An EMT’s primary job during a suspected stroke is not to treat the stroke itself but to recognize it fast, keep the patient stable, and get them to the right hospital as quickly as possible. The clot-busting drugs and surgical procedures that actually reverse stroke damage can only be given in a hospital, so nearly everything an EMT does in the field is aimed at shrinking the gap between symptom onset and definitive care. That makes the EMT’s role deceptively simple on paper and extraordinarily consequential in practice, because decisions made in the ambulance directly shape whether the patient arrives in time for treatments that are sharply time-sensitive.
How EMTs Recognize a Stroke in the Field
The backbone of prehospital stroke recognition is a structured screening scale. The most widely used is the Cincinnati Prehospital Stroke Scale, which checks for facial droop, arm drift, and slurred speech. When EMS crews document their use of it, their sensitivity for detecting stroke jumps dramatically. One study found that crews who used the Cincinnati scale correctly identified strokes with about 85% sensitivity, compared to roughly 31% when they did not use it. Documentation of the scale was independently associated with a twelve-fold increase in the odds of correctly recognizing a stroke.1PubMed. Clinical predictors of accurate prehospital stroke recognition The public-facing version of this approach is the FAST mnemonic (Face, Arms, Speech, Time), which compresses the same checks into a format easy enough for bystanders to remember.
Beyond simple yes-or-no stroke detection, a newer challenge is figuring out whether a stroke involves a large vessel occlusion, a blockage of one of the brain’s major arteries. These patients need endovascular thrombectomy, a catheter-based clot retrieval procedure available only at comprehensive stroke centers. Several scales exist to flag large vessel occlusions in the field. A systematic review and meta-analysis of 43 studies found that the Rapid Arterial Occlusion Evaluation score (at a cutoff of 5 or higher) had a pooled sensitivity of about 75% and specificity of about 76%, while the Field Assessment Stroke Triage for Emergency Destination scale performed similarly.2PubMed Central. Stroke Scales for Large Vessel Occlusion in the Prehospital Emergency Setting: A Systematic Review and Meta-Analysis A pooled analysis of over 2,300 patients confirmed that the RACE scale performed best overall in decision curve analysis, meaning it offered the greatest net benefit when weighing the cost of missing a large vessel occlusion against the cost of overtriaging patients who don’t have one.3PubMed Central. Prehospital Large-Vessel Occlusion Stroke Detection Scales: A Pooled Individual Patient Data Analysis of 2 Prospective Cohorts
No single scale is perfect for every clinical scenario. Among mild or minor strokes, the National Institutes of Health Stroke Scale captured the most cases (95% sensitivity) but had very low specificity, flagging many non-stroke patients, while the Los Angeles Prehospital Stroke Screen had the opposite problem, missing over half of mild strokes but rarely calling a false positive.4PubMed Central. Presenting symptoms and diagnostic accuracy of prehospital stroke scales for patients with suspected mild minor stroke EMTs work within whichever scale their local protocol mandates, but understanding that every tool has blind spots is part of why clinical judgment still matters alongside the checklist.
Where the Standard Screening Misses Strokes
The biggest known gap in FAST-based screening is posterior circulation stroke. Strokes in the back of the brain affect the brainstem and cerebellum, and they often present with dizziness, vision changes, and coordination problems rather than the classic facial droop or arm weakness that FAST is designed to catch. Over a third of patients with posterior circulation strokes are delayed or misdiagnosed because they lack those “typical” symptoms.5Journal of Paramedic Practice. Neurological assessment with FAST to better detect posterior circulation stroke A review of consecutive stroke records found that about 14% of stroke patients had no FAST symptoms at all when they first presented. Among those FAST-negative strokes, 42% had gait imbalance or leg weakness, 40% had visual symptoms, and 70% had at least one of the two.6PubMed. BE-FAST (Balance, Eyes, Face, Arm, Speech, Time): Reducing the Proportion of Strokes Missed Using the FAST Mnemonic
This is why an expanded mnemonic called BE-FAST has gained traction. It adds “Balance” and “Eyes” to the front of the original FAST checks. Some services also use supplemental tests like the finger-to-nose test or additional mnemonic frameworks (the “5 Ds” or “DANISH”) to pick up the vertigo, diplopia, and ataxia that characterize posterior strokes.7British Paramedic Journal. Not so FAST: pre-hospital posterior circulation stroke Adoption varies widely by region. Many EMS systems still use plain FAST, and the evidence base for how much these expanded tools improve real-world field outcomes is still developing. But the direction is clear: the original three-check approach leaves a meaningful number of strokes undetected, and EMTs trained to look for balance and vision problems catch more of them.
