Poor prediction of stroke mimics using validated stroke mimic scales in a large academic telestroke network

Abstract

Introduction

Telestroke enables timely and remote evaluation of patients with acute stroke syndromes. However, stroke mimics represent more than 30% of this population. Given the resources required for the management of suspected acute ischemic stroke, several scales have been developed to help identify stroke mimics. Our objective was to externally validate four mimic scales (Khan Score (KS), TeleStroke Mimic Score (TS), simplified FABS (sFABS), and FABS) in a large, academic telestroke network.

Methods

This is a retrospective, Institutional Review Board-exempt study of all patients who presented with suspected acute stroke syndromes and underwent video evaluation between 2019 and 2020 at a large academic telestroke network. Detailed chart review was conducted to extract both the variables needed to apply the mimic scales, the final diagnosis confirmed by final imaging, and discharge diagnosis (cerebral ischemic vs stroke mimic). Overall score performance was assessed by calculating the area under curve (AUC). Youden cutpoint was established for each scale and used to calculate sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy.

Results

A total of 1043 patients were included in the final analysis. Final diagnosis of cerebral ischemia was made in 63.5% of all patients, and stroke mimic was diagnosed in 381 patients (36.5%). To predict stroke mimic, TS had the highest AUC (68.3), sensitivity (99.2%), and NPV (77.3%); KS had the highest accuracy (67.5%); FABS had the highest specificity (55.1%), and PPV (72.5%).

Conclusions

While each scale offers unique strengths, none was able to identify stroke mimics effectively enough to confidently apply in clinical practice. There remains a need for significant clinical judgment to determine the likelihood of stroke mimic at presentation.

Keywords

Telemedicine telestroke stroke mimic stroke mimic scales ehealth

Introduction

Telestroke enables timely and remote evaluation of patients with acute stroke syndromes. Our previous literature supports that the diagnostic accuracy of cerebral ischemia is high at the initial telestroke encounter.¹ This in combination with the cost-effectiveness of the telestroke health care delivery model^2–4 supports the continued use of telestroke for the expedited assessment of patients suspected to have acute ischemic stroke syndromes.

However, stroke mimics represent up to 30% of this population.⁵ These patients with suspected acute stroke syndromes at initial evaluation but a final diagnosis of stroke mimic consume a significant amount of resources including cost and personnel to ensure timely expert evaluation and treatment recommendation.⁶ Additionally, misdiagnosis of acute ischemic stroke in the initial telestroke setting may come with additional costs including resources required to transfer patients to higher-level hospitals capable of performing the required diagnostics and interventions such as CT angiogram, MRI brain, and thrombectomy. We must also note that although the risk of intracerebral hemorrhage and additional severe side effects of IV thrombolysis is low in stroke mimics, it is still present and should be avoided if possible.^7,8

Therefore, several scales based on clinical factors available at the initial encounter have been developed to help identify stroke mimics. These include the Khan Score (KS),⁹ TeleStroke Mimic Score (TS),¹⁰ simplified FABS (sFABS),¹¹ and the FABS.¹²

These scales were validated in unique clinical settings and diverse patient cohorts. For example, although both the TS¹⁰ and FABS¹² validation included all consecutive patients being evaluated for acute stroke syndromes, TS was validated in a telestroke setting, whereas FABS was validated in patients being evaluated in-person at tertiary stroke settings. Comparatively, both sFABS¹¹ and KS⁹ are validated in patients who were evaluated in-person and ultimately received IV thrombolysis as part of acute management. Therefore, the external validation of these scales in alternative patient populations, such as the telestroke setting, is necessary to determine if these tools could be used to assist with clinical decision-making.

The objective of this study was to externally validate these four stroke mimic scales (KS, TS, sFABS, and FABS) in all consecutive patients evaluated for cerebral ischemia in a large, academic telestroke network.

Methods

This is a retrospective, Institutional Review Board-exempt study with inclusion criteria as follows: all adult consecutive patients evaluated for cerebral ischemia by video telestroke consultation in a large, academic, hub-and-spoke telemedicine network from January 2019 to December 2020.

Patients were excluded if they were <18 years old, did not have video as a part of consultation, did not have valid research authorization, or had intracerebral hemorrhage at the time of presentation.

