First Aid Stroke Recognition (FA): Systematic Review

Conflict of Interest Declaration

The ILCOR Continuous Evidence Evaluation process is guided by a rigorous ILCOR Conflict of Interest policy. The following Task Force members and other authors were recused from the discussion as they declared a conflict of interest: none applicable

The following Task Force members and other authors declared an intellectual conflict of interest and this was acknowledged and managed by the Task Force Chairs and Conflict of Interest committees: none applicable

CoSTR Citation

Meyran D, Cassan P, Singletary E, Zideman D on behalf of the International Liaison Committee on Resuscitation (ILCOR) First Aid Task Forces.

Collaborators: Bendall J, Berry D, Borra V, Carlson J, Chang W-T, Charlton N, Djarv T, Douma M, Epstein J, Hood N, Markenson D, Orkin A, Sakamoto T, Woodin J

First Aid Stroke Recognition Consensus on Science with Treatment Recommendations [Internet] Brussels, Belgium: International Liaison Committee on Resuscitation (ILCOR) First Aid Task Force, 2020 January 2. Available from: http://ilcor.org

Methodological Preamble

The continuous evidence evaluation process for the production of this Consensus on Science with Treatment Recommendations (CoSTR) started with a systematic review of stroke recognition for first aid providers (PROSPERO submitted) conducted by Daniel Meyran and Pascal Cassan with involvement of the First Aid Task Force clinical content experts. Evidence from published literature was sought and considered by the First Aid Task Force. These data were taken into account when formulating the Treatment Recommendations.

PICOST

The PICOST (Population, Intervention, Comparator, Outcome, Study Designs and Timeframe)

Population: Adults with suspected acute stroke

Intervention: Use of a rapid stroke scoring system or scale (or test) (as FAST, LAPSS, CPSS, OPSS, KPSS, MASS or others)

Comparison: Basic first aid assessment without the use of a scale

Outcomes:

• Change time to treatment (e.g. symptom onset to hospital/emergency
  department arrival or hospital admission (9-Critical)
• Recognition of stroke: (5-Important)
  • high number considered beneficial for observational study
• high sensitivity and high specificity considered beneficial for diagnosis study
• Discharge with favorable neurologic status (increase considered beneficial) (5-Important)
• Survival with favorable neurologic outcome (increase considered beneficial) (5-Important)
• Increased public/layperson recognition of stroke signs (5-Important)

Study Designs: Randomized controlled trials (RCTs) and non-randomized studies (non-randomized controlled trials, interrupted time series, controlled before-and-after studies, cohort studies) were eligible for inclusion. Unpublished studies (for example, conference abstracts, trial protocols, posters) were excluded.

Timeframe and Languages: All years and all languages were included provided there was an English abstract

PROSPERO Registration Submitted Oct 26 2019

Consensus on Science

Time to treatment

For the critical outcome of time to treatment we identified four observational studies {Wojner-Alexandrov 2005 1512}{O’Brien 2012 241}{Iguchi 2011 51}{Chenkin 2009 153} evaluating four different stroke scales (KPSS, LAPSS, OPSS, FASTER).

For the Kurashiki Prehospital Stroke Scale (KPSS), we identified very low certainty evidence (downgraded for risk of bias and indirectness) from one retrospective observational study{Iguchi 2011 51} enrolling 430 participants with suspected acute stroke, showing benefit for the use of KPSS in regard to time to treatment, as measured by the number of patients whose time from symptom onset to hospital arrival was within 3 hours. Of patients who had the KPSS applied, 62.9% arrived within 3 hours compared with 52.3% who did not have the scale applied (RR 1.2 [95%CI 1.01 – 1.43]) (Evidence profile table 1). This same study showed a significantly shorter elapsed time from symptom onset to hospital admission with the use of KPSS in the prehospital setting (Mean time 2.1 hours [1.0 - 6.2]) in comparison with no prehospital KPSS use (Mean time 2.7 hours [1.2 - 9.7]) (p=0.024)).

