Supplementary information for analyses

Angus IG Ramsay; Jean Ledger; Sonila M Tomini; Claire Hall; David Hargroves; Patrick Hunter; Simon Payne; Raj Mehta; Robert Simister; Fola Tayo; Naomi J Fulop

TABLE 11

Characteristics of studies included in the systematic review

First author and year	Location	Study type	Intervention	Population/review scope	Outcomes	Findings
Barrett et al., 2017⁵⁹	USA	Feasibility study¹ Simulation and ‘live’ phase	Audio-visual communication (iPad) EMS clinician supporting ambulance clinician assessment of patient	Stroke patients (live phase)	Completion of NIHSS assessment Time to conduct NIHSS assessment Staff satisfaction	NIHSS completed in 13/14 cases Video signal dropped periodically NIHSS assessment completed in average of 7.6 minutes Both neurologists and ambulance clinicians reported high satisfaction with prehospital video triage
Bergrath et al., 2011⁶¹	Germany	Feasibility study	Transmission of vital signs and 12-lead ECG data Audio-visual communication EMS physician in teleconsultation centre supporting assessment of patient in ambulance	157 emergency ambulance calls	Completion of assessments Clinicians assessed technical performance of system (quality of audio and visual signal, background noise, etc.) Degree of cooperation between ambulance and hospital EMC physicians	97% of video calls completed successfully Quality of video signal rated highly EMS clinician contributed to decisions on several occasions
Bergrath et al., 2013⁶⁰	Germany	Prospective, observational study	Transmission of blood pressure, 3-lead ECG, 12-lead ECG data EMS physician in teleconsultation centre supporting assessment of patient in ambulance (either by ambulance clinician or EMS physician)	35 emergency ambulance calls (i.e. many not stroke), used audio-visual consultation	Duration of consultation Patient conditions addressed Stage at which consultation commenced (e.g. without onsite EMS physician, while awaiting one, or once one arrived)	Mean duration of consultations 24.9 minutes 3/35 calls involved a diagnosis of neurological emergency 26/35 calls commenced without or while awaiting on-scene EMS clinician Conclusion that this approach is feasible for a range of emergency conditions
Chapman Smith et al., 2016⁶³	USA	Feasibility study Simulation	Audio-visual communication (using iPad) between trained NIHSS assessors and either prehospital clinicians or study investigators	Actors performing scripted stroke scenarios	Signal reliability Completion of NIHSS assessment Reliability of NIHSS scoring (ambulance vs. bedside)	All 27 calls and NIHSS assessments completed without significant interruption High correlation of ambulance and bedside NIHSS ratings (0.96)
Chapman Smith et al., 2019⁶²	USA	Feasibility study Simulation	Audio-visual communication between neurologists using tablets and ambulance fitted with dedicated audio-visual system	Actors performing scripted stroke scenarios	Interruptions of signal Clinician ratings of audio-visual quality and acceptability of service Obtaining clinical data and NIHSS score (compared with bedside/script-based NIHSS assessments)	91% of calls were completed Variable assessment of signal quality stability, and safety Moderate agreement between remote vs. script-based NIHSS scores
Eadie et al., 2015⁶⁴	Scotland	Feasibility study	Audio-visual communication to support stroke clinicians conduct remote assessments in different parts of the rural highlands, with a focus on mobile vs. static calls	Healthy volunteers performing scripted stroke scenarios	Reliability of signal Accuracy of decision about eligibility thrombolysis Time to decision about thrombolysis	Drops in signal in 47% of calls, but resumed in all cases, with delays of 2–3 minutes in assessment No difference in accuracy or time taken in completing remote assessments in mobile and static calls
Felzen et al., 2017⁶⁵	Germany	Feasibility study	Transmission of vital signs and 12-lead ECG data Consultation report printout in ambulance Audio-visual communication and still images to allow EMS physician in teleconsultation centre supporting assessment of patient in ambulance (either by ambulance clinician or EMS physician)	539 ambulance emergency calls (33 stroke cases)	Reliability of different transmissions (e.g. malfunctions, impact on quality of communication) Clinical value of communication sources, quality of audio-visual signals, background noise	Transmissions of all information types successful in > 90% of cases Clinicians rated images and videos as being of significant clinical value
Joseph et al., 2021⁶⁶	USA	Feasibility study Simulation	Audio-visual communication (via laptops at each end) to let a neurologist and nurse support ambulance clinician in simulated assessment	13 simulated stroke consultations	Communication (verbal/non-verbal) between ambulance clinician, neurologist, nurse and patient Types of interaction (requests for information/confirmation, giving instruction, repetition, camera movement, hand or facial gestures) Responses to disruptions in communication	82% communication verbal Neurologist, ambulance clinician and patient equally involved in discussion Disruption in 8% of all interactions, with 44% of these being communication-related