Sorting Real Strokes From Mimics
Not every patient who looks like they are having a stroke actually is. A single-center study that tracked nearly 950 patients transported by ambulance with suspected stroke found that about 43% turned out to be stroke mimics.8PubMed. Stroke Mimics Transported by Emergency Medical Services to a Comprehensive Stroke Center: The Magnitude of the Problem That is a strikingly high false-positive rate, and it reflects the genuine difficulty of telling strokes apart from other conditions in the field without imaging.
The most common neurological mimics were seizures, migraines, and peripheral nerve problems. The most common non-neurological mimics were cardiovascular conditions and psychiatric presentations.9PubMed Central. Stroke mimics: incidence, aetiology, clinical features and treatment One especially tricky mimic is severe low blood sugar, which can produce focal neurological deficits that look identical to a stroke, including one-sided neglect. Case reports document patients whose glucose was undetectable when EMS arrived, and whose stroke-like symptoms resolved after receiving glucose.10PubMed Central. Hypoglycemic hemineglect a stroke mimic This is why checking blood glucose is a standard part of every prehospital stroke assessment. It takes seconds, it costs nothing, and it can prevent a patient from being treated for a stroke they are not having.
EMTs are not expected to definitively distinguish a stroke from a mimic. Only brain imaging can do that. But knowing that mimics are common helps explain why protocols emphasize a systematic assessment rather than jumping straight to stroke-specific care the moment a patient has facial droop. Ruling out hypoglycemia, noting seizure activity, and documenting what you observe all contribute to better triage decisions.
What EMTs Actually Do Before the Hospital
Prehospital stroke care is mostly supportive. The treatments that reverse strokes, intravenous thrombolytics for ischemic strokes and endovascular thrombectomy for large vessel occlusions, require hospital-based imaging to confirm the diagnosis before they can be given. So the EMT’s medical interventions are focused on keeping the patient in the best possible shape to receive those treatments.
The basics start with the airway. Stroke patients, especially those with severe deficits, are at high risk of airway compromise. Roughly 38% of patients with acute hemiparetic stroke aspirate, and most of that aspiration is silent, meaning the patient inhales food or saliva into the lungs without coughing or showing obvious signs.11PubMed. Aspiration in patients with acute stroke EMTs position the patient to protect the airway, suction secretions, and intervene if breathing becomes inadequate. Keeping the patient nil by mouth is standard practice.
Supplemental oxygen is given if the patient’s oxygen saturation drops below 94%. The evidence for giving oxygen to stroke patients who are already breathing normally is weak. A randomized trial of routine low-flow oxygen in normoxic acute stroke patients showed no survival or disability benefit, and one subgroup of more severe strokes actually fared worse with supplemental oxygen.12PubMed Central. Acute Stroke: Current Evidence-based Recommendations for Prehospital Care – Section: Supplemental Oxygen The current standard is straightforward: correct hypoxia, but don’t blanket every stroke patient with oxygen they don’t need.
Starting an IV line is recommended for any patient with suspected stroke who may receive thrombolytics at the hospital, because those drugs carry a bleeding risk that could require rapid fluid or blood product infusion. But protocols are clear that transport should never be delayed to establish IV access, and multiple failed attempts should not be pursued in the field.13PubMed Central. Acute Stroke: Current Evidence-based Recommendations for Prehospital Care – Section: Fluid Assessment and Vascular Access
Blood Pressure in the Ambulance
One of the more complex questions in prehospital stroke care is what to do about high blood pressure. Many stroke patients arrive with markedly elevated readings. The instinct might be to bring it down, but the answer depends entirely on which kind of stroke the patient is having, and in the field, EMTs usually don’t know that yet.