Notably, the telestroke program at Mayo Clinic was established in 2007 to provide emergency care for suspected acute ischemic stroke cases in a hub-and-spoke model around comprehensive stroke centers in Minnesota, Florida, and Arizona. The program has grown to include 27 sites in 2021, of which 17 are Mayo Clinic Health System sites in rural Minnesota and Wisconsin. All sites have the ability to perform CT angiography, but only three sites perform CT perfusion studies. Telestroke activation criteria in our network include all patients presenting with acute neurologic deficits within 24 h from last known well. Other details concerning the structure, personnel, staffing, workflow, operations, administration, and technologies have already been published.^13–15

Detailed chart review was conducted by five individuals of the study team: three medical students, one senior neurology resident, and one vascular neurologist. A second complete review was done by the senior neurology resident to ensure initial and final diagnoses were correctly reported. The initial suspected diagnosis was extracted from the initial telestroke encounter. The final diagnosis from the discharge summary was supported by diagnostic studies and was taken as the gold standard diagnosis. Additionally, variables needed to apply the various mimic scores were also extracted including demographic data, pre-existing vascular risk factors, pertinent medical history (seizure, migraine, and psychiatric history), and NIHSS at presentation.

Cerebral ischemia was defined as AIS and transient ischemic attack (TIA), as these diagnoses are difficult to differentiate at the time of telestroke encounter, and both represent a continuum of neurovascular ischemia. All other diagnoses were defined as stroke mimics, which were subsequently coded into discrete categories.

The extracted data was applied retrospectively to the four different stroke mimic scales (TS, KS, FABS, sFABS). Overall score performance was assessed by calculating the area under curve (AUC). Both pre-established cutpoints and Youden cutpoint were used to calculate sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy. Youden cutpoint was determined by optimizing the equally weighted aggregate of sensitivity and specificity. Pre-established cutpoint was based on the original validation data of each of the various scales.^9–12. Please see the Supplementary materials for the details of each of the four scales considered. Additionally, overall score performance was compared to the performance of stroke neurologists, as demonstrated in our previous study.¹

Results

Electronic records of 1073 patients were manually reviewed. Twelve patients were excluded due to the presence of ICH on arrival, and 18 patients were excluded as their initial telestroke encounter did not have a video. Therefore, 1043 patients met inclusion criteria and were included in the final analysis. This same cohort was used in our previous study to determine the overall diagnostic accuracy of cerebral ischemia the stroke neurologist at the initial telestroke encountered.¹

Overall, average age was 69.1 years (SD 16.0), 52.1% were female, and 94.5% identified as white. Final diagnosis of cerebral ischemia was made in 63.5% of all patients (539/1043 with AIS 123/1043 with TIA). Stroke mimic was diagnosed in 36.5% of patients (381/1043). Table 1 displays the demographic data by final diagnosis.

Table 1.

Demographic data by final diagnosis.

	Final reference standard diagnosis (gold standard)
	Stroke mimic (n = 381)	Cerebral ischemia (n = 662)	Total (n = 1043)	P-value
Age at telestroke consult				<0.0001^a
Mean (SD)	63.9 (17.3)	72.1 (14.4)	69.1 (16.0)
Gender, n (%)				<0.0001^b
Male	152 (39.9%)	348 (52.6%)	500 (47.9%)
Female	229 (60.1%)	314 (47.4%)	543 (52.1%)
Race, n (%)				0.1995^b
White	355 (93.2%)	631 (95.3%)	986 (94.5%)
Black or African American	7 (1.8%)	10 (1.5%)	17 (1.6%)
Asian	5 (1.3%)	7 (1.1%)	12 (1.2%)
American Indian/Native Alaskan	2 (0.5%)	1 (0.2%)	3 (0.3%)
Pacific Islander	0 (0.0%)	1 (0.2%)	1 (0.1%)
Other	9 (2.4%)	4 (0.6%)	13 (1.2%)
Unknown/choose not to disclose	3 (0.8%)	8 (1.2%)	11 (1.1%)

Kruskal–Wallis P-value; ^bChi-square P-value.

Patients with a final diagnosis of cerebral ischemia were more likely to have pre-existing hypertension (78.5% vs 67.7% P = <0.001), hyperlipidemia (68.4% vs 50.4% P = <0.001) atrial fibrillation (23.7% vs 15.7% P = 0.0023) and coronary artery disease 29.0% vs 22.8% P = 0.03). Patients with a final diagnosis of stroke mimic were more likely to have a history of seizure (12.9% vs 5.4% P = <0.001), migraine (20.5% vs 6.2% P = <0.001, or psychiatric diagnosis (34.9% vs 18.1% P = <0.001). There was no difference in diabetes mellitus (30.2% stroke mimic vs 30.1% cerebral ischemia P = 0.97) or stroke history (19.4% stroke mimic vs 23.9% cerebral ischemia P = 0.10) between the groups. Table 2 displays baseline comorbidities by final diagnosis.