Lippman et al., 2016⁶⁷	USA	Feasibility study Simulation	Audio-visual communication (via iPad) with portable Wi-Fi. Permitting communication between hospital and ambulance to permit NIHSS assessment	30 simulation sessions	Sustained audio-visual signal Audio-visual signal quality	Signal was sustained and audio-visual signal was of sufficient quality to permit NIHSS assessment
Mort et al., 2016⁶⁸	Scotland	Feasibility study Simulation; preliminary findings	Ambulance technology to send live transcranial ultrasound images and audio-visual stream to stroke unit (to assess potential for prehospital thrombolysis)	23 simulation sessions	Feasibility of transmission of ultrasound and audio-visual signals	Variable data transfer rate, but suggesting transmission feasible
Rogers et al., 2021⁶⁹	USA	Feasibility study Simulation	Audio-visual communication to let neurologist, nurse and ambulance clinician conduct full neurological assessment in ambulance	13 simulation sessions	Observation by three human factors experts assessing usability and performance of system from human factors perspective, including error types	Issues with usability of interface (e.g. incorrect data entry, clicking the wrong buttons, inappropriate menu selections); errors most commonly made by nurses, then neurologists Recommendations for clearer system design, including clear labelling, highlighting errors, interface layout, system automation
Torres Zenteno et al., 2016⁷⁰	Spain	Feasibility study Simulation	System to let ambulance clinicians alert stroke team, including videoconferencing to let neurologist perform NIHSS assessment	3 simulation sessions	Capturing times taken for each stage of information-sharing	System seems feasible; technical issues include mobile network coverage
Valenzuela Espinoza et al., 2016⁷¹	Belgium	Feasibility study	Audio-visual communication letting stroke specialists provide 24/7 remote assessment of suspected stroke patients in the ambulance	187 emergency calls (16 with suspected stroke)	Completion of calls Consultation time Information transfer Diagnosis	94% of calls completed successfully Median consultation lasted 9 minutes (IQR 8–13 minutes) 12 patients identified as having stroke or TIA; 10 confirmed by in-hospital diagnosis Information was transferred safely and reliably from ambulance to hospital
Van Hooff et al., 2013⁷²	Belgium	Feasibility study Simulation	Audio-visual communication using laptops and cameras to let stroke clinician perform remote assessment of suspected stroke using UTSS	41 simulated calls from ambulance exhibiting different symptoms	Audio-visual and network signal reliability Examination times Consistency of scoring of stroke symptoms	Signal mostly reliable (5 video freezes, but nothing preventing the assessment) Mean examination using UTSS was 3.1 minutes High inter-rater reliability
Yperzeele et al., 2014³⁹	Belgium	Feasibility study	Audio-visual communication between ambulance and hospital using laptops, plus transmission of patient vital signs	41 emergency calls of which 5 were suspected stroke (3 confirmed stroke)	Data transfer of patient characteristics Patient diagnosis (prehospital and in-hospital)	Prehospital diagnosis reached in 37/41 cases High agreement between prehospital and hospital diagnosis No adverse incidents or safety issues reported
Wu et al., 2021⁷⁸	China	Observational study	Smartphone platform ‘Green’ to enable prehospital notification and communication between prehospital and ED clinicians	8457 acute ischaemic stroke patients who underwent thrombolysis	Time from arrival at hospital to administration of thrombolysis	Significant reductions in time to administration of thrombolysis observed in patients transferred using Green system vs. patients who arrived by themselves (i.e. without prehospital notification)
Al Kasab et al., 2021⁷³	USA	Pilot study	Audio-visual communication to let stroke clinicians assess suspected stroke patient remotely, including medical history, demographics, and NIHSS	67 stroke patients who received thrombolysis: 15 patients who underwent prehospital video triage compared with 52 patients who underwent standard transfer processes (which included telephone consultation)	Stroke onset to hospital Time from arrival at hospital to thrombolysis decision Time from arrival at hospital to thrombolysis administration	Patients who underwent prehospital video triage had no significant difference in time to hospital, but had significantly shorter time to thrombolysis decision and administration of thrombolysis
Belt et al., 2016⁷⁴	USA	Pilot study	Audio-visual communication, using high-definition camera, microphone and screen to let stroke clinicians work with ambulance clinicians to assess suspected stroke patients and coordinate care with acute setting	89 suspected stroke patients	Reliability of connection Length of remote consultation Time from when patient was last well to thrombolysis administration Time from arrival at hospital to thrombolysis administration	Connection adequate for all but two cases Remote consultations lasted around 7 minutes for thrombolysis cases and around 4 minutes for non-thrombolysis cases Patients who underwent prehospital video triage had shorter times to thrombolysis
