atrial fibrillation, hypertension, diabetes, dyslipidemia, smoking, alcohol consumption, and time of arrival (during or outside business hours) as covariates showed that D2N time was slightly shortened during the training stage; however, the difference was not significant (exponentiated regression coefficient: exp[β], 0.879 [95% CI, 0.730–1.058]; P=0.171). Moreover, D2N time was significantly reduced in the TCS stage (exp[β], 0.701 [95% CI, 0.588–0.836]; P<0.001) compared with the original stage (Figure 4A). The geometric mean D2P time in the 3 stages was 93.8, 96.7, and 88.5 minutes, respectively. In multivariable analysis, D2P time was slightly prolonged during the training stage; however, the difference was not significant (exp[β], 0.879 [95% CI, 0.733–1.098]; P=0.29). In the TCS stage, however, the geometric mean D2P time was significantly reduced to 88.5 minutes (exp[β], 0.718 [95% CI, 0.574–0.899]; P=0.004) compared with the original stage (Figure 4B).

Figure 4.

The outcome measure's changes in the training and TCS stages compared with the original stage.

A, Door‐to‐needle time. B, Door‐to‐brain‐image time. C, Door‐to‐puncture time. D, Door‐to‐recanalization time. All results are expressed as geometric means and geometric standard deviations. Adjusted for age, sex, facility, initial National Institutes of Health Stroke score, prestroke modified Rankin Scale score, large vessel occlusion, atrial fibrillation, hypertension, diabetes, dyslipidemia, smoking, alcohol consumption, and time of arrival (during or outside business hours) as covariates. ^*Significant difference at P<0.05. ^**Significant difference at P<0.01. TCS indicates Task Calc. (Calculation) Stroke.

Secondary Outcomes

The geometric mean D2I time was reduced from 16.9 minutes in the original stage to 14.1 minutes in the training stage and 13.9 minutes in the TCS stage. In multivariable analysis adjusted for the same covariates as in the primary analysis, the reduction in D2I time was not significant in both the training (exp[β], 0.882 [95% CI, 0.742–1.048]; P=0.153) and TCS (exp[β], 0.881 [95% CI, 0.741–1.047]; P=0.149) stages compared with the original stage (Figure 4C). The geometric D2R times in the 3 stages were 158.4, 160.2, and 131.7 minutes, respectively. In multivariable analysis, D2R time did not change in the training stage (exp[β], 0.947 [95% CI, 0.796–1.126]; P=0.535). However, D2R time reduced significantly in the TCS stage (exp[β], 0.713 [95% CI, 0.589–0.864]; P=0.001) compared with the original stage (Figure 4D). We showed the arithmetic means and standard deviations alongside the geometric means and geometric standard deviations for the workflow variables in Table 3.

Table 3.

Geometric and arithmetic means of workflow variables

Variable	Overall	Original stage	Preparation stage	TCS stage
Door‐to‐needle time	N = 244	N = 117	N = 63	N = 66
Geometric mean (SD)	54.2 (1.6)	58.0 (1.5)	54.6 (1.6)	47.8 (1.6)
Mean (SD)	59.9 (27.5)	63.5 (27.8)	60.8 (28.3)	52.8 (25.0)
Door‐to‐puncture time	N = 161	N = 81	N = 35	N = 46
Geometric mean (SD)	92.8 (1.7)	93.8 (1.8)	96.7 (1.7)	88.5 (1.5)
Mean (SD)	106.5 (67.0)	111.8 (83.6)	108.4 (49.2)	95.7 (39.9)
Door‐to‐Brain image time	N = 316	N = 154	N = 78	N = 84
Geometric mean (SD)	15.4 (1.8)	16.9 (1.9)	14.1 (1.7)	13.9 (1.6)
Mean (SD)	18.4 (15.3)	21.3 (19.9)	16.2 (9.5)	15.2 (6.8)
Door‐to‐recanalization time	N = 136	N = 68	N = 31	N = 41
Geometric mean (SD)	150.2 (1.5)	158.4 (1.7)	160.2 (1.4)	131.7 (1.4)
Mean (SD)	165.6 (83.6)	180.5 (104.5)	170.9 (64.2)	138.2 (44.6)

Data are shown as number, geometric mean (geometric standard deviation), and mean (standard deviation). SD indicates standard deviation; TCS, Task Calc. Stroke.