A large randomized trial of intensive ambulance-delivered blood pressure reduction in hyperacute stroke found starkly different effects depending on stroke type. In patients with hemorrhagic stroke (a bleed in the brain), prehospital blood pressure lowering improved functional outcomes. But in patients with ischemic stroke (a clot blocking blood flow), the same intervention increased the odds of a poor outcome.14PubMed. Intensive Ambulance-Delivered Blood-Pressure Reduction in Hyperacute Stroke The problem is obvious: ischemic strokes are far more common than hemorrhagic ones, so aggressively lowering blood pressure in undiagnosed stroke patients will harm more patients than it helps. This is why most EMS protocols call for monitoring blood pressure and reporting it to the receiving hospital but not treating it aggressively in the ambulance unless readings are dangerously extreme. The threshold for prehospital intervention varies by local protocol and is generally set very high.
Documenting the “Last Known Normal” Time
Perhaps the single most important piece of information an EMT gathers is the “last known normal” time, the last moment when someone can confirm the patient was acting like their usual self. Thrombolytics for ischemic stroke have a time window, typically up to 4.5 hours from symptom onset, and whether a patient falls inside or outside that window depends on this timestamp. Getting it wrong in either direction has real consequences.
An analysis of stroke code patients found that among those who received clot-busting medication and whose preliminary last-known-normal time was earlier than the corrected time, roughly 29% would have had the treatment inappropriately withheld if the preliminary time had been used. In the other direction, among patients excluded from treatment for being outside the window whose preliminary time was later than the corrected time, about 70% would have been inappropriately treated outside the safe window.15PubMed Central. Accuracy of First Recorded “Last Known Normal” Times of Stroke Code Patients In other words, inaccurate timestamps lead to wrong decisions in both directions: treating people who should not be treated and withholding treatment from people who should get it.
EMTs are trained to pin this time down as precisely as possible, ideally by talking to a witness. “When did you last see them acting normally?” is the key question. “When did you notice something was wrong?” is a different and less useful answer, because the patient may have been symptomatic for a while before anyone noticed. If the patient woke up with symptoms, the last known normal time is when they were last seen awake and well, not when they went to bed. These distinctions sound pedantic, but they routinely determine whether someone receives a treatment that can mean the difference between recovery and permanent disability.
Getting to the Right Hospital, Not Just the Closest One
Not all hospitals can handle all strokes. Primary stroke centers can administer intravenous thrombolytics but typically cannot perform endovascular thrombectomy. Comprehensive stroke centers can do both. When an EMT’s field assessment suggests a large vessel occlusion, the protocol in many systems is to bypass the nearest primary stroke center and head directly to a comprehensive center, provided the additional transport time is reasonable.
One triage study used a severity score alongside the Cincinnati Prehospital Stroke Scale to sort patients: those scoring high enough on the severity screen were recommended for transport to the closest comprehensive stroke center if the added travel time was no more than 15 minutes.16JAMA Neurology. Prehospital Comprehensive Stroke Center vs Primary Stroke Center Triage in Patients With Suspected Large Vessel Occlusion Stroke Geographic modeling has estimated that bypassing the closest hospital for the closest comprehensive stroke center adds a median of about 8 minutes of transport time.17PubMed. Geographic Modeling to Quantify the Impact of Primary and Comprehensive Stroke Center Destination Policies That is a modest delay that, for the right patient, is more than offset by getting immediate access to thrombectomy rather than arriving at a hospital that will need to transfer them anyway.
The tricky part is the field-level decision. The screening tools for large vessel occlusion have moderate accuracy at best. Sending every suspected stroke patient to a comprehensive center would overwhelm those facilities, while sending everyone to the nearest hospital delays thrombectomy for the patients who need it most. This tension is one of the hardest operational problems in stroke systems of care, and every region balances it differently based on local geography, hospital capacity, and transport times.