Johansson et al., 2019⁷⁵	Sweden	Pilot study	Audio-visual communication to let neurologist and prehospital emergency nurses work together to support assessment of suspected stroke patients, e.g. using NIHSS	11 suspected stroke cases	Clinician perceptions of image and sound quality, and consistency of assessment Qualitative research with nurses	Clinicians felt images were good or very good and were confident in uniformity of assessments Nurses split on whether to develop service further owing to operational interference and unclear efficacy
Liman et al., 2012⁷⁶	Germany	Pilot study Simulation	Audio-visual communication system built into ambulance (including head and body camera) to permit remote assessment by stroke physician	Actors displaying different stroke symptoms	Completion of NIHSS Reliability of remote NIHSS assessment vs. stroke physician assessment of video recorded in-ambulance	NIHSS completed in 12 of 30 calls Moderate reliability of remote NIHSS assessment in remaining 12 calls
Wu et al., 2014⁷⁷	USA	Pilot study Simulation	Audio-visual communication to let vascular neurologists conduct assessment (e.g. using NIHSS) of suspected stroke patients while being conveyed by fire department	40 simulated scenarios performed by actors in ambulance	Reliability of technology (i.e. to permit completion of assessment) Reliability of assessment (i.e. agreement between remote assessments vs. independent neurologist assessment of scripted scenarios)	85% of assessments completed without major technological interruption 90% agreement between remote assessment and independent assessment of scripts
Bergrath et al., 2012⁷⁹	Germany	Prospective study	Audio-visual communication, plus transmission of still pictures and vital data, letting emergency physician support assessment of patient in ambulance	18 stroke patients undergoing prehospital video triage; 46 control stroke patients attended by a prehospital emergency car and ambulance	Timings of assessments Technical assessment of system Diagnosis	No significant difference between patients undergoing prehospital triage and control group in terms of time onsite, time to hospital and time to scan
Drenck et al., 2019⁸⁰	Denmark	Prospective study	Prehospital assessment of suspected stroke patients, including onsite ECG, consultation with neurologist in hospital	520 suspected stroke cases	Factors contributing to increased/decreased on-scene time	On-scene time lower when ECG conducted in hospital, other processes conducted in transit rather than onsite and when communication with hospital neurologists was rated as good
Eder et al., 2021⁸¹	Germany	Prospective study	Audio-visual communication (using hand held and desktop PCs) to support communication between stroke clinicians and ambulance clinicians in assessing suspected stroke patients	845 acute ischaemic stroke patients	Time from arrival at hospital to brain scan Time from arrival at hospital to thrombolysis administration	Arrival to brain scan was significantly lower in people who underwent prehospital video triage No significant difference in arrival to thrombolysis times
Mazya et al., 2020⁸⁴	Sweden	Prospective study	Telephone-only consultation between ambulance and stroke clinicians, to support diagnosis of potential for mechanical thrombectomy and decision-making on patient destination (i.e. to primary stroke centre or comprehensive stroke centre)	2905 suspected stroke patients	Accuracy of diagnosis of large-artery occlusion (i.e. suitable for mechanical thrombectomy) Delivery of mechanical thrombectomy (time from onset) Delivery of thrombolysis (time from onset)	87% accuracy of diagnosis of large-artery occlusion Prehospital telephone triage associated with significantly lower time from onset to thrombectomy, no significant change in onset to thrombolysis
Felzen et al., 2019⁸²	Germany	Retrospective study	24/7 service, including transmission of vital signs and 12-lead ECG data. Audio-visual communication and still images to allow EMS physician in teleconsultation centre supporting assessment of patient in ambulance (either by ambulance clinician or EMS physician)	6265 emergency calls (of which 1049 neurological conditions; stroke unspecified)	Number of teleconsultations Number of complications Number of transmission malfunctions	Use of prehospital triage increased over time Only 6 adverse events reported Transmission malfunctions were uncommon, with the highest occurring in audio communication (1.9%)
Quadflieg et al., 2020⁸⁵	Germany	Retrospective study	24/7 service, including transmission of vital signs and 12-lead ECG data. Audio-visual communication and still images to allow emergency physician in teleconsultation centre supporting assessment of patient in ambulance (either by ambulance clinician or EMS physician)	1218 emergency calls (including 584 patients who underwent prehospital video triage and 634 patients treated by onsite emergency physician)	Concordance of prehospital diagnosis with diagnosis at end of hospital stay	No significant difference between concordance of diagnoses provided by onsite emergency physician and remote emergency physician