For the Los Angeles Prehospital Stroke Scale (LAPSS), we identified very low certainty evidence (downgraded for indirectness) from one observational ‘before and after’ study{Wojner-Alexandrov 2005 1512}; 1518 participants with a suspected acute stroke showing no benefit for the use of LAPSS for the time, in minutes, from symptom onset to emergency department arrival. The mean time was 358 minutes for those who had a LAPSS screening tool applied (pre-intervention phase) compared with 226 minutes for those without use of a LAPSS screening tool (post-intervention phase) (MD 132.00 min [95%CI 14.68 – 249.32]) (Evidence profile table 2). This same study did not find a significant benefit from use of the LAPSS in a prehospital setting for the rate of patients admitted within 120 min (RR 1.15 [95%CI 0.68 – 1.94]) (Evidence profile table 2).

For the Ontario Prehospital Stroke Scale (OPSS), we identified very low certainty evidence (downgraded for risk of bias) from one observational study{Chenkin 2009 153} enrolling 861 participants suspected of acute stroke showing benefit for the use of OPSS in regard to the number of patients with time from symptom onset to hospital arrival within 3 hours. Of patients who had the OPSS applied, 32,1% arrived within 3 hours compared with 22.5% who did not have the scale applied (RR 1.42 [95%CI 1.12 – 1.82]) (Evidence profile table 3).

For the Face, Arm, Speech, Time, Emergency Response Protocol (FASTER) we identified very low certainty evidence (downgraded for risk of bias and imprecision) from one observational study {O’Brien 2012 241} enrolling 115 participants showing benefit the use of FASTER with regard to time from symptom onset to treatment with Tissue Plasminogen Activator (tPA). (MD -32 min P =0.005). This same study showed benefit for the use of FASTER with regard to door to computerized tomography (CT) time for patients receiving tPA (MD -46 min P =0.001). No significant differences were found between the two groups (with and without stroke screening tool applied) in regard to time from symptom onset to hospital arrival for patients receiving tPA. (MD 17 min P =0.180) (Evidence profile table 4).

We did not identify any comparative studies evaluating the other scales (ROSIER, MASS, CPSS, MedPACS and PreHAST) for the critical outcome of “time to treatment”.

Recognition of stroke: intervention studies

For the important outcome of recognition of stroke (interventional studies, outcome defined as definitive stroke diagnosis or administration of thrombolytic), we identified four observational studies{Harbison 2003 71}{Iguchi 2011 51}{O’Brien 2012 241}{You 2013 1699} evaluating four different stroke scales (FAST, KPSS, FASTER, OPSS, LAPSS).

For the Face, Arm, Speech, Time (FAST) scale, we identified low-certainty evidence (downgraded for serious risk of bias and imprecision) from one observational study{Harbison 2003 71} enrolling 356 participants with suspected of stroke, showing benefit for the use of FAST in regard to the number of patients with confirmed stroke or TIA who were admitted within 3 hours following symptom onset. Of those participants who had the scale applied 48.2% were diagnosed compared to 14.6% who did not have the scale applied, (RR 3.3 [95%CI 2.29 – 4.75]) (Evidence profile table 5).

For the Kurashiki Prehospital Stroke Scale (KPSS), we identified low-certainty evidence (downgraded for risk of bias and indirectness) from one observational study{Iguchi 2011 51} enrolling 430 participants suspected of stroke showing equal benefit for both use and non-use of KPSS for the number of patients who were diagnosed with stroke and received thrombolytic therapy. Of participants who had the KPSS scale applied, 13.7% were diagnosed and received tPA compared to 14.4% who did not have the scale applied (RR 0.95 [95%CI 0.59 – 1.53]) (Evidence profile table 1).

