Clinical Practice Guideline for the Prehospital Stage in Acute Stroke : I. Use of Emergency Medical Services Assessment Tools
Article information
, Dongwook Seo2,*, Jinwoo Jeong3, Kyoung-Chul Cha4, Yong Soo Cho5, Su Jin Kim6, Jongkyu Park7, Won-Sang Cho8, Se Won Oh9, Jang Hun Kim10, Hyeong Jin Lee11, Hong Suk Ahn12, Yuna Jo13, Jung-Jae Kim14, Kyoung Min Jang15, Gi-Yong Yun16, Jong Min Lee17, Hoon Kim11, Young Woo Kim1, Tae Gon Kim18, Sung-kon Ha19, Sukh Que Park2, Soon Chan Kwon,17
Abstract
Accurate and early identification of stroke and large vessel occlusion (LVO) in emergency settings is essential for improving patient outcomes and ensuring the efficient allocation of medical resources. This clinical practice guideline systematically reviews domestic and international literature and conducts meta-analyses to evaluate the utility and diagnostic accuracy of stroke assessment tools used in prehospital emergency medical services (EMS). We developed a guideline based on evidence from systematic reviews and meta-analyses via a de novo process. A systematic literature review was conducted to evaluate the usefulness of diagnostic EMS assessment tools for diagnosing stroke and LVO. Overall, 70 non-randomized control studies were selected for this study. A meta-analysis was conducted with a subgroup analysis to distinguish between patients with stroke and those with LVO. EMS tools demonstrated high sensitivity but low specificity for diagnosing stroke. In the prehospital setting, using validated EMS stroke assessment tools is recommended for the early identification of stroke and LVO. Upon hospital arrival, stroke specialists should conduct further evaluation and triage to confirm the diagnosis and guide appropriate management. Delays in diagnosing LVO are frequently unacceptable. While experts advocate for the use of EMS assessment tools to facilitate early identification of LVO, these tools alone lack adequate sensitivity. Therefore, further diagnostic evaluations and consultation with stroke specialists upon hospital arrival are recommended.
INTRODUCTION
Early identification of stroke in an emergency setting is crucial for improving patient outcomes. However, in real-world situations, first responders often face challenges in determining symptoms that strongly suggest stroke, especially when large vessel occlusion (LVO) is involved. Missing critical cases can cause severe harm to patients. Moreover, even after arriving at the hospital, patients with stroke may need to be transferred to another treatment-capable center, causing significant delays in care.
Consequently, various emergency medical service (EMS) assessment tools have been developed to enhance the detection of stroke and LVO in prehospital settings. Additionally, numerous studies have evaluated the diagnostic accuracy of these tools; however, the results have been inconsistent. Therefore, a systematic review and meta-analysis is essential to assess the validity of these tools and determine whether they should be recommended for use in identifying stroke and LVO in emergency settings.
In Korea, emergency department overcrowding has become a major social issue, particularly in tertiary hospitals. These circumstances cause patients with mild stroke to be triaged alongside severe cases, contributing to treatment delays for the more critical cases. This highlights a gap in the allocation of medical resources. Ideally, suspected severe stroke cases should be transported directly to stroke centers, while non-urgent patients should be directed to regional hospitals. In cases of acute cerebral infarction with LVO, timely thrombolysis and mechanical thrombectomy are essential for effective treatment. Recent studies have actively explored the mothership and drip-and-ship models [10,17]. According to our meta-analysis, direct transport to a thrombectomy-capable hospital has been associated with improved functional outcomes. As the rapid identification of LVO becomes increasingly critical and important, the diagnostic accuracy of EMS assessment tools for detecting LVO has been reported. However, owing to the lack of meta-analyses and formal guidelines on the diagnostic accuracy of these tools, further research is warranted.
Through this guideline, we aimed to promote education and awareness among the public, emergency responders, and healthcare professionals, thereby facilitating appropriate resource allocation and clinical decision making. This study provides a practical and evidence-based guideline tailored to various clinical and regional settings in Korea. Particularly, we sought to establish a national protocol that would be continuously updated in accordance with the emerging evidence and changes in healthcare delivery systems by systematically reviewing the current literature and incorporating expert consensus.