Schröder et al., 2021⁸⁶	Germany	Retrospective study	24/7 service, including transmission of vital signs and 12-lead ECG data. Audio-visual communication and still images to allow emergency physician in teleconsultation centre to support assessment of patient in ambulance (either by ambulance clinician or EMS physician)	10,362 emergency calls using prehospital video triage Of these, 2007 life-threatening calls (of which 890 involved stroke)	Change in vital signs, pre and post prehospital video triage consultation	Significant improvements in vital signs overall 96% of 890 stroke cases could be managed by the remote emergency physician
Verma et al., 2010⁸⁷	Canada	Retrospective study	‘Patching’ to let stroke clinicians support ambulance clinicians in decision-making about ‘false positives’, including stroke patients who are not eligible for thrombolysis	2966 stroke patients conveyed by ambulance	Proportion of ‘false positive’ stroke patients, comparing patients where there was prehospital patch with stroke clinician and where ambulance clinicians applied assessment alone	Proportion of false positives was significantly higher when ambulance clinicians did not receive prehospital input from stroke clinicians
Valenzuela Espinoza et al., 2017⁴¹	Belgium	Cost-effectiveness model	Prehospital video triage allowing prenotification of stroke clinicians and a range of patient data from the ambulance	2282 stroke patients from Brussels stroke registry	Cost-utility model measuring costs and quality-adjusted life-years, driven by assumptions about reduced time to key interventions and resultant improvements in patient outcomes	If time gains are greater than 6 minutes, the model estimates that prehospital video triage is cost-effective. If time gains exceed 12 minutes, it becomes dominant (i.e. saving cost and improving outcomes)
Hölscher et al., 2013⁸³	–	Concept paper	Discussion of potential of prehospital transcranial ultrasound scans Brief discussion of evidence related to prehospital video triage	–	–	Proposal that prehospital ultrasound may be less costly than mobile stroke units and less technologically challenging than prehospital video triage
Seah et al., 2019⁸⁸	–	Concept paper	Description of development of online platform to allow all clinicians associated with stroke pathway to communicate	–	–	Outline of communication, principally involving instant text-messaging, but with function of sharing images or videos Summary of key roles across stroke pathway
Amadi-Obi et al., 2014⁸⁹	–	Review	Telemedicine (not just prehospital video triage) related to trauma, myocardial infarction, and stroke	Studies from 1970–2014	Outcomes of interest included cost-effectiveness, feasibility and clinical outcome	5 studies covered communication between ambulance and hospital Review found limited conclusive evidence for effectiveness of telemedicine for emergency generally
Aude Bert et al., 2013⁴⁷	–	Review	Prehospital management of stroke patients (not just video triage)	Review process unclear	Current approaches to prehospital stroke management, noting risks and potential benefits	Alongside other interventions, prehospital video triage discussed as feasible, but noting the limitations of 3G networks and the potential of 4G to carry a stable signal between ambulance and stroke clinicians
Chowdhury et al., 2021⁹⁰	–	Review	Improved triage for thrombolysis Identification of patients with large arterial occlusion (i.e. eligible for diversion to centre that provides thrombectomy) Mobile stroke units	Systematic review and meta-analysis of papers from 2005	Analysed impact on time to key interventions, including thrombolysis (time from onset; time from arrival at hospital) and patient outcomes (function; mortality)	27 articles included in analysis Improved triage for thrombolysis associated with increased thrombolysis rates, reduced time to thrombolysis No impact of interventions on functional outcome or mortality
Fassbender et al., 2013⁹¹	–	Review	Approaches to improve timely access to acute stroke care, principally thrombolysis	Review process unclear	Role of patients/public, ambulance services (e.g. ambulance clinician triage, prenotification), future directions (including prehospital video triage)	Prehospital video triage is seen as of potential value, but mobile transmission from ambulance is identified as a significant obstacle
Fassbender et al., 2020⁹²	–	Review	Approaches to improve timely delivery of mechanical thrombectomy	Review process unclear	Role of different stakeholders in timely pathway to thrombectomy, including patients/public, ambulance services and acute hospitals	Prehospital video triage discussed in relation to prenotification, noting potential for two-way communication between ambulance and stroke clinicians. The technology is described as reliable, although impact on outcomes yet to be established
Guzik et al., 2021⁹³	–	Review	Use of telemedicine in different parts of the stroke pathway during the COVID-19 pandemic	Review process unclear	Use of telemedicine from prehospital through to long-term rehabilitation	Prehospital triage by remote stroke clinician has potential to support screening process, permit ambulance clinicians to focus more on COVID-19 control (e.g. PPE), and help ensure appropriate patient destination. However, evidence is required on how best to deliver these services