The overall distributions of the mRS scores at discharge for the 3 stages are shown in Figure 5. In the ordinal logistic regression analysis adjusted for the same covariates as in the primary analysis, the proportion of patients with an mRS score ≤2 at discharge significantly increased as the stage progressed (from 46.6% during the original stage to 52.0% during the training stage and 64.7% during the TCS stage; P=0.003). The mortality rate at discharge decreased slightly from 9.7% during the original stage and 10.0% during the training stage to 4.6% during the TCS stage; however, the difference was not significant (P=0.444).

Figure 5.

Distribution of mRS scores at discharge during each study stage.

The severity of disability significantly decreased across the mRS score range (adjusted odds ratio, 0.36 [95% CI, 0.18–0.7]; P=0.003). The frequency of mortality (mRS score of 6) at discharge decreased slightly as the study progressed, but the change was not statistically significant (P=0.444) between the 3 stages. The scores on the mRS indicate the following: 0, no symptoms; 1, no significant disability despite the presence of symptoms (able to carry out all usual duties and activities); 2, slight disability (unable to carry out all previous activities but able to look after own affairs without assistance); 3, moderate disability (requiring some help but able to walk without assistance); 4, moderately severe disability (unable to walk without assistance and unable to attend to own bodily needs without assistance); 5, severe disability (bedridden, incontinent, and requiring constant nursing care and attention); and 6, death. Adjusted for age, sex, facility, initial National Institutes of Health Stroke score, prestroke mRS score (created during 3‐step TCS preparation program), large vessel occlusion, atrial fibrillation, hypertension, diabetes, dyslipidemia, smoking, alcohol consumption, and time of arrival (during or outside business hours) as covariates. mRS indicates modified Rankin Scale; and TCS, Task Calc. (Calculation) Stroke.

Discussion

The current study demonstrates that a mobile task management application for AIS care, TCS, can reduce the time from patient arrival to reperfusion therapy. Furthermore, there was a significant improvement in the clinical outcome at discharge as the stage progressed.

There have been several previous attempts to improve AIS care using ICT applications. ⁴ Although these ICT applications are good at sharing image and text information, they do not have the specialized multitasking progress management capabilities of TCS. ¹⁷ , ¹⁸ To the best of our knowledge, this study is the first report to show that purely ICT‐based multitasking progress management can shorten the workflow time for stroke treatment and improve the prognosis of patients. Although these applications are good at sharing images and text information, most are not designed to manage AIS care multitasking. ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ Unlike previous applications, TCS was designed and developed to promote parallel task completion among stroke center staff by supporting communication and task coordination. ²⁹ , ³⁰ The following 4 TCS features may be crucial for reducing door‐to‐treatment times: the notification feature alerts staff about patient arrival, a synchronous timer feature enables all team members to estimate the start time of their tasks for the target patient, the visual task management feature shares the tasks required among team members and tracks parallel completion, and the history analysis feature monitors team performance to identify areas of improvement.

Achieving the full potential of TCS depends on the precise training of stroke center staff. Therefore, we established a training team to introduce TCS to individual stroke centers, consisting of QM, stroke, and ICT experts. In fact, D2N time showed a statistically significant reduction during the training stage. It has been reported that organization of stroke care, review of medical results, and medical simulation with expert support can reduce D2N time. ²⁷ , ²⁸ In this study as well, the training itself may have promoted the reduction in D2N time independently of TCS use. The unique feature of simulation training using TCS is that the entire team can easily simulate various patterns of stroke care simultaneously, even if the entire team is not in the exact location, by simply inputting virtual stroke care patterns and expected patient arrival times. Furthermore, the D2N time during the TCS period was further reduced. It is assumed that this is attributed to the function of the TCS itself and the continuous training effect contributed by the repeated use of the TCS application during the TCS period.