For the Los Angeles Prehospital Stroke Scale (LAPSS), we identified moderate-certainty evidence (downgraded for indirectness) from one observational ‘before and after’ study{Wojner-Alexandrov 2005 1512} enrolling 1518 participants showing benefit for the use of LAPSS by paramedics in regard to the number of correct initial diagnoses of stroke determined by neurologist. Of participants who had the LAPSS applied, 79.21% had a correct diagnosis by paramedics compared with 61.3% who did not have the scale applied (RR 2.41 [95%CI 1.83 – 3.17]) (Evidence profile table 2). The same study showed no benefit for the use of LAPSS in the rate of treatment with intravenous tPA of confirmed stroke cases between the two phases. 12.0% patients who had the LAPSS applied had treatment with intravenous tPA for a confirmed stroke compared to 16,60% who did not have the scale applied (RR 1.15 [95%CI 0.68 – 1.94]) (Evidence profile table 2).

For the Ontario Prehospital Stroke Scale (OPSS), we identified low-certainty evidence (downgraded for risk of bias) from one observational study{Chenkin 2009 153} enrolling 861 participants suspected of stroke showing no benefit with the use of OPSS for the identification of the rate of ischemic stroke. Of participants who had the OPSS applied 52.3% had confirmed ischemic stroke compared with 47.23% who did not have the scale applied (RR 1.23 [95%CI 0.93 - 1.62]) (Evidence profile table 3). This study showed benefit for the use of OPSS for the rate of thrombolytic therapy of all ischemic stroke cases. Of participants who had the OPSS scale applied, 10.10% received thrombolytic therapy compared with 5.86% who did not have the scale applied (RR 1.72 [95%CI 1.03 – 2.88]) (Evidence profile table 3). This same study showed benefit for the use of OPSS for rate of thrombolytic therapy for ischemic stroke patients arriving within 3 hours. 32.13% patients with ischemic stroke who had the scale applied who received thrombolytic therapy within 3 hours of arriving compared to 22.46% who have not the scale applied (RR 143 [95%CI 1.12 – 1.82]) (Evidence profile table 3).

For the Face, Arm, Speech, Time, Emergency Response protocol (FASTER), we identified very-low-certainty evidence (downgraded for serious risk of bias) from one observational study{O’Brien 2012 241} including 34 participants with suspected acute stroke showing benefit for the use of FASTER in the number of patients who received thrombolytic therapy. Of patients who had the scale applied, 19.1% received thrombolytic therapy compared with 7.5% who did not have the scale applied RR 2.56 [95%CI 1.02 - 6.45]) (Evidence profile table 4).

Recognition of stroke: diagnostic studies

For the important outcome of recognition of stroke (diagnostic studies, outcome defined as correct stroke diagnosis), we identified 18 observational studies{Berglund 2014 212}{Berg 2010 2}{Fothergill 2013 2007}{Pickam 2019 195}{Asimos 2014 509}{Bray 2005 28}{Chen 2013 e70742}{Kidwell 2000 71}{Chenkin 2009 153}{Bray 2010 1363}{English 2018 919}{Frendl 2009 754}{Greenberg 2017 1}{Kim 2017 867}{Ramanujan 2008 307}{Studnek 2013 348}{Vanni 2011 499}{Andsberg 2017 1} including 8153 participants, studying 9 different screening tools (FAST, LAPSS, OPSS, CPSS, ROSIER, MASS, BEFAST, Med-PACS, Pre-HAST) (table 5). All studies evaluated the accuracy of scales in a prehospital setting with the test being performed by paramedics or nurses (fig. 14). All studies used the same positivity threshold for each scale (’1 or greater’).

For the FAST scale, we identified very-low certainty evidence (downgraded for risk of bias, indirectness inconsistency) from 4 observational prospective studies{Berglund 2014 212}{Berg 2010 2}{Fothergill 2013 2007}{Pickam 2018 195} including 1585 participants suspected of acute stroke. The mean prevalence of stroke/TIA was (827/1585) 52.18% and ranged from 44.5% to 59.87%. The reported diagnostic sensitivities were 0.64{Berglund 2014 212}, 0.95{Berg 2010 2}, 0.76{Pickam 2019 195}, 0.97{Fothergill 2013 3007} and the report specificities were 0.75 {Berglund 2014 212}, 0.33{Berg 2010 2}, 0.46{Pickam 2019 195} and 0.13{Fothergill 2013 3007}. The summary estimate sensitivity was 0.86 (95%CI 0.69-0.94) and the summary estimate diagnostic specificity was 0.38 (95%CI 0.16-0.66) (fig. 1, 2 & Evidence profile table 6).