Hubert et al., 2014⁵³	–	Review	Prehospital management of stroke in developing countries. Interventions include mobile stroke units and remote prehospital triage	Systematic review of papers from 2014 onward	In the context of prehospital video triage, main outcome was reliability of signal	25 articles included in analysis Studies reported variable stability of signal, but noted potential of 4G networks
Lumley et al., 2020¹⁶	–	Review	Technology to support ambulance clinicians in assessing suspected stroke Interventions included blood biomarkers, prehospital imaging and ‘mobile telestroke’ (including prehospital video triage)	Scoping review of papers between 2000 and 2019	In terms of prehospital video triage, accuracy of diagnosis, acceptability of technology, impact on delivery of interventions and patient outcomes	19 articles included in analysis Most studied found good agreement between prehospital and final assessments Most studies reported high satisfaction with technology and a view that it was reliable No robust evidence on patient acceptance reported Some evidence suggested that time to interventions either improved through, or was not affected by, prehospital triage Little evidence reported on outcomes
Martinez-Gutierrez et al., 2019⁹⁴	–	Review	Approaches to improve time to thrombectomy, focusing on detection, evaluation, triage and transport of patients	Review process unclear	In terms of prehospital video triage, reliability of signal and quality of communication between ambulance and stroke services	Signal reliability reported as poor in earlier years, but indications of improvement with advances in technology Describes importance of appropriate transfer to thrombectomy-capable centres, and the potential of prehospital triage to support this
Mazighi et al., 2010⁹⁵	–	Review	Approaches to improve timely access to acute stroke care, principally thrombolysis Main focus on inter-hospital communication and prenotification of stroke services from ambulance Limited focus on prehospital triage	Review process unclear	In terms of prehospital video triage, there is a brief focus on feasibility	Purely identified as a future development, dependent on improvements in image quality and signal reliability
Rogers et al., 2017⁴⁸	–	Review	Telemedicine to support prehospital emergency care, including stroke	Systematic review of articles between 2000 and 2016	Feasibility, diagnostic accuracy, and impact on treatment timings (time to thrombolysis)	23 articles were selected, of which 6 focused on stroke Generally positive views of usability and quality of images High accuracy of remotely conducted NIHSS ratings Prehospital assessments associated with shorter time to thrombolysis
van Gaal et al., 2018⁹⁶	–	Review	Discusses approaches to improve access to thrombectomy, importantly noting that diagnosis is ‘primarily a means to support transport decision-making, not an end’ Range of interventions discussed, including real-time involvement of stroke clinicians	Review process unclear	Potential impact on appropriate diagnosis (via NIHSS), wider eligibility for thrombectomy and destination decision-making	Describes no evidence conducted on thrombectomy settings Draws on previous research on thrombolysis to suggest prehospital triage services are likely to be reliable and support accurate decision-making Makes a general point for implementation, relating to local requirements for sensitivity and specificity of processes, and the need to consider under- and over-triage Notes that prehospital triage is likely to disrupt ambulance and stroke clinicians’ other duties
Weber et al., 2015⁹⁷	–	Review	Prehospital stroke care, focusing on telemedicine, thrombolysis and outcomes	Systematic review covering 2004–2015	In terms of prehospital video triage, focus on reliability of signal to permit assessment (e.g. via NIHSS)	Reported many issues with 3G-based systems, but greater reliability of assessment via 4G systems
Winburn et al., 2018⁴⁰	–	Review	Prehospital telemedicine, focusing on different types of care. Interventions include audio communication, transmission of data (images, patient vital data), videoconferencing, and remote monitoring	Systematic review covering 2000–2017	Covers the trends, scope and type of telemedicine used in prehospital emergency medicine. Outcomes were not discussed	68 papers included, of which 19 related to stroke and 12 related to prehospital video triage (i.e. videoconferencing more common in stroke than other conditions)
Yperzeele et al., 2014³⁹	–	Review	Approaches to prehospital care to improve access to stroke care. Interventions include education of different stakeholders and prehospital triage	Review process unclear	In terms of prehospital video triage, accuracy of decision-making, reliability of connection, journey times and time to clinical intervention	Reports limited evidence on prehospital video triage, with most studies simulations with healthy volunteers and relatively few ‘real-world’ studies Video consultations more accurate than telephone alone Improvements in technology suggest it is likely that prehospital triage may be delivered reliably Notes clear potential of prehospital video triage, but that data security, privacy and regulatory issues remain to be addressed

