The appropriate timeframe for starting endovascular therapy in an eligible stroke patient runs from the first minutes after symptom onset to as late as 24 hours in carefully selected cases. The practical rule splits into three windows: within 6 hours for most anterior-circulation large-vessel occlusions, 6 to 24 hours only when advanced imaging shows salvageable brain tissue, and as early as possible in every window because outcomes worsen with each passing hour. The patient must have a confirmed large-vessel occlusion, a measurable deficit, and no hemorrhage on imaging before any clock matters at all.
How the 2015 Landmark Trials Set the First Six-Hour Standard
Endovascular stroke care became standard practice after 2015, when five randomized trials published together in the New England Journal of Medicine showed that mechanical thrombectomy plus standard medical care beat intravenous alteplase alone in anterior-circulation large-vessel occlusion. The trials β MR CLEAN, ESCAPE, EXTEND-IA, SWIFT PRIME, and REVASCAT β enrolled patients mostly within 6 hours of symptom onset and consistently demonstrated better 90-day functional outcomes with the procedure.
ESCAPE stood out for its imaging-based selection: patients needed a small core infarct and evidence of good collateral circulation on CT angiography. That trial was halted early for efficacy, a signal that the treatment effect was large enough that continuing randomization would have been unethical. The number needed to treat for a good functional outcome in some of these trials ran as low as a handful of patients, which is rare in acute stroke care.
Those results turned a 6-hour door-to-groin timeline into a quality benchmark. Hospitals built stroke systems around rapid transfer, prehospital notification, and direct-to-angio protocols. The 6-hour window is still the foundation of every guideline that followed.
One detail the trial summaries often flatten: the trials did not test a fixed deadline so much as a population. Patients enrolled at 5 hours 45 minutes were included; those at 6 hours 30 minutes generally were not. The 6-hour line is a regulatory and design boundary, not a biological cliff.
Why the Window Moved to 24 Hours for Select Patients
Two trials published in 2018 in the New England Journal of Medicine β DAWN and DEFUSE 3 β reopened the question. DAWN enrolled patients 6 to 24 hours after last known well who had a clinical-imaging mismatch: a substantial neurological deficit but a small infarct core on diffusion-weighted MRI or CT perfusion. DEFUSE 3 used a perfusion-based mismatch profile between 6 and 16 hours.
Both trials showed benefit, and DAWN's treatment effect was striking in the 12-to-24-hour group. The results shifted guidelines from the American Heart Association and American Stroke Association, which now recommend thrombectomy up to 24 hours in selected patients who meet advanced imaging criteria. The 24-hour figure is not a green light for every late presenter.
Selection is the whole point. DAWN required a mismatch between the severity of the clinical deficit, measured by the NIH Stroke Scale, and the volume of already-dead tissue. A patient with a large core and a mild deficit does not qualify, no matter how early they arrive. The imaging does the gatekeeping.
In practice, this means a community hospital without CT perfusion or MRI mismatch protocols cannot simply extend its window by decree. It needs the imaging capability, the interpretation, and often a transfer arrangement to a comprehensive stroke center. The 24-hour window is a systems question as much as a clock question.
The Hourly Cost of Waiting, Quantified
Speed matters within every window, and the magnitude is measurable. A frequently cited analysis in the journal Stroke estimated that every 15-minute delay to reperfusion costs roughly one month of healthy life, and that each minute of delay translates to approximately 1.9 days of healthy life lost. Those figures came from pooled trial data and are used in quality-improvement literature, not as a bedside promise.
The arithmetic is worth sitting with. A 30-minute transfer delay is not a scheduling inconvenience; by that estimate it costs about two months of disability-free life. A one-hour delay costs roughly four months. That is the scale that justifies helicopter transfers and prehospital triage algorithms.
Set that against the alternative. Intravenous alteplase, when eligible, has a treatment window of up to 4.5 hours from onset per the FDA label and AHA/ASA guidance, and it is given before thrombectomy in most eligible patients rather than instead of it. The two treatments are sequenced, not competing.
The practical translation for a stroke team: if a patient is eligible, the correct initiation time is now. The 6-hour and 24-hour figures are inclusion criteria for trials and guidelines, not targets to plan around. No protocol says "wait until hour five."
How Advanced Imaging Decides Who Qualifies Late
Late-window eligibility hinges on distinguishing dead tissue from tissue at risk. Non-contrast CT rules out hemorrhage, but CT perfusion and MRI diffusion-weighted imaging do the finer work. CT perfusion produces maps of cerebral blood flow, cerebral blood volume, and Tmax, the time to maximum concentration of contrast, and the mismatch between them approximates the salvageable penumbra.
The DEFUSE 3 criteria specify a core of less than 70 milliliters and a mismatch ratio of at least 1.8 between the hypoperfused tissue and the core. DAWN used a different mismatch definition tied to age and NIHSS. These are not interchangeable, and a center should know which protocol it is applying.
Here is where real-world friction appears. CT perfusion adds processing time and requires trained interpretation. A 2021 study in Stroke found substantial variability among clinicians reading the same perfusion scans, which is why automated software such as RAPID has become common. The software does not replace judgment, but it standardizes the numbers the guidelines rely on.