Prenotification and Structured Handover
Calling ahead to the receiving hospital is one of the simplest interventions an EMT performs, and the evidence for its impact is surprisingly strong. A multicenter study found that when EMS prenotified the hospital, the median time from the patient’s arrival to their first CT scan dropped from 19 minutes to 13 minutes. The door-to-needle time for thrombolytic treatment also trended shorter, though the difference did not reach statistical significance in that study.18PubMed Central. Effect of prehospital notification on acute stroke care: a multicenter study Six minutes shaved off the time to imaging may not sound dramatic, but in stroke care, where brain tissue dies at a rate of roughly two million neurons per minute during a large ischemic event, those minutes are not trivial.
What the EMT communicates during that call also matters. A study testing a structured handover protocol between EMS and emergency department providers found that using an organized checklist significantly improved the reporting of key information, including the last known normal time, current anticoagulant or antiplatelet medications, neurological deficits, baseline mental status, and medical history.19PubMed. Optimizing handover in patients with stroke symptoms utilizing an organized protocol between emergency department providers and emergency medical personnel Without a structured format, critical details tend to get lost in the rush. Whether the patient is on blood thinners, for instance, directly affects whether thrombolytics can be given safely.
Scene Time and Protocol Strictness
How long an ambulance crew spends on scene before departing for the hospital is a controllable variable that protocols try to minimize. An evaluation of EMS stroke protocols found that protocols specifying an explicit time limit on scene reduced the 90th percentile scene time by about two minutes compared to protocols with no time instructions at all. Protocols that gave only general instructions to minimize scene time, without a specific number, did not produce a statistically meaningful reduction.20PubMed Central. An Evaluation of Emergency Medical Services Stroke Protocols and Scene Times The takeaway for system design is blunt: vague instructions to hurry don’t change behavior. Concrete time targets do. Most modern stroke protocols set an on-scene goal of roughly 15 minutes or less, though local variation exists.
Mobile Stroke Units and Telestroke
The most advanced evolution of prehospital stroke care is the mobile stroke unit, a specially equipped ambulance carrying a CT scanner, point-of-care lab equipment, and a physician or advanced-practice provider capable of interpreting imaging and initiating thrombolysis on the spot. Two large controlled clinical trials, B_PROUD and BEST-MSU, demonstrated in 2021 that treatment aboard mobile stroke units was safe and led to improved functional outcomes compared to conventional emergency care. The benefit appeared to be driven primarily by more frequent “golden hour” thrombolysis, meaning treatment delivered within 60 minutes of symptom onset.21PubMed. Mobile Stroke Units: Evidence, Gaps, and Next Steps
An earlier randomized trial of prehospital stroke treatment via mobile stroke unit found that the median time from alarm to treatment decision fell from 76 minutes with conventional care to 35 minutes with the mobile unit, a gain of about 41 minutes.22The Lancet Neurology. Prehospital stroke management in a specialized mobile stroke unit versus hospital-based treatment: a randomised controlled trial Reviews of the field consistently confirm that mobile stroke units reduce delays and enable better triage to the appropriate hospital level.23The Lancet Neurology. Streamlining of prehospital stroke management: the golden hour
Mobile stroke units remain expensive and are concentrated in major metropolitan areas. For regions where deploying one is not feasible, telestroke offers a middle ground. Pilot studies have tested real-time video links between ambulance crews and remote neurologists, allowing a stroke specialist to observe the patient and guide the EMT through a neurological exam during transport. One feasibility study found that 85% of teleconsultations were completed without major technical problems, and the neurological severity scores obtained remotely matched the scripted scores with high agreement.24PubMed Central. Pre-hospital Utility of Rapid Stroke evaluation Using In-ambulance Telemedicine (PURSUIT): A Pilot Feasibility Study Rural EMS systems have also demonstrated that the technology is workable even in areas with challenging connectivity, though clinical reliability in live-patient settings still needs more rigorous testing.25PubMed Central. Mobile Telestroke During Ambulance Transport Is Feasible in a Rural EMS Setting: The iTREAT Study
Neither mobile stroke units nor ambulance-based telestroke changes the fundamental EMT protocol. They layer specialist capability on top of it. The EMT still performs the initial screening, still documents the last known normal time, still manages the airway and monitors vitals, and still drives. What these technologies do is compress the gap between field recognition and hospital-level decision-making, sometimes eliminating it entirely. For the moment, they represent the leading edge of where prehospital stroke care is headed rather than the everyday reality for most ambulance crews.