: 24/7, 24 hours per day, 7 days per week; 3G, third generation broadband cellular network technology; UTSS, Unassisted TeleStroke Scale.

: Note

: Feasibility = technical studies and pilots (e.g. simulations, usability evaluations and technical performance assessments). Some publications are from the same trial or research study, for example there are several publications originating from the Prehospital Stroke Study at the Universitair Ziekenhuis Brussel (PreSSUB) in Belgium and the implementation of EMS telemedicine project in Aachen, Germany.

TABLE 12

Overview of survey responses

Item/option	NC London, n (%)	East Kent, n (%)	Total, n (%)
Q2: How long have you been an ambulance clinician for?
Less than 1 year	8 (5.0)	3 (4.1)	11 (4.7)
1–2 years	33 (20.8)	12 (16.2)	45 (19.3)
3–4 years	44 (27.7)	15 (20.3)	59 (25.3)
5 years or more	74 (46.5)	44 (59.5)	118 (50.6)
Q4: How often have you used prehospital video triage?
1–2 times	11 (6.9)	5 (6.8)	16 (6.9)
3–9 times	106 (66.7)	32 (43.2)	138 (59.2)
10 times or more	41 (25.8)	36 (48.6)	77 (33.0)
Don’t know	1 (0.6)	1 (1.4)	2 (0.9)
Q5: Were you satisfied with ease of use?
Strongly agree	104 (65.4)	20 (27)	124 (53.2)
Agree	46 (28.9)	38 (51.4)	84 (36.1)
Neither agree or disagree	4 (2.5)	4 (5.4)	8 (3.4)
Disagree	1 (0.6)	6 (8.1)	7 (3.0)
Strongly disagree	4 (2.5)	6 (8.1)	10 (4.3)
Q6: Were sound and vision of sufficient quality?
Strongly agree	54 (34.0)	19 (25.7)	73 (31.3)
Agree	79 (49.7)	27 (36.5)	106 (45.5)
Neither agree or disagree	14 (8.8)	12 (16.2)	26 (11.2)
Disagree	10 (6.3)	9 (12.2)	19 (8.2)
Strongly disagree	2 (1.3)	4 (5.4)	6 (2.6)
Don’t know	0 (0.0)	3 (4.1)	3 (1.3)
Q7: Which problems have you experienced?
No issues	60 (37.7)	14 (18.9)	74 (31.8)
Poor video	26 (16.4)	13 (17.6)	39 (16.7)
Poor sound	57 (35.8)	26 (35.1)	83 (35.6)
Poor sound and video	14 (8.8)	13 (17.6)	27 (11.6)
Wi-Fi issues	23 (14.5)	13 (17.6)	36 (15.5)
Other	15 (9.4)	8 (10.8)	23 (9.9)
No answer	18 (11.3)	49 (66.2)	67 (28.8)
Q8: Has it influenced conveyance decisions?
Yes, frequently	50 (31.4)	17 (23.0)	67 (28.8)
Yes, infrequently	93 (58.5)	41 (55.4)	134 (57.5)
No, never	15 (9.4)	12 (16.2)	27 (11.6)
Not sure	1 (0.6)	4 (5.4)	5 (2.1)
Q9: How many attempts needed to contact stroke team?
1– 2	156 (98.1)	57 (77.0)	213 (91.4)
3– 4	1 (0.6)	13 (17.6)	14 (6.0)
5 or more	0 (0.0)	3 (4.1)	3 (1.3)
Don’t know	2 (1.3)	1 (1.4)	3 (1.3)
Q10: What if there is no immediate response?
Keep trying until call picked up	11 (6.9)	14 (18.9)	25 (10.7)
Try up to 2 times, then revert to standard patient conveyance	87 (54.7)	39 (52.7)	126 (54.1)
Other	39 (24.5)	18 (24.3)	57 (24.5)
Don’t know	22 (13.8)	3 (4.1)	25 (10.7)
Q11: Any safety concerns about prehospital video triage?
Yes	13 (8.2)	22 (29.7)	35 (15.0)
No	145 (91.2)	46 (62.2)	191 (82.0)
Don’t know	1 (0.6)	4 (5.4)	5 (2.1)
Prefer not to say	0 (0.0)	2 (2.7)	2 (0.9)
Q12: Is it an improvement on what went before?
Strongly agree	99 (62.3)	24 (32.4)	123 (52.8)
Agree	51 (32.1)	26 (35.1)	77 (33.0)
Neither agree or disagree	4 (2.5)	10 (13.5)	14 (6.0)
Disagree	1 (0.6)	9 (12.2)	10 (4.3)
Strongly disagree	2 (1.3)	5 (6.8)	7 (3.0)
Don’t know	2 (1.3)	0 (0.0)	2 (0.9)
Q13: Would you like prehospital triage to continue?
Yes	153 (96.2)	52 (70.3)	205 (88.0)
I don’t mind	3 (1.9)	7 (9.5)	10 (4.3)
Not unless aspects are improved	2 (1.3)	10 (13.5)	12 (5.2)
No, definitely not	1 (0.6)	4 (5.4)	5 (2.1)
Don’t know	0 (0.0)	1 (1.4)	1 (0.4)
Q14: Should it be considered for other conditions?
Yes	127 (79.9)	29 (39.2)	156 (67.0)
No	10 (6.3)	23 (31.1)	33 (14.2)
Not sure	22 (13.8)	22 (29.7)	44 (18.9)
Q15: Did you receive sufficient training?
Yes	145 (91.2)	31 (41.9)	176 (75.5)
No	12 (7.5)	35 (47.3)	47 (20.2)
Don’t know	2 (1.3)	8 (10.8)	10 (4.3)
Total (all items were completed by all respondents)	159 (100.0)	74 (100.0)	233 (100.0)