MATERIALS AND METHODS
Clinical practice guideline development process
Two key questions were established, and two recommendations were derived. For areas where recommendations could not be drawn, a systematic review was conducted to summarize the findings. Continuous revisions and further guideline development will be performed in subsequent phases. We reported this guideline in accordance with Institutional Review Board approval (UC24ZISE0069), and the requirement for informed consent was waived. Further details of the participating members and progress in developing this guideline are provided in Supplementary Material 1.
Rationale for developing the guideline process
Stroke treatment is categorized into the following three phases : prehospital care, hospital treatment, and post-treatment management. Most of the previously established Korean guidelines have focused on hospital-based treatment, and many were created by modifying or adapting existing international guidelines. In healthcare systems, overcrowded emergency rooms have a risk of delayed intervention for patients with stroke and LVO. Consequently, rapid triage—particularly the prompt recognition of suspected stroke at first medical contact—and immediate transfer to a designated stroke center are critical. We propose a set of prehospital guidelines that evaluate the initial diagnostic accuracy and feasibility of stroke assessment tools in emergency settings.
Selection of key questions
The selection of key questions for the clinical practice guideline topic began with a survey conducted among all members of the society. Following the survey and the collection of topics by the working committee, two rounds of meetings among the development committee members were held to finalize the key questions. The key questions for each recommendation were formulated on the basis of the Population, Intervention, Comparison, Outcome (PICO) framework, and a systematic literature search was conducted accordingly. Our key questions were as follows : early and appropriate intervention is critical in managing stroke; thus, using validated EMS assessment tools is recommended to facilitate the early identification of patients with stroke and support timely and appropriate treatment decisions. 1) Is the use of EMS diagnostic tools for stroke assessment at the prehospital stage beneficial in emergency situations? And 2) is the use of EMS diagnostic tools for LVO assessment at the prehospital stage beneficial in emergency situations? We evaluated the effectiveness of the EMS assessment tools in undiagnosed emergency patients before hospital arrival. The final diagnosis was confirmed after hospital admission. The intervention group comprised patients assessed using the EMS assessment tools, whereas the control group included those who were not assessed with the EMS assessment tools. The outcomes were as follows : diagnostic accuracy of the final stroke diagnosis, including sensitivity and specificity; and diagnostic performance for identifying LVO, which was assessed using sensitivity and specificity. A summary of each key question and its corresponding recommendation is provided in Table 1.
Summary of the key questions and recommendations
Literature search and selection
The literature search was conducted through discussions between methodology experts and the development committee members. An initial set of search terms was formulated, and a search strategy was established accordingly. Based on this strategy, searches were conducted across the following four major databases : MEDLINE (PubMed), Embase, Cochrane Library, and KoreaMed. The search was performed on November 11, 2024. The detailed search strategy is provided in Supplementary Material 1.
Relevant studies were selected separately for each key question, with three development committee members assigned depending on the volume of literature available for each topic. Duplicate exclusion and literature selection followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses flowchart (Fig. 1). The inclusion and exclusion criteria for each key question were formulated and applied on the basis of the PICO framework.
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow of the systematic review and meta-analysis. LVO : largevessel occlusion.
Publication and future updates of the clinical practice guideline
This clinical practice guideline will be published in medical journals affiliated with the Korean Neurological Society, Korean Society of Emergency Medicine, Korean Society of Cerebrovascular Surgeons, and Korean NeuroEndovascular Society (KoNES). Additionally, it will be made publicly available on the society’s official website, Instagram, and other online platforms, ensuring that physicians, healthcare professionals, and the public can easily download and access the guideline.
Furthermore, the guideline will be registered on the official website of KoNES for promotional purposes. A summary booklet will also be produced and actively distributed at future academic conferences and events.