Table 2.

Baseline comorbidities by final diagnosis.

	Final reference standard diagnosis (gold standard)
	Stroke mimic(n = 381)	Cerebral ischemia(n = 662)	Total(n = 1043)	P-value
Hypertension, n (%)	258 (67.7%)	520 (78.5%)	778 (74.6%)	0.0001^a
Hyperlipidemia, n (%)	192 (50.4%)	453 (68.4%)	645 (61.8%)	<.0001^a
Diabetes mellitus n (%)	115 (30.2%)	199 (30.1%)	314 (30.1%)	0.9667^a
Tobacco use, n (%)				0.0024^a
Former	112 (29.5%)	249 (37.7%)	361 (34.7%)
Current	89 (23.4%)	105 (15.9%)	194 (18.6%)
Atrial fib., n (%)	60 (15.7%)	157 (23.7%)	217 (20.8%)	0.0023^a
Stroke history, n (%)	74 (19.4%)	158 (23.9%)	232 (22.2%)	0.0965^a
CAD, n (%)	87 (22.8%)	192 (29.0%)	279 (26.7%)	0.0302^a
Seizure history, n (%)	49 (12.9%)	36 (5.4%)	85 (8.1%)	<0.0001^a
Migraine history, n (%)	78 (20.5%)	41 (6.2%)	119 (11.4%)	<0.0001^a
Psychiatric history, n (%)	133 (34.9%)	120 (18.1%)	253 (24.3%)	<0.0001^a

Chi-square P-value.

Mean systolic blood pressure overall at presentation was 154.7 (SD 27.1), and diastolic blood pressure overall was 86.6 (SD 16.8). Systolic blood pressure was significantly higher in the cerebral ischemia group (P = 0.0047), although diastolic blood pressure was similar between groups (P = 0.6878). Mean blood glucose overall was 138.7 (SD 72.9) and was similar between the groups (P = 0.24).

Median NIHSS score at presentation was 3 (IQR 1.8) and was significantly higher in the cerebral ischemia group (P = <0.001).

Stroke mimics

The most common stroke mimics were metabolic encephalopathy (13.7%, 52/381), migraine (10.8%, 41/381), and seizure (10.5%, 40/381). Please see Table 3 for stroke mimics.

Table 3.

Stroke mimics.

Stroke mimic	Frequency
Metabolic Encephalopathy	52
Migraine	41
Seizure	40
Peripheral nerve disorders	34
Neurologic spell of unknown etiology	30
Functional Neurologic spell	27
Infection/sepsis	20
Vestibular disorder	19
Other systemic Disorder	27
Recrudescence	16
Cranial nerve disorders	15
Hypertensive emergency	15
Transient global amnesia	7
Hypoglycemia	5
Brain mass, primary	4
Dementia	4
Syncope	4
Psychiatric spell	3
Brain mass, metastatic	2
CNS infection	2
Medication side effect related	2
Orthostatic hypotension	2
RCVS	2
Alcohol-related	1
Demyelinating disease	1
Hyperglycemia	1
Primary ocular disorder	1
PRES	1
Spinal cord disorder	1
Vertebrobasilar insufficiency	1

Application of scores using pre-established cutpoint

When applying pre-established cutpoints to the stroke mimic scales, TS had the highest AUC (68.3), sensitivity (99.2%), and NPV (77.3%); KS had the highest accuracy (67.5%); FABS had the highest specificity (55.1%) and PPV (72.5%). Please see Table 4.

Table 4.

Statistical performance of stroke mimic scales using pre-established cutpoints.