A failure mode worth naming: a patient arrives at hour 10 with a favorable perfusion profile but the CT perfusion software fails or the scan is uninterpretable due to patient motion. The team then faces a decision without the qualifying imaging. Guidelines do not authorize late thrombectomy on clinical grounds alone, so the safest path is usually transfer or a documented multidisciplinary judgment, not a solo call.
What the Numbers Look Like at Each Cutoff
| Time from last known well | Guideline position | What must be true | Typical selection tool |
|---|---|---|---|
| 0β6 hours | Thrombectomy recommended for anterior large-vessel occlusion | Confirmed occlusion, no hemorrhage, measurable deficit | Non-contrast CT plus CT angiography |
| 6β16 hours | Thrombectomy recommended in selected patients | Favorable perfusion mismatch | CT perfusion or MRI |
| 16β24 hours | Thrombectomy recommended in selected patients | DAWN-type clinical-imaging mismatch | MRI diffusion-weighted imaging or CT perfusion |
| Beyond 24 hours | Not established by randomized evidence | Individualized, usually trial or registry context | Not standardized |
The table compresses a lot of nuance, and the rows are not equally solid. The 0-to-6-hour row rests on five randomized trials. The 16-to-24-hour row rests substantially on DAWN, a single trial with a modest sample. That asymmetry matters when consenting a family at 2 a.m.
Notice also what the table does not say: there is no row where "the patient looked bad" substitutes for imaging. Clinical impression alone never satisfies the criteria.
Why the Door-to-Groin Clock Is a Different Number
The timeframe question has two clocks, and confusing them causes real harm. The first is symptom onset to treatment, the window discussed above. The second is door-to-groin puncture, the hospital's internal speed once the patient arrives. Guidelines from the AHA/ASA set a target door-to-groin puncture time of 90 minutes or less for direct arrivals and 60 minutes or less for transferred patients, per the 2019 guideline update.
The 60-minute transfer target is aggressive, and many centers miss it. A 2020 analysis in the Journal of NeuroInterventional Surgery found that only a minority of transferred patients met the 60-minute benchmark, with median times closer to 90 minutes or longer depending on the system. That gap is where quality improvement budgets should go.
The two clocks interact. A patient arriving at hour 5 with a 90-minute door-to-groin time still falls within the 6-hour window. The same patient arriving at hour 5 with a 3-hour door-to-groin time does not. This is why rapid imaging and clear transport protocols change eligibility, not just speed.
For hospital administrators, the arithmetic is blunt. Shaving 20 minutes off door-to-groin for the eligible population is worth more, in functional outcomes, than adding a new scanner that sits idle overnight.
The Exception Cases That Break the Rules
Posterior circulation strokes β basilar artery occlusion β do not fit neatly into either window. The 2015 trials largely excluded them because the anatomy and clinical course differ. The BEST and BASICS trials, published in the New England Journal of Medicine in 2020 and 2021, yielded mixed results on thrombectomy for basilar occlusion, and the field has leaned toward treating selected patients well beyond 24 hours when the deficit is severe and imaging is favorable. No guideline gives a clean cutoff here.
Wake-up strokes and strokes of unknown onset are another edge case. The last known well may be hours before the patient woke, so the clock is uncertain. MRI-based selection, specifically DWI-FLAIR mismatch, is used in some protocols to identify patients who likely had onset within 4.5 hours and may still qualify for thrombolysis, per the WAKE-UP trial published in the New England Journal of Medicine in 2018. For thrombectomy, the DAWN and DEFUSE 3 approach β imaging-based rather than clock-based β is the practical answer.
Pediatric stroke is a third gap. Thrombectomy in children is not supported by large randomized trials, and the 24-hour window has not been validated in that population. Protocols exist at select centers, but the general rule does not transfer.
A fourth case: the patient with a large established core and a severe deficit who presents at hour 8. Late-window trials largely excluded large-core patients initially, though the SELECT2 and ANGEL-ASPECT trials published in 2023 in the New England Journal of Medicine suggested benefit in selected large-core patients. That area is still moving.
How Prehospital Triage Changes the Effective Window
The clinical window starts when the patient is last known well, but the effective window β the time available to actually get treatment β depends on how fast the system moves. Mobile stroke units, which carry CT scanners to the scene, shorten time to thrombolysis but not necessarily to thrombectomy, since the angio suite is still at the hospital. A 2021 analysis in the New England Journal of Medicine of mobile stroke units in Germany and the US found faster thrombolysis but no clear improvement in functional outcomes in one of the larger trials, a result that took some of the shine off the concept.
Prehospital large-vessel occlusion screening scales, such as the RACE scale, help paramedics route suspected large-vessel occlusions directly to comprehensive centers rather than the nearest hospital. Getting that right shortens the effective window without changing the formal criteria.
The Los Angeles Motor Scale and the Cincinnati Prehospital Stroke Severity Scale serve similar purposes. None has perfect sensitivity and specificity, and over-triage can overwhelm comprehensive centers, so the choice is a system-level