The KoNES Academic Committee and the Clinical Practice Guideline Committee will continue to monitor research findings in specific areas, including early diagnosis and transfer of emergency patients with stroke and treatment selection for subarachnoid hemorrhage. Additionally, the committee will track updates from international clinical guidelines. If new research accumulates, warranting additions or modifications to the recommendations, a Clinical Guideline Revision Committee under the society will be formed to revise the guideline every 5 years.
RESULTS
We searched the literature to identify articles evaluating the usefulness of the diagnostic tools for distinguishing stroke at the prehospital stage. Ultimately, 70 articles were selected for this study. Of these, 15 publications focused on stroke as the final diagnosis, and 55 were specific to LVO. Most studies were highly heterogeneous. A meta-analysis was conducted with a subgroup analysis to distinguish between stroke- and LVO-related conditions.
Early selection and diagnosis of patients with stroke
Using EMS assessment tools for stroke diagnosis at the prehospital stage is beneficial. When suspected stroke occurs before hospital arrival, it is recommended to use a stroke diagnostic screening tool in the prehospital setting for early selection and diagnosis of patients with stroke (level of evidence [LOE], moderate; grade of recommendation [GOR], A).
Fifteen studies were found to use prehospital tools for stroke diagnosis. Overall, the use of these tools improved sensitivity but did not enhance the specificity of the diagnosis. The sensitivity and specificity for stroke diagnosis were 0.83 (95% confidence interval [CI], 0.72–0.90; Fig. 2) and 0.61 (95% CI, 0.48–0.73; Fig. 3), respectively. Thus, the tools demonstrated high sensitivity but low specificity in diagnosing stroke. The overall area under the curve (AUC) was 0.775, with moderate heterogeneity observed across studies (I2=35.5%; Fig. 4).
Meta-analysis of sensitivity for diagnosing stroke. GLMM : generalized linear mixed model, CI : confidence interval, FAST : Face Arm Speech Test, NA : not available, CPSS : Cincinnati prehospital stroke scale, LAMS : Los Angeles motor scale.
Meta-analysis of specificity for diagnosing stroke. GLMM : generalized linear mixed model, CI : confidence interval, FAST : Face Arm Speech Test, NA : not available, CPSS : Cincinnati prehospital stroke scale, LAMS : Los Angeles motor scale.
Summary receiver operating characteristic (SROC) of emergency medical service assessment tools for diagnosing stroke and large vessel occlusion. LVO : large vessel occlusion.
Regarding sensitivity, the Face Arm Speech Test (FAST) and the Cincinnati prehospital stroke scale (CPSS) showed the highest values of 0.87 and 0.79, respectively, whereas the Los Angeles motor scale (LAMS) had a sensitivity of 0.68. The LAMS showed the highest specificity at 0.84, followed by CPSS at 0.59 and FAST at 0.57. A higher sensitivity indicates a lower likelihood of missing patients with stroke during the initial assessment. Higher specificity helps to correctly exclude patients without stroke and minimize false positives. Therefore, tools with high sensitivity, including FAST and CPSS, are advantageous for identifying stroke in the prehospital setting. Conversely, the LAMS is more useful to accurately rule out patients without stroke owing to its higher specificity.
The FAST and CPSS assess the presence of facial palsy, arm weakness, and speech disturbance. By contrast, the LAMS assigns scores based on arm drift and grip strength. These differences in tool compositions may result in varying sensitivities and specificities for stroke diagnosis.
In clinical settings, misdiagnosing a patient with stroke during an emergency can lead to a poor prognosis, and a high number of false positives can result in emergency department overcrowding and the unnecessary consumption of medical resources. However, considering that patients with stroke are often not transferred to hospitals equipped with emergency medicine, neurology, or neurosurgery departments, and that early diagnosis and proper treatment significantly improve outcomes, we concluded that EMS assessment tools in the prehospital stage are appropriate for stroke diagnosis. Furthermore, based on experts’ opinions, EMS tools with higher sensitivity were deemed to be more suitable for use in prehospital settings. The result of the tool depends on the characteristics of each tool, resulting in varying sensitivity and specificity. Since emergency department conditions and regional settings also differ, the appropriate tool should be selected on the basis of the local context. When possible, tools with a higher sensitivity are preferred.