Scoring System	AUC (95% CI)	Pre-established cutpoint	Sensitivity (95% CI)	Specificity (95% CI)	PPV (95% CI)	NPV (95% CI)	Accuracy (95% CI)
TS	68.3 (64.9, 71.8)	≤5	99.2 (98.2, 99.7)	4.8 (2.8, 7.5)	64.9 (61.8, 68)	77.3 (54.6, 92.2)	65.2 (62.2, 68.2)
KS	67.7 (64.1, 71.2)	>5	97.3 (95.7, 98.4)	15.7 (12.2, 19.8)	66.7 (63.7, 69.7)	76.9 (66, 85.7)	67.5 (64.6, 70.3)
sFABS	66.2 (62.8, 69.5)	≥3	79.6 (76.3, 82.7)	43.5 (38.3, 48.9)	71.5 (68, 74.8)	54.6 (48.6, 60.5)	66.6 (63.6, 69.6)
FABS	63.9 (60.5, 67.4)	≥3	66.5 (62.7, 70.2)	55.1 (49.7, 60.3)	72.5 (68.6, 76.1)	48 (43.1, 53)	62.4 (59.3, 65.4)

Application of scores using Youden cutpoint

When applying Youden cutpoints to the stroke mimic scales, TS had the highest AUC (68.3), NPV (55.3%), and accuracy (67.4%); KS had the highest specificity (58.3%) and PPV (74%); and sFABS had the highest sensitivity (79.6%). Please see Table 5.

Table 5.

Statistical performance of stroke mimic scales using Youden cutpoints.

Scoring system	AUC (95% CI)	Youden cutpoint	Sensitivity (95% CI)	Specificity (95% CI)	PPV (95% CI)	NPV (95% CI)	Accuracy (95% CI)
TS	68.3 (64.9, 71.8)	≤15	77.6 (74.1, 80.8)	49.4 (44.1, 54.8)	73.2 (69.7, 76.5)	55.3 (49.7, 60.9)	67.4 (64.4, 70.4)
KS	67.7 (64.1, 71.2)	≥3	68.4 (64.7, 72)	58.3 (53.1, 63.3)	74 (70.4, 77.5)	51.5 (46.7, 56.3)	64.7 (61.7, 67.6)
sFABS	66.2 (62.8, 69.5)	≥3	79.6 (76.3, 82.7)	43.5 (38.3, 48.9)	71.5 (68, 74.8)	54.6 (48.6, 60.5)	66.6 (63.6, 69.6)
FABS	63.9 (60.5, 67.4)	≥3	66.5 (62.7, 70.2)	55.1 (49.7, 60.3)	72.5 (68.6, 76.1)	48 (43.1, 53)	62.4 (59.3, 65.4)

Please see Figures 1–5 for ROC curves.

Figure 1.

ROC curves comparing all stroke mimic scores and expert.

Figure 2.

ROC curve for FABS.

Figure 3.

ROC curve for sFABS.

Figure 4.

ROC curve for TS.

Figure 5.

ROC curve for Khan Score.

Comparison of scales to stroke neurologist (experts)

Based on our previous publication which determined the overall diagnostic accuracy of the stroke neurologist to diagnose cerebral ischemia,¹ we were able to determine the diagnostic accuracy of the stroke neurologist to diagnose stroke mimic at the initial telestroke encounter. It was determined that the stroke neurologists had the highest AUC (89.4), sensitivity (97.1%), specificity (81.4%), PPV (90.1%), NPV (94.2%), and accuracy (91.4%). The stroke neurologist had a statistically significant higher AUC compared to all scales on contrast analysis (P < 0.001). Please see Tables 6 and 7.

Table 6.

Contrast analysis comparing scales to expert.

Scoring system	AUC (95% CI)	Contrast (95% CI)	P-value
TS	68.3 (64.9, 71.8)	−21.1 (−25.1, −17.1)	<0.0001
KS	67.7 (64.1, 71.2)	−21.7 (−25.8, −17.6)	<0.0001
sFABS	66.2 (62.8, 69.5)	−23.2 (−27.1, −19.4)	<0.0001
FABS	63.9 (60.5, 67.4)	−25.5 (−29.5, −21.5)	<0.0001
Expert	89.4 (87.3, 91.5)	Reference

Table 7.

Statistical performance of stroke mimic scales using pre-established cutpoints versus experts.