: Note

: Items 1 (confirming consent) and 3 (confirming host organisation) excluded from table.

: Participants could provide multiple responses to item 7, meaning total percentages may exceed 100% for this item.

Postcode	NC London, n			East Kent, n
Postcode	30–45 minutes	45–60 minutes	> 60 minutes	Postcode	30–45 minutes	45–60 minutes
E17	2			CT10	10	1
E4	1	2	2	CT11	3
E5	1			CT12	1
EN1	8	3	3	CT14	3
EN2	8	2	4	CT16	1
EN3	16	3	2	CT21	4
EN4	7	2	3	CT5	1
EN5	9	1		CT7	1
EN9	1			CT9	6
HA8	1	2	1	TN23	4
N10	1			TN24	2
N11		1		TN25	1
N12	2		1	TN26	3
N13	1	1	1	TN27	3
N14	3	2		TN28	3	2
N15	3	1		TN29	7	1
N16	2			TN30	1	3
N17	4	4		TN31		1
N18	1	2		CT10	10	1
N19	3			CT11	3
N2		2		CT12	1
N20	4	1		CT14	3
N21	2	1		CT16	1
N22	3	2		CT21	4
N3		1		CT5	1
N5		1		CT7	1
N6	2	1		CT9	6
N8	4			TN23	4
N9	4	5	1	TN24	2
NW1	1			TN25	1
NW11	1			TN26	3
NW2	2			TN27	3
NW3	4	1		TN28	3	2
NW4		2		TN29	7	1
NW6	2			TN30	1	3
NW7	1	1		TN31		1
W10	1
W8	1
W9	2	1