Early selection and diagnosis of patients with LVO
Using EMS assessment tools to diagnose LVO at the prehospital stage is beneficial. However, in clinical settings, delays in diagnosing LVO are frequently unacceptable. Owing to the diverse clinical presentations of LVO, relying solely on prehospital EMS assessment tools may be insufficient for accurate diagnosis. While experts advocate for the use of EMS assessment tools to facilitate early identification of LVO, these tools alone lack adequate sensitivity. Further diagnosis should be made using the National institutes of health stroke scale (NIHSS) and medical imaging findings after hospital arrival, and evaluations and consultation with specialists are strongly recommended (LOE, moderate; GOR, B).
Our meta-analysis showed that the usefulness of prehospital diagnostic tools for LVO diagnosis had a sensitivity and specificity of 0.68 (95% CI, 0.64–0.72; Fig. 5) and 0.78 (95% CI, 0.75–0.81; Fig. 6), respectively. Overall, the sensitivity of the LVO diagnosis was low, whereas its specificity was high. The overall AUC was 0.8, with low heterogeneity observed across studies (I2=13%). This indicates that the EMS assessment tools maintain high diagnostic accuracy within the clinically relevant range of specificity. EMS assessment tools for stroke (AUC, 0.76) and LVO (AUC, 0.80) showed good diagnostic accuracy, supporting their use in various prehospital settings. Therefore, these tools should be employed to aid early identification and appropriate triage in any suspected stroke case (Fig. 4). Notably, LVO has been evaluated using various scales because it is a highly demanded topic in clinical practice. Therefore, a subgroup analysis was performed.
Meta-analysis of sensitivity for diagnosing large vessel occlusion. GLMM : generalized linear mixed model, CI : confidence interval, CPSS : Cincinnati prehospital stroke scale, NA : not available, PASS : Prehospital acute stroke severity, RACE : Rapid Arterial oCclusion Evaluation, LAMS : Los Angeles motor scale, FAST-ED : Field Assessment Stroke Triage for Emergency Destination, VAN : Vision, aphasia, neglect, FAST : Face Arm Speech Test.
Meta-analysis of specificity for diagnosing large vessel occlusion. GLMM : generalized linear mixed model, CI : confidence interval, CPSS : Cincinnati prehospital stroke scale, NA : not available, PASS : Prehospital acute stroke severity, RACE : Rapid Arterial oCclusion Evaluation, LAMS : Los Angeles motor scale, FAST-ED : Field Assessment Stroke Triage for Emergency Destination, VAN : Vision, aphasia, neglect, FAST : Face Arm Speech Test.
The CPSS showed the highest sensitivity at 0.91, indicating the greatest potential for correctly identifying patients with LVO. In contrast, Rapid Arterial oCclusion Evaluation (RACE) and Field Assessment Stroke Triage for Emergency Destination (FAST-ED) had moderate sensitivities of 0.69 and 0.72, respectively, suggesting a reasonable level of reliability. LAMS had the highest specificity at 0.85, followed by Prehospital acute stroke severity (PASS) (0.80), FAST-ED (0.80), and RACE (0.80), all of which showed high specificity. Conversely, the CPSS, which had the highest sensitivity, showed the lowest specificity of 0.27, indicating a high likelihood of failing to exclude patients without LVO. Therefore, it may produce a high rate of false positives. The characteristics of the disease and the clinical setting should be considered before recommending the use of diagnostic tools. LVO requires prompt and aggressive treatment because it has a poor prognosis. Consequently, patients are preferably directly transported to a stroke center capable of performing thrombolysis or thrombectomy. Prehospital diagnostic tools are more likely to be used by trained paramedics or nonmedical personnel rather than physicians. If a scale is relatively complex or includes several items, its practicality decreases; hence, simpler scales are preferred. Consequently, we recommend tools that have high sensitivity and are easy to use. However, these tools tend to have high false-positive rates. After reviewing expert opinions, we concluded that it is reasonable to prioritize the use of highly sensitive tools for LVO diagnosis. Moreover, combining these tools in clinical settings, when possible, would be beneficial owing to the limitations of each tool. For example, combining a highly sensitive tool such as CPSS with a highly specific tool, including LAMS or prioritizing the use of tools such as FAST-ED or RACE, which have balanced sensitivity and specificity, could enhance the accuracy of LVO diagnosis.