For the LAPSS, we identified low certainty-evidence (downgraded for very serious risk of bias) from 5 observational studies including 2692 participants suspected of stroke{Asimos 2014 509}{Bergs 2010 2}{Bray 2005 28}{Chen 2013 e70742}{Kidwell 2000 71}. Four studies were prospective{Bergs 2010 2}{Bray 2005 28}{Chen 2013 e70742}{Kidwell 2000 71} and one study was retrospective{Asimos 2014 509}. The mean prevalence of stroke/TIA was (1786/2692) 66.34% and ranged from 42.81% to 89.88%. The reported diagnostic sensitivities were 0.74{Asimos 2014 509}, 0.74{Bergs 2010 2}, 0.78{Bray 2005 28}, 0.78{Chen 2013 e70742}, 0.91{Kidwell 2000 71} and the reported diagnostic specificities were 0.48{Asimos 2014 509}, 0.83{Bergs 2010 2}, 0.85{Bray 2005 28}, 0.90{Chen 2013 e70742}, 0.97{Kidwell 2000 71}. The summary estimated diagnostic sensitivity was 0.78 (95%CI 0.75-0.81) and the summary estimated diagnostic specificity was 0.86 (95%CI 0.67-0.95) (fig. 3, 4 & Evidence profile table 7).

For the OPSS, we have identified low-certainty evidence (downgraded for risk of bias) from 1 observational study{Chenkin 2009, 153} including 554 participants with suspected of stroke. The prevalence of stroke/TIA was (214/554) 38.63%. The reported diagnostic sensitivity was 0.87 (95%CI 0.82-0.92) and the reported diagnostic specificity was 0.59 (95%CI 0.54-0.65) (fig. 5).

For the CPSS scale, we identified very low-certainty evidence (downgraded for risk of bias and inconsistency) from 12 observational studies{Asimos 2014 509}{Berg 2010 2}{Bray 2010 1363}{Bray 2005 28}{Frendl 2009 754}{Greenberg 2017 1}{Kidwell 2000 71}{Ramanujan 2008 307}{English 2018 919}{Kim 2017 867}{Vanni 2011 499}{Studnek 2013 348}. Six studies were prospective{Bergs 2010 2}{Bray 2010 1363}{Bray 2005 28}{Kidwell 2000 71}{Kim 2017 867}{Vanni 2011 499} and six studies were retrospective{Asimos 2014 509}{Frendl 2009 754}{Ramanujan 2008 307}{English 2018 919}{Greenberg 2017 1}{Studnek 2013 348} including 5202 participants suspected of stroke. The mean prevalence of stroke/TIA was (2154/5202) 41.41% and ranged from 31.91% to 50.90%. The reported sensitivities were 0.77{Asimos 2014 509}, 0.95{Bergs 2010 2}, 0.88{Bray 2010 1363}, 0.95{Bray 2005 28}, 0.70{Frendl 2009 754}, 0.46{Kidwell 2000 71}, 0.44{Ramanujan 2008 307}, 0.75{English 2018 919}, 0.93{Kim 2017 867}, 0.79{Studnek 2013 348}, and the report specificities were 0.48{Asimos 2014 509}, 0.33{Bergs 2010 2}, 0.79{Bray 2010 1363}, 0.56{Bray 2005 28}, 0.52{Frendl 2009 754}, 0.44{Kidwell 2000 71}, 0,.3{Ramanujan 2008 307}, 0.21{English 2018 919}, 0.73{Kim 2017 867}, 0.24{Studnek 2013 348}. The summary estimated diagnostic sensitivity was 0.80 (95%CI 0.67-0.88) and the summary estimated diagnostic specificity was 0.52 (95%CI 0.40-0.64) (fig. 6, 7 & Evidence profile table 8). Two additional studies were identified{Greenberg 2017 1}{Vanni 2011 499} but they provided incomplete data and the results were unable to be estimated.