Scoring system	AUC (95% CI)	Pre-established cutpoint	Sensitivity (95% CI)	Specificity (95% CI)	PPV (95% CI)	NPV (95% CI)	Accuracy (95% CI)
TS	68.3 (64.9, 71.8)	≤5	99.2 (98.2, 99.7)	4.8 (2.8, 7.5)	64.9 (61.8, 68)	77.3 (54.6, 92.2)	65.2 (62.2, 68.2)
KS	67.7 (64.1, 71.2)	>5	97.3 (95.7, 98.4)	15.7 (12.2, 19.8)	66.7 (63.7, 69.7)	76.9 (66, 85.7)	67.5 (64.6, 70.3)
sFABS	66.2 (62.8, 69.5)	≥3	79.6 (76.3, 82.7)	43.5 (38.3, 48.9)	71.5 (68, 74.8)	54.6 (48.6, 60.5)	66.6 (63.6, 69.6)
FABS	63.9 (60.5, 67.4)	≥3	66.5 (62.7, 70.2)	55.1 (49.7, 60.3)	72.5 (68.6, 76.1)	48 (43.1, 53)	62.4 (59.3, 65.4)
EXPERT	89.4 (87.3, 91.5)		97.1 (95.6, 98.3)	81.4 (77.1, 85.1)	90.1 (87.6, 92.2)	94.2 (91.1, 96.5)	91.4 (89.5, 93)

Discussion

Our data demonstrates that although each stroke mimic scale offers unique strengths, none was able to identify stroke mimics effectively enough to confidently apply in routine clinical practice. When comparing the various stroke mimic scales to the expert, the expert is able to make the diagnosis of stroke mimic more accurately across all measures. This is likely a result of the ability of the stroke neurologist to integrate various points of data available in the clinical setting, which may not be included in the stroke mimic scales, to form a final suspected diagnosis. Poon et al. who also evaluated the diagnostic accuracy of stroke mimic at the initial telestroke encounter found similar PPV (85%) and NPV (90%) to what we report (PPV 90%, NPV 94%).¹⁶ This underscores the importance of the need for these telestroke mimic scales to remain a part of a set of clinical decision-making tools that assists the physician rather than be used alone to make a diagnosis. As the use of telestroke grows, we may consider the use of stroke mimic scales as a strategy to triage consultations from spokes to the hub hospital in order to optimize stroke neurologist workflows. The high accuracy of our experts compared to stroke mimic scales to diagnose a stroke mimic is likely also related to the subspecialty training of the vascular neurologists included in our telestroke system. Therefore, the use of stroke mimic scales may also be considered as a clinical decision-making tool in areas where stroke neurologists are not readily available.

The stroke mimic rate in the validation cohorts of the stroke mimic scales are as follows: sFABS 12.8%, FABS 41%, SM 22.9%, and Khan 25%. Our stroke mimic rate (36.5%) is higher than most of the validation cohorts, except for sFABS. sFABS likely had the lowest number of stroke mimics as they only included patients who received tPA within the 4.5-h window, which likely meant the treating physician had a low suspicion that the patient's clinical presentation is likely due to a stroke mimic. Our higher stroke mimic rate is likely a result of our study being real-life and including all consecutive patients presenting with acute neurologic deficits within 24 h from last known well.

In our external telestroke cohort, TS achieved the highest AUC (68.3), sensitivity (99.2%), and NPV (77.3%). This may be because both of our cohorts include video-based telestroke evaluations of patients presenting with suspected acute ischemic syndromes. Therefore, validation of the TS score in various external telestroke populations, especially those without access to a vascular neurologist 24/7, may serve as a triage tool to determine which patients should be seen more urgently at a higher-level center.

Tu et al. externally validated the performance of these four mimic scales in patients who received intravenous thrombolysis for suspected acute ischemic stroke in the telestroke setting.¹⁷ Their results similarly note that TS had the highest AUC (0.75 95% CI = 0.63–0.8), which was higher than our AUC of 0.68.3. They also note TS had the highest sensitivity (91.3%) and KS had highest specificity (88.2). This contrasts with our data where TS had the highest sensitivity (99.2%) and FABS had the highest specificity (55.1%). The differences in our results, despite both including telestroke populations, are likely a result of our cohort including all patients with neurologic deficits within 24 h of symptom onset. It is likely that within the cohort of Tu et al., many obvious stroke mimics in which a neurologist had excluded prior to consideration of IV tPA had already been eliminated from the cohort, which skews the results. Notably, their stroke mimic rate was significantly lower than that in our cohort (6.6% vs 36.5%).

To our knowledge, this is the first external validation cohort for the included stroke mimic scales of all patients presenting within 24 h of last known normal in the telestroke setting. The limitations of our study include its retrospective design. It should also be noted that our results cannot be generalized to other patient populations, including those who are evaluated in-person, past the 24-h window or those evaluated by non-vascular trained neurologists. It is also unclear if any of the stroke neurologists used these stroke mimic scales as part of their evaluation, as this was not particularly asked about or retrieved from the chart analysis.