Based on the meta-analyzed studies and our emergency department data, the following cutoff values were considered for LVO diagnosis. If each EMS assessment tool exceeds the following cutoff values, the patient can be considered to have a high suspicion of LVO or a high risk of stroke at the point of emergency occurrence. A CPSS score of 1 point, Vision, aphasia, neglect (VAN) score ≥1, Cincinnati prehospital stroke severity scale (CPSSS) and PASS scores ≥2, FAST-ED and LAMS scores ≥4, and RACE score ≥5 can be considered indicative of stroke or LVO. In certain regions of South Korea, there are established protocols that mandate the direct transport of patients with suspected stroke who present with high prehospital stroke severity scores to thrombectomy-capable centers. Differentiating the destination hospital based on the prehospital stroke severity score is intended to expedite the identification and treatment of LVO.
According to the consensus of all experts, relying solely on prehospital diagnostic tools to identify and differentiate LVO from other conditions is clinically challenging and may pose a significant risk if misdiagnosed. Since patients with LVO can present with various clinical symptoms, it is essential to rule out other conditions, including hemorrhagic diseases or transient ischemic attack, before confirming the diagnosis of LVO. Therefore, an LVO diagnosis should be accompanied by additional diagnostic tools and imaging findings after hospital arrival.
Consequently, although the use of prehospital diagnostic tools can be recommended for initial suspicion of LVO, their standalone accuracy is likely to be limited. Therefore, a precise diagnosis should be made using the NIHSS and medical imaging after hospital arrival, and prompt and aggressive treatment should be provided within the therapeutic time window (LOE, moderate; GOR, B).
Characteristics and results of the EMS assessment tools
CPSS
The CPSS, first introduced by Kothari in 1999, is a screening tool developed to screen for severe stroke and LVO, derived from selected components of the NIHSS [6,7]. The CPSS was originally developed to identify patients with suspected stroke who may be eligible for thrombolytic therapy, and to provide an initial assessment of stroke severity. In the initial study, a score of ≥1 was proposed as indicative of possible stroke. The CPSS is also useful for the prehospital identification of LVO [16]. Subsequently, various studies have investigated the optimal cutoff value, with most recommending a threshold of 2 or 3 points for LVO suspicion [16,20]. A key advantage of the CPSS is its simplicity and feasibility, making it particularly suitable for use by EMS personnel in prehospital settings [16].
The CPSS assesses the following three clinical findings : facial palsy, arm drift, and speech disturbances. Each component has a score of 1 point, yielding a total score ranging from 0 to 3. The CPSSS comprises the following three items : gaze deviation scored as 2 points; if the patient answers at least one of the two orientation questions (age and current month) incorrectly and fails at least one of the two commands (close your eyes and make a fist), as 1 point; and inability to hold the arm up for >10 seconds (NIHSS ≥2) is scored as 1 point. A total score of ≥2 is considered positive for stroke and LVO. The CPSSS was retrospectively derived from a cohort of 624 patients from the two National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group studies and was validated using a cohort of 650 patients from the Interventional Management of Stroke (IMS) III trial [2,11]. In patients with severe stroke defined as NIHSS ≥15, the CPSSS showed sensitivity, specificity, and AUC of 92%, 51%, and 0.83, respectively. In contrast, for patients with LVO, it demonstrated sensitivity, specificity, and AUC of 83%, 40%, and 0.67, respectively [6]. Although the CPSSS shows high sensitivity, it has limitations, including a low AUC for LVO detection and the exclusion of isolated M2 occlusions in its derivation and validation cohorts.