For the ROSIER scale, we identified moderate-certainty evidence (downgraded for risk of bias) from 1 observational prospective study{Fothergill 2013 2007} including 295 participants suspected of acute stroke. The prevalence of stroke/TIA was (177/295) 60%. The reported diagnostic sensitivity was 0.97 (95%CI 0.93-0.99) and the reported diagnostic specificity was 0.18 (95%CI 0.11-0.26) (fig. 8).

For the MASS scale, we identified low-certainty evidence (downgraded for risk of bias) from 3 observational studies including 981 participants suspected of stroke{Berg 2010 2}{Bray 2005 28}{Bray 2010 1363}. Two studies were prospective{Berg 2010 2}{Bray 2005 28} and one study was retrospective{Bray 2010 1363}. The mean prevalence of stroke/TIA was (291/981) 29.66% and ranged from 23.41% to 61.29%. The reported sensitivities were 0.74{Berg 2010 2}, 0.90{Bray 2005 28}, 0.83{Bray 2010 1363} and the report specificities were 0.67{Berg 2010 2}, 0.74{Bray 2005 28}, 0.50{Bray 2010 1363} (fig. 22). The summary estimated diagnostic sensitivity was 0.85 (95%CI 0.79-0.90) and the summary estimated diagnostic specificity was 0.82 (95%CI 0.69-0.91) (fig. 8, 9 & Evidence profile table 9).

For the BEFAST scale, we identified low-certainty evidence (downgraded for risk of bias) from one observational prospective study including 359 patients with a suspected stroke{Pickham 2018 195}. The prevalence of stroke/TIA was (223/445) 45%. The reported sensitivity was 0.91 (95%CI 0.86-0.95) and the report specificity was 0.26 (95%CI 0.20-0.33) (fig. 11).

For the Med-PACS, we identified very-low-certainty evidence (downgraded for risk of bias) from one observational retrospective study{Studnek 2013 348} including 416 participants with a suspected acute stroke. The prevalence of stroke/TIA was (186/416) 44.7%. The reported diagnostic sensitivity was 0.74 (95%CI 0.67-0.80) and the reported diagnostic specificity was 0.33 (95%CI 0.27-0.39) (fig. 12).

For the Pre-HAST, we have identified low-certainty evidence from one observational prospective study{Andsberg 2017 1} including 69 participants suspected of stroke. The prevalence of stroke/TIA was (26/69) 37.7%. The reported sensitivity was 1.00 (95%CI 0.87-1.00) and the report specificity was 0.40 (95%CI 0.25-0.56) (fig. 13).

These studies can be divided into subgroups based on whether the stroke scales included blood glucose measurement or not. For the studies that used stroke scales including blood glucose measurement (LAPSS, OPSS, ROSIER, MASS, Med-PACS), the ranged estimate sensitivity was between 0.74 to 0.97 compared with 0.80 to 1.00 for the studies with stroke scales without the use of blood glucose measurement (FAST, CPSS, Pre-HAST, BEFAST) (fig. 15). The ranged estimate specificity of the studies that used stroke scales including blood glucose measurement was between 0.26 to 0.72 compared with 0.18 to 0.86 for those scales without blood glucose measurement (PreHAST, FAST, CPSS, BEFAST) (fig. 16).

Increased public/layperson recognition of signs of stroke

For the important outcome of increased public/layperson recognition of the signs of stroke, we identified very low certainty evidence (downgraded for risk of bias) from one human study{Wall 2008 A49} enrolling 72 participants (members of the public). This study showed benefit for the use of training in the recognition of stroke. 55/72 (76.4%) of participants were able to identify signs of stroke before training on a stroke screening assessment system compared with 68/72 (94.4%) immediately after training (RR, 1.24; 95%CI, 1.07 - 1.42), and 63/65 (96.9%) of participants were able to identify the signs of stroke 3 months after training (RR, 1.27; 95% CI, 1.11 - 1.45).

No comparison studies were identified for the important outcomes of discharge with favorable neurologic status and survival with favorable neurologic outcome.