However, the D2P and D2R times did not decrease significantly during the training stage, but they did during the TCS stage when the staff used the application without external support. To start MT, it is necessary to arrange an angiography room and recruit radiology technicians, nurses, and an endovascular surgeon. Thus, as the number and complexity of tasks increases, it may be more difficult to shorten the workflow time by training alone. Rather, substantial reduction in the D2P time may require experience with the use of the TCS application among multiple team members. Indeed, the D2N time was further reduced during the TCS stage. Therefore, the continued use of TCS by the AIS care team reduced workflow time, suggesting that TCS alone may have further enhanced QM activities.

In contrast to the D2N and D2P times, the D2I time was not significantly reduced by TCS, possibly because it was already only 16.9 minutes at all 4 stroke centers, substantially lower than the 25 minutes required by the American Heart Association/American Stroke Association guidelines. ⁷ The geometric mean image‐to‐needle (I2N) times in the 3 stages were 38.2, 37.9, and 32.6 minutes, respectively. In the multivariable analysis, the I2N time did not change in the training stage (exp[β], 0.879 [95% CI, 0.73–1.058]; P=0.171). However, the I2N time was reduced significantly in the TCS stage (exp[β], 0.701 [95% CI, 0.588–0.836]; P<0.001). The geometric mean image‐to‐puncture (I2P) times in the 3 stages were 73.1, 85.7, and 71.6 minutes, respectively. In the multivariable analysis, the I2P time did not change in the training stage (exp[β], 0.983 [95% CI, 0.786–1.228]; P=0.877). However, the I2P time was significantly reduced in the TCS stage (exp[β], 0.692 [95% CI, 0.543–0.883]; P=0.003). Therefore, we considered that the reason the I2N and I2P times, but not the D2I time, was shortened when using the TCS was attributed to the difference in the number of team members involved in task processing. Tasks performed between the I2N and I2P times require the cooperation of more team members than tasks performed between the D2I times. It can thus be said that TCS is more effective in tasks involving more team members.

Of greater importance than the reduced workflow times, the severity of disability at discharge was reduced in the TCS stage compared with the pre‐TCS period. Although the causal relationships between TCS use, reduced key processing times, and mid‐ to long‐term clinical outcomes, such as mRS score at 3 months, remain unclear, these findings provide a strong rationale for larger scale application studies with more extensive follow‐up.

TCS is a mobile task management application that can be downloaded from anywhere in the world, and institutions can begin using it on readily available smartphones or tablets after a brief 3‐step training program. In this study, the research team visited each facility to conduct on‐site training before starting AIS care using TCS, but we are now developing a remote training program using web conferencing and video materials to facilitate wider implementation. The positive effect of using ICT in stroke care can be gained not only by task processing at the stroke center, as in this study, but also by prehospital task processing, such as selecting the nearby hospitals that have the provision of reperfusion therapy. ²⁸ Therefore, by extending the function of TCS to prehospital task processing, it may be possible to achieve faster AIS care.

This study has several limitations. First, institutional staffing and available resources will likely influence the practical utility of TCS. Certain facilities have paging systems that allow ED physicians to page out stroke codes to the entire stroke staff at once rather than calling each person individually. Because the 4 facilities that participated in the study did not have a paging system, the notification function of the TCS had a significant influence on the decrease in D2N time or D2P time in this study. Therefore, TCS might not be that much more advantageous in certain institutions. Second, this study did not focus on the impact of training, including the lectures and simulations. Some key processing times were already significantly reduced during the training stage before introducing TCS, suggesting that the training program was sufficiently effective in some institutions. Third, we have still not provided clear evidence for improved clinical outcomes attributed to TCS use (eg, data of ≈3‐month mRS scores were not obtained). Finally, the duration of the 3 stages was different in each facility because of seasonal or facility variations, which might have affected the results. We are planning a larger, prospective, observational study that will further clarify the impact of TCS implementation on key processing times and clinical outcomes in patients with AIS.

In conclusion, TCS application can improve hospital QM and reduce the workflow times for reperfusion therapy by facilitating parallel task completion. Furthermore, preliminary evidence suggests that implementation can improve clinical outcome.

Sources of Funding

This work was supported by a Grant‐in‐Aid for Scientific Research (JP‐16K10727, JP‐21H03173).

Disclosures

None.