FAST
The FAST was developed in the United Kingdom in 1998 as a stroke identification instrument and was later established as a stroke scale [4]. The FAST is one of the earliest tools proposed during the initial development of stroke scales and was designed to identify patients with stroke who may benefit from thrombolytic therapy [4]. Along with the CPSS and Los Angeles prehospital stroke screen (LAPSS), FAST has been widely used in various studies comparing stroke assessment tools. Owing to the nature of its components, FAST is advantageous for identifying anterior circulation stroke. With the growing emphasis on the identification of LVO, particularly in the context of the expanded use of endovascular treatment for ischemic stroke, modified versions of the FAST have emerged. These include the G-FAST, which incorporates gaze evaluation for improved detection of posterior circulation stroke [3], and FAST-ED, which includes eye deviation and neglect as assessment items [3]. The FAST evaluates the following three symptoms : facial palsy, arm weakness, and speech impairment. Any of these findings are considered FAST positive.
VAN
The VAN scale was introduced by Teleb in 2016 as a screening tool specifically developed for emergent LVO detection [18]. The VAN scale assesses the functional neurovascular anatomy and comprises the following four components : arm weakness, visual disturbance, aphasia, and neglect. Arm weakness is classified as absent, mild, moderate, or severe. However, only the presence or absence of weakness is considered in the final assessment. Visual disturbances are evaluated based on the presence of field cuts, double vision, or sudden blindness. Aphasia can be categorized as expressive, receptive, or mixed. Neglect is assessed using indicators such as forced gaze, the inability to feel both sides, or ignoring one side. A VAN-positive result was defined as the presence of arm weakness along with at least one abnormal finding in either vision, aphasia, or neglect. The VAN test is simple to administer, yields dichotomous (positive/negative) results, and can be performed rapidly. Consequently, initial studies have demonstrated the inter-rater reliability of this tool [12]. VAN is classified as a cortical symptom-based stroke scale, similar to FAST-ED and RACE [1,12]. The VAN may have higher sensitivity but potentially lower specificity than screening tools that do not assess cortical signs. Some studies have suggested that its dichotomous structure contributes to its reduced specificity [12].
LAMS
The LAMS was developed by Llanes in 2004 as a simplified modification of the LAPSS, aiming to identify patients with stroke and simultaneously assess stroke-related motor deficits [9]. Initially, no specific cutoff value was proposed; however, it was noted to correlate well with the NIHSS. Later, in 2008, Nazliel suggested that a LAMS score of 4 or more could indicate suspected LVO, and subsequent studies have reported similar findings [13,14]. LAMS evaluates the following three components : facial droop, arm drift, and grip strength. Facial droop was scored as either 0 or 1, based on its presence. Arm drift and grip strength were scored from 0 to 2, depending on their presence and severity. The total score ranged from 0 to 5.
RACE
The RACE scale was developed to screen for LVO [15] and comprises the following five components : facial palsy is scored as follows : 0 points, absent; 1 point, if mild (NIHSS, 1); and 2 points, if moderate-to-severe (NIHSS, 2–3); arm motor function is scored as follows : 0, if absent or mild (NIHSS, 0–1), 1 point if moderate (NIHSS, 2), and 2 points if severe (NIHSS, 3–4); leg motor function is evaluated using the same criteria and scoring system as those for the arm; gaze deviation is scored as 1 point if present (NIHSS, 1–2); and, language and neglect are assessed depending on the side of hemiparesis. Aphasia is evaluated using two commands (close your eyes and make a fist) for right-sided weakness : 0 points if both are correctly performed, 1 point if only one is completed, and 2 points if neither is performed. Agnosia is assessed using two tasks (“Whose arm is this?” while showing the paralyzed arm, and “Clap your hands” while raising both arms) for left-sided weakness : 0 points if both are answered correctly, 1 point if only one is answered correctly, and 2 points if neither is answered correctly. A total score of ≥5 is considered positive for LVO. The RACE scale was retrospectively derived from the data of 654 patients admitted to a stroke unit at a single institution between January 2006 and March 2010, and prospectively validated in a cohort of 885 patients admitted between February 2011 and March 2013. The RACE scale demonstrated sensitivity, specificity, and AUC of 85%, 68%, and 0.82, respectively, for diagnosing LVO. Despite high sensitivity, nearly half of the LVO diagnoses were based on transcranial duplex ultrasound, which may have reduced diagnostic accuracy.