Treatment Recommendations

We recommend that first aid providers use stroke assessment systems for individuals with suspected acute stroke (strong recommendation, low certainty evidence).

For first aid, we suggest the use of, FAST, MASS, CPSS or LAPSS for stroke assessment (weak recommendation, low certainty evidence).

For first aid, we suggest the use of stroke assessment systems that include blood glucose measurement, when available, such as MASS or LAPSS to increase specificity of stroke recognition (weak recommendation, low certainty evidence).

For first aid, we suggest the use of FAST or CPSS stroke assessment systems in the absence of the ability to undertake a blood glucose measurement (weak recommendation, low certainty evidence).

Justification and Evidence to Decision Framework Highlights

The original literature search for the 2015 ILCOR CoSTR was completed in January 2014 {Singletary 2015 S269}{Zideman 2015 e225}. This literature search was rerun in September 2019 to capture the new evidence between 2014 and 2019. Four additional studies were added {Berglund 2014 212}{Pickam 2019 195}{English 2018 919}{Greenberg 2017 1] and incorporated into the consensus on science and grade tables to update the 2015 consensus on science and treatment recommendations.

The Task Force considers that an ideal stroke assessment system for first aid use must have few steps, to be easily understood with a good retention, and to be sensitive and last must take a minimal time to be completed. These considerations were taken into account in the choice of tests that were studied.

Early treatment can minimize a potentially devastating neurologic injury. In making this recommendation, the Task Force emphasizes that in the initial management of the individual, the assessment using a stroke recognition score compared to no stroke score assessment assists in early stroke recognition, a reduced time from symptom onset to arrival at a hospital emergency department or hospital admission, and faster treatment of confirmed stroke patients.

The Task Force recognizes that in most studies, the stroke scale assessment was performed by paramedics or nurses and not by lay people or first aid providers. The benefit of training first aid providers in stroke assessment systems outweighs the risks, which is largely limited to false-positive identification by first aid providers. The Task Force considers that the lay public or first aid providers should use the stroke scale assessment protocol that provides the highest sensitivity and the lowest number of false negatives.

Four scales have been the subject of several studies and therefore a large number of participants have been tested (FAST, CPSS, LAPSS, MASS). Four scales (OPSS, ROSIER, BEFAST, MED-PACS) were each the subject of a single study with between 250 and 600 participants{Chenkin 2009 153}{Fotherhill 2013 2007}{Studneck 2013 348}{Pickham 2018 195}. The PreHAST scale provided a high level of sensitivity was only tested in a single study, with a low number of participants{Andsberg 2017 1}. The Task Force decided:

• to limit its conclusions concerning those scales with a larger number of patients
• to exclude scales with only a single study and low numbers of participants.

In this review of the literature, the stroke assessment systems include various components, such as looking for specific signs and measuring blood glucose levels. Our review found that stroke assessment systems that included blood glucose measurement had similar sensitivity but increased specificity to accurately identify stroke compared with the systems that did not include blood glucose measurement. We recognize that first aid providers may not have access to or the skill to use a properly calibrated glucose measurement device. Although use of blood glucose measurement is not routinely included in first aid training, glucose measurement devices are commonly available and used by the public.

The cost of the intervention is estimated as low. However, the Task Force took into account that assessment scales which included blood glucose measurement would require the acquisition of measurement devices that can be expensive for some countries. Furthermore, for some countries, the use of glucose measurement devices by first aid providers is not authorized by law.

Those developing local guidelines for first aid providers can use the results of this review to determine if the benefit of increased specificity with systems that include glucose measurement would be desirable in their settings, compared with using simpler stroke assessment systems that do not include glucose measurement, which have similar sensitivity but lower specificity.

Knowledge Gaps

• There were no studies identified that evaluated rapid stroke scoring systems or scales in the pediatric setting.
• Future studies should document if use of a rapid stroke scoring system or scale increases survival rate, CPC level, or survival to hospital discharge and after 30 days.
• There are no randomized controlled trials comparing the intervention with standard care in any patient population.