Characteristic	Region
	RoE			NC London			East Kent			DiD
	Before^a	After	Difference	Before^a	After	Difference	Before^a	After	Difference	NC London minus RoE (SE)	East Kent minus RoE (SE)
Number of patients admitted to HASU, n	44,122.00	43,336.00	–786	717.00	667.00	–50	451.00	319.00	–132	69.50 (793.16)	–13.00 (793.48)
Number of patients admitted to non-HASU, n	415.00	870.00	455	62.50	79.00	17	1.50	172.00	171	–438.50* (7.91)	–284.50* (–69.33)

Clinical interventions	DiD, % (SE)
	Adjusted outcomes for age		Adjusted outcomes for sex (female)		Adjusted outcomes for NIHSS		Adjusted outcomes for type of stroke (infarction)		Adjusted outcomes for number of comorbidities (2 or more)		Adjusted outcomes for arriving by ambulance
	NC London minus RoE	East Kent minus RoE	NC London minus RoE	East Kent minus RoE	NC London minus RoE	East Kent minus RoE	NC London minus RoE	East Kent minus RoE	NC London minus RoE	East Kent minus RoE	NC London minus RoE	East Kent minus RoE
Admitted to stroke unit within 4 hours	42.74 (12.59)	40.29 (5.30)	51.348 (6.38)	42.611 (5.55)	43.10 (10.96)	40.95* (5.27)	47.91 (5.41)	24.68 (12.63)	51.42 (7.30)	46.47 (8.26)	45.68 (13.37)	33.24 (3.58)
Brain scan within 1 hour	1.38 (6.79)	24.03* (0.74)	7.453 (0.67)	20.88* (0.86)	2.7 (5.91)	23.15* (0.74)	5.44 (0.85)	43.48** (0.28)	7.61* (0.16)	15.10 (2.06)	4.83 (4.47)	27.87 (9.40)
Thrombolysis
Given thrombolysis (all stroke types)	–5.43 (3.19)	3.16* (0.12)	–2.018 (0.23)	4.374* (0.16)	–4.25 (2.77)	3.50* (0.12)	–2.96* (0.09)	–4.28 (0.54)	–1.96 (0.47)	6.61 (0.56)	–3.33 (2.50)	1.83 (3.73)
Eligible patients for thrombolysis	–5.25 (2.13)	5.28 (0.49)	–2.69 (0.62)	7.13 (0.57)	–4.15 (1.85)	5.80 (0.49)	–3.30 (0.49)	–6.10** (0.01)	–2.66 (0.78)	10.50 (1.29)	–3.62 (1.99)	2.99 (5.45)
Eligible patients given thrombolysis	0.35 (0.67)	–7.48 (0.74)	0.45 (0.85)	–6.65 (0.74)	0.65 (0.58)	–7.25 (0.74)	0.53 (0.77)	–12.32 (2.59)	0.47 (0.88)	–5.06 (0.78)	0.69 (0.85)	–7.66 (3.12)
Thrombolysis (1 hour)	–11.32 (3.86)	22.17* (0.99)	0.84 (1.23)	20.66* (0.97)	–1.80 (3.36)	21.75* (0.98)	–0.27 (0.99)	31.09 (3.75)	0.89 (1.52)	17.79* (0.91)	–0.85 (3.68)	22.81 (5.43)
Assessed by a specialist
Assessed by stroke physician (24 hours)	6.16* (1.10)	32.16** (0.12)	8.17* (0.17)	32.99** (0.15)	7.4* (0.96)	32.40** (0.12)	7.84** (0.12)	27.10** (0.26)	8.18* (0.25)	34.51** (0.44)	7.70 (0.93)	31.23 (2.51)
Assessed by a stroke nurse within 24 hours	2.22* (16.01)	16.01* (0.37)	3.91 (0.42)	16.88* (0.36)	3.7** (0.37)	16.25* (0.37)	3.82 (0.39)	10.77 (1.63)	3.93 (0.40)	18.53** (0.24)	3.85* (0.09)	15.45 (3.01)
Swallow screen within 4 hours	19.06 (5.30)	31.27 (2.59)	23.69 (3.09)	32.24 (2.70)	20.30* (4.62)	31.55 (2.57)	22.28 (2.65)	24.67 (6.29)	23.71 (3.47)	33.81 (3.98)	21.29 (5.84)	27.99* (1.22)

Appendix 1Supplementary information for analyses

TABLE 11

TABLE 12

TABLE 13

TABLE 14

TABLE 15

Publication Details

Copyright

Publisher

NLM Citation

TABLE 11

TABLE 12

TABLE 13

TABLE 14

TABLE 15