FAST-ED
The FAST-ED scale was developed as a screening tool for LVO [8] and comprises the following five components : facial palsy scored as 0 points if absent or mild (NIHSS, 0–1) and 1 point if partial or complete (NIHSS, 2–3); arm weakness scored as 0 points if absent (NIHSS, 0), 1 point if the arm cannot be held up but can resist gravity (NIHSS, 1–2), and 3 points if the arm cannot overcome gravity (NIHSS, 3–4); speech impairment scored as 0 points if absent (NIHSS, 0), 1 point if mild-to-moderate (NIHSS, 1), or 2 points if severe (NIHSS, 2–3); gaze deviation scored as 0 if absent (NIHSS, 0), 1 point if partial (NIHSS, 1), or 2 points if complete (NIHSS, 2); and neglect scored as 0 if absent (NIHSS, 0), 1 point if extinction is present on double simultaneous stimulation (NIHSS, 1), or 2 points if the patient is unaware of one side or direction (NIHSS, 2). A total score of ≥4 is considered positive for LVO. The FAST-ED scale was derived from the data of 741 patients enrolled in the Screening Technology and Outcomes Project in Stroke (STOPStroke) cohort and demonstrated sensitivity, specificity, and AUC of 61%, 89%, and 0.813, respectively. Although the FAST-ED scale is characterized by a high specificity, its low sensitivity is a limitation. One advantage of this scale is that it stratifies the LVO risk into three distinct groups based on the score as follows : a score of 0–1 corresponds to an LVO probability of <15%, 2–3 to approximately 30%, and ≥4 to approximately 60%.
PASS
The PASS scale was developed as a screening tool for LVO [5] and comprises the following three items : the patient is given 1 point if they fail to answer either their age or the current month; gaze palsy or deviation is scored 1 point; and arm weakness is scored as 1 point if present (NIHSS ≥1). A total score of ≥2 is considered positive. The PASS scale was retrospectively derived from two-thirds of a dataset of 5864 patients registered in the Danish Stroke Registry, validated using the remaining one-third of patients [19]. The PASS scale demonstrated sensitivity, specificity, and AUC of 61%, 83%, 0.73, respectively, and an odds ratio of 7.63 (95% CI, 5.71–10.20) [5] for LVO prediction. Although the PASS scale has high specificity, its low sensitivity limits its use as a screening tool.
Prehospital use of EMS assessment tools in the diagnosis and triage of stroke
It is impossible to accurately diagnose stroke or LVO before hospital arrival. Owing to the nature of these conditions, delayed diagnosis, misdiagnosis, or delayed transfer to a specialized stroke center may have serious consequences for both patients. Therefore, these guidelines aim to provide practical recommendations to members of the public, emergency medical personnel, hospital staff, and government health authorities for using stroke diagnostic tools during emergency transportation. In conclusion, among the currently available tools, some demonstrate high sensitivity, whereas others show reasonably balanced sensitivity and specificity. Consequently, we strongly recommend the use of EMS assessment tools to aid in the early identification of stroke at the prehospital stage (GOR, A). Additionally, tools for the prehospital identification of LVO may be utilized; however, owing to the inherent difficulty in diagnosing LVO and the wide variability in patient presentation, such tools should not be solely relied upon. Rather, their use should be followed by further evaluation using other validated clinical and imaging modalities upon arrival at the hospital (GOR, B). If a stroke or LVO is strongly suspected in an emergency patient, the patient should be transported to a designated stroke center. We aim to provide emergency medical personnel with proper information on hospitals capable of receiving and treating patients with stroke.