Attachments

Evidence-to-Decision Table: FA-801 Stroke Recognition

1. Andsberg G, Esbjörnsson M, Olofsson A, Lindgren A, Norrving B, Euler von M. PreHospital Ambulance Stroke Test - pilot study of a novel stroke test. Scand J Trauma Resusc Emerg Med 2017;25(1):1–6. DOI : 10.1186/s13049-017-0377-x
2. Asimos AW, Ward S, Brice JH, Rosamond WD, Goldstein LB, Studnek J. Out-of-hospital stroke screen accuracy in a state with an emergency medical services protocol for routing patients to acute stroke centers. Ann Emerg Med 2014;64(5):509–15. DOI : 10.1016/j.annemergmed.2014.03.024
3. Berglund A, Svensson L, Wahlgren N, Euler von M. Face Arm Speech Time Test Use in the Prehospital Setting, Better in the Ambulance than in the Emergency Medical Communication Center. Cerebrovasc Dis 2014;37(3):212–6. DOI : 10.1159/000358116
4. Bergs J, Sabbe M, Moons P. Prehospital stroke scales in a Belgian prehospital setting: a pilot study. Eur J Emerg Med 2010;17(1):2–6. DOI : 10.1097/MEJ.0b013e32831040ec
5. Bray JE, Coughlan K, Barger B, Stroke CB, 2010. Paramedic diagnosis of stroke: examining long-term use of the Melbourne Ambulance Stroke Screen (MASS) in the field. Stroke. 2010;41:1363-1366. DO : 10.1161/STROKEAHA.109.571836
6. Bray JE, Martin J, Cooper G, Barger B, Bernard S, Bladin C. Paramedic Identification of Stroke: Community Validation of the Melbourne Ambulance Stroke Screen. Cerebrovasc Dis 2005;20(1):28–33. DOI: 10.1159/000086201
7. Chen S, Sun H, Lei Y, et al. Validation of the Los Angeles Pre-Hospital Stroke Screen (LAPSS) in a Chinese Urban Emergency Medical Service Population. PLoS ONE 2013;8(8):e70742. DOI: 10.1371/journal.pone.0070742
8. Chenkin J, Gladstone DJ, Verbeek PR, et al. Predictive Value of the Ontario Prehospital Stroke Screening Tool for the Identification of Patients with Acute Stroke. Prehosp Emerg Care 2009;13(2):153–9. DOI : 10.1080/10903120802706146
9. English SW, Rabinstein AA, Mandrekar J, Klaas JP. Rethinking Prehospital Stroke Notification: Assessing Utility of Emergency Medical Services Impression and Cincinnati Prehospital Stroke Scale. J Stroke Cerebrovasc Dis 2018;27(4):919–25. DOI : 10.1016/j.jstrokecerebrovasdis.2017.10.036
10. Fothergill RT, Williams J, Edwards MJ, Russell IT, Gompertz P. Does use of the recognition of stroke in the emergency room stroke assessment tool enhance stroke recognition by ambulance clinicians? Stroke 2013;44(11):3007–12. DOI : 10.1161/STROKEAHA.13.000851
11. Frendl DM, Strauss DG, Underhill BK, Goldstein LB. Lack of Impact of Paramedic Training and Use of the Cincinnati Prehospital Stroke Scale on Stroke Patient Identification and On-Scene Time. Stroke 2009;40(3):754–6. DOI : 10.1161/STROKEAHA.108.531285
12. Greenberg K, Lesenskyj A, Eichorn D, Maxwell CR, et al. (2017). Change Takes Time: EMS as the Spark Plug for Faster Acute Ischemic Stroke Care. Mathews J Emergency Med. 2(2): 025.
13. Harbison J, Hossain O, Jenkinson D, Davis J, Louw SJ, Ford GA. Diagnostic Accuracy of Stroke Referrals From Primary Care, Emergency Room Physicians, and Ambulance Staff Using the Face Arm Speech Test. Stroke 2003;34(1):71–6. DOI : 10.1161/STROKEAHA.13.000851