Given the significant resources required for suspected acute ischemic stroke patients at the initial telestroke encounter, there is a need for optimizing the diagnosis of stroke mimic. At this time, our data supports that the most accurate diagnosis of stroke mimic is made by the stroke neurologist as opposed to any individual stroke mimic scale. However, future research evaluating the usefulness of these scales to supplement clinicians’ skills and to triage patients in busy stroke centers, hub-and-spoke telestroke networks, or in regions without stroke neurologists, should be considered. There remains a need for a significant clinical judgment to determine the likelihood of stroke mimic at presentation.

Supplemental Material

sj-docx-1-jtt-10.1177_1357633X241273762 - Supplemental material for Poor prediction of stroke mimics using validated stroke mimic scales in a large academic telestroke network

Supplemental material, sj-docx-1-jtt-10.1177_1357633X241273762 for Poor prediction of stroke mimics using validated stroke mimic scales in a large academic telestroke network by Nikita Chhabra, Stephen W English, Richard J Butterfield, Nan Zhang, Abigail E Hanus, Rida Basharath, Monet Miller and Bart M Demaerschalk in Journal of Telemedicine and Telecare

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Nikita Chhabra

Stephen W English

Data availability statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Supplemental material

Supplemental material for this article is available online.

References

Chhabra

English

Miller

, et al. Demonstration of high diagnostic accuracy for cerebral ischemia in a large academic hub-and -spoke telestroke network. Mayo Clinic Proc: Digital Health 2023; 1: 490–497.

Demaerschalk

Hwang

H-M

Leung

. Cost analysis review of stroke centers, telestroke, and rt-PA. Am J Manag Care 2010; 16: 537–544.

Nelson

Saltzman

Skalabrin

, et al. The cost-effectiveness of telestroke in the treatment of acute ischemic stroke. Neurology 2011; 77: 1590–1598.

Nelson

Okon

Lesko

, et al. The cost-effectiveness of telestroke in the pacific northwest region of the USA. J Telemed Telecare 2016; 22: 413–421.

Pohl

Hesszenberger

Kapus

, et al. Ischemic stroke mimics: A comprehensive review. J Clin Neurosci 2021; 93: 174–182.

Goyal

Male

Al Wafai

, et al. Cost burden of stroke mimics and transient ischemic attack after intravenous tissue plasminogen activator treatment. J Stroke Cerebrovasc Dis 2015; 24: 828–833.

Nguyen

Chang

. Stroke mimics and acute stroke evaluation: Clinical differentiation and complications after intravenous tissue plasminogen activator. J Emerg Med 2015; 49: 244–252.

Ali-Ahmed

Federspiel

Liang

, et al. Intravenous tissue plasminogen activator in stroke mimics. Circ Cardiovasc Qual Outcomes 2019; 12: e005609.

Khan

Chaku

Goehl

, et al. Novel algorithm to help identify stroke mimics. J Stroke Cerebrovasc Dis 2018; 27: 703–708.

10.

Ali

Viswanathan

Singhal

, et al. The TeleStroke mimic (TM)-score: A prediction rule for identifying stroke mimics evaluated in a telestroke network. J Am Heart Assoc 2014; 3: e000838.

11.

Qin

Zhao

Yin

, et al. Validation of simplified FABS scale to predict stroke mimics in a Chinese population undergoing intravenous thrombolysis. Clin Neurol Neurosurg 2017; 161: 1–5.

12.

Goyal

Tsivgoulis

Male

, et al. FABS: An intuitive tool for screening of stroke mimics in the emergency department. Stroke 2016; 47: 2216–2220.

13.

Huang

Greenway

MRF

Nasr

, et al. Telestroke in the time of COVID-19: The Mayo Clinic experience. Mayo Clin Proc 2020; 95: 1704–1708.

14.

Dumitrascu

Demaerschalk

. Telestroke. Curr Cardiol Rep 2017; 19: 85.

15.

Kreofsky

BLH

Blegen

Lokken

, et al. Sustainable telemedicine: Designing and building infrastructure to support a comprehensive telemedicine practice. Telemed J E Health 2018; 24: 1021–1025.

16.

Poon

Tkach

de Havenon

, et al. Telestroke consultation can accurately diagnose ischemic stroke mimics. J Telemed Telecare 2023: 29: 444–450.

17.

Tan

Saffari

, et al. External validation of stroke mimic prediction scales in the emergency department. BMC Neurol 2020; 20: 269.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.38 MB