Based on our research, currently available stroke diagnostic tools have some limitations. No single tool demonstrated consistently high accuracy for stroke and LVO detection. This is reflected in the variation of sensitivity and specificity observed across the tools. With the rapid global advancement of telemedicine, new diagnostic approaches will likely emerge for use in the prehospital setting. Future studies are needed to explore, validate, and potentially integrate these technologies to improve early identification of stroke and LVO.
The KoNES has established certification systems for endovascular neurosurgeons and stroke intervention-certified hospitals capable of performing acute management of stroke (https://www.konesonline.or.kr/certification/certiList.php; Supplementary Table 4). As of November 29, 2024, South Korea had 311 certified stroke interventionists and 106 certified institutions. In Japan, acute stroke treatment can be administered at centers certified by academic societies. Thus, when an emergency occurs, paramedics can use the EMS assessment tool; if the score meets a certain threshold, stroke or LVO is suspected. In such cases, the patient can be transported to a stroke intervention-certified hospital within 1 hour of travel time. As South Korea has many stroke centers, patients with suspected stroke can be transported to nearby facilities and receive appropriate treatment within an established infrastructure. If proper transport guidelines are in place and consistently followed, the quality of care for patients with stroke is likely to improve, and the redistribution of medical resources can be achieved more efficiently. In the future, KoNES plans to investigate and monitor emergency room visits of patients with stroke using these guidelines. These guidelines should be actively introduced to paramedics, emergency room medical staff, stroke specialists, and the public. Through nationwide education, this effort is expected to enable the faster identification of patients with stroke and contribute to the appropriate redistribution of medical resources.
Notes
Conflicts of interest
Won-Sang Cho has been editorial board of JKNS since September 2021 and Sukh Que Park has been editorial board of JKNS since May 2017. They were not involved in the review process of this original article. No potential conflict of interest relevant to this article was reported.
Informed consent
This type of study does not require informed consent.
Author contributions
Conceptualization : JSO; Data curation : JSO, DS; Formal analysis : JSO; Funding acquisition : JSO; Methodology : JSO; Project administration : JSO, JML, HSA, JJK, KMJ, GYY, JHK, DS, HJL, JJ, KCC, YSC, SJK, JP, WSC, SWO, HK, YWK, SKH, SQP, SCK; Visualization : JSO; Writing-original draft : JSO, DS; Writing-review & editing : JSO, YJ
Data sharing
This guideline is based on a meta-analysis of previously published studies. No new data were generated or analyzed in this study, and therefore, data sharing is not applicable.
Preprint
None
Acknowledgements
We would like to express our gratitude to Prof. Hyun-Jung Kim for contributing to the development of the literature search strategy and for providing insightful guidance.
This research was supported by the Korean Neuroendovascular Society (KoNES), Patient-Centered Clinical Research Coordinating Center (PACEN) (RS-2024-00399351), Bio & Medical Technology Development Program of the National Research Foundation funded by the Korean government (NRF-2023R1A2C100531), and Ministry of Health & Welfare of Republic of Korea (RS-2024-00439915) and Uijeongbu St. Mary’s Hospital of The Catholic University of Korea. The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Supplementary materials
The online-only data supplement is available with this article at https://doi.org/10.3340/jkns.2025.0106.
Progess of guideline
Key clinical question
Grade of level of evidence (LOE)
Grade of recommendation (GOR)
Institutions of KoNES certified stroke center
Certainly of recommendation of an emergency medical service tool for stroke diagnosis of emergent patients. aHeterogeneity >75. CI : confidence interval, CoE : certainty of evidence.
Certainly of recommendation of an emergency medical service tool for large vessel occlusion diagnosis of emergent patients. aHeterogeneity >75. CI : confidence interval, CoE : certainty of evidence.