14. Iguchi Y, Kimura K, Watanabe M, Shibazaki K, Aoki J. Utility of the Kurashiki Prehospital Stroke Scale for Hyperacute Stroke. Cerebrovasc Dis 2011;31(1):51–6. DOI : 10.1159/000320854
15. Kidwell CS, Starkman S, Eckstein M, Weems K, Saver JL. Identifying stroke in the field. Prospective validation of the Los Angeles prehospital stroke screen (LAPSS). Stroke 2000;31(1):71–6. DO I: 10.1161/01.STR.31.1.71
16. Kim DH, Kim SW, Jun SM, Kim JK. Accuracy of prehospital stroke recognition by paramedics and TPA therapy rate after transportation. Journal of the Neurological Sciences 2017;381:867. DOI : 10.1016/j.jns.2017.08.2444
17. O'Brien W, Crimmins D, Donaldson W, et al. FASTER (Face, Arm, Speech, Time, Emergency Response): experience of Central Coast Stroke Services implementation of a pre-hospital notification system for expedient management of acute stroke. J Clin Neurosci 2012;19(2):241–5. DOI : 10.1016/j.jocn.2011.06.009
18. Pickham D, Valdez A, Demeestere J, et al. Prognostic Value of BEFAST vs. FAST to Identify Stroke in a Prehospital Setting. Prehospital Emergency Care 2018;23(2):195–200. DOI : 10.1080/10903127.2018.1490837
19. Ramanujam P, Guluma KZ, Castillo EM, et al. Accuracy of Stroke Recognition by Emergency Medical Dispatchers andParamedics—San Diego Experience. Prehosp Emerg Care 2008;12(3):307–13. DOI : 10.1080/10903120802099526
20. Singletary EM, Zideman DA, De Buck ED, Chang WT, Jensen JL, Swain JM, Woodin JA, Blanchard IE, Herrington RA, Pellegrino JL, Hood NA, Lojero-Wheatley LF, Markenson DS, Yang HJ; Part 9: First Aid: 2015 International Consensus on First Aid Science with Treatment Recommendations. Circulation. Oct 20 2015;132 (16 Suppl 1):S269-311. DOI: 10.1161/CIR.0000000000000278
21. Studnek JR, Asimos A, Dodds J, Swanson D. Assessing the Validity of the Cincinnati Prehospital Stroke Scale and the Medic Prehospital Assessment for Code Stroke in an Urban Emergency Medical Services Agency. Prehosp Emerg Care 2013;17(3):348–53. DOI : 10.3109/10903127.2013.773113
22. Vanni S, Polidori G, Pepe G, et al. Use of biomarkers in triage of patients with suspected stroke. J Emerg Med 2011;40(5):499–505. DOI : 10.1016/j.jemermed.2008.09.028
23. Wall HK, Beagan BM, O'Neill J, Foell KM, Boddie-Willis CL. Addressing stroke signs and symptoms through public education: the Stroke Heroes Act FAST campaign. Prev Chronic Dis 2008;5(2):A49.
24. Wojner-Alexandrov AW, Alexandrov AV, Rodriguez D, Persse D, Grotta JC. Houston Paramedic and Emergency Stroke Treatment and Outcomes Study (HoPSTO). Stroke 2005;36(7):1512–8. DOI: 10.1161/01.STR.0000170700.45340.39
25. You JS, Chung SP, Chung HS, et al. Predictive value of the Cincinnati Prehospital Stroke Scale for identifying thrombolytic candidates in acute ischemic stroke. Am J Emerg Med 2013;31(12):1699–702. DOI : 10.1016/j.ajem.2013.08.029
26. Zideman DA, Singletary EM, , De Buck EDJ, Chang WT, Jensen JL, Swain JM, Woodin JA, Blanchard IE, Herrington RA, Pellegrino JL, Hood NA, Lojero-Wheatley LF, Markenson DS, Yang HJ; on behalf of the First Aid Chapter Collaborators. Part 9: First aid: 2015 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations. Resuscitation. 2015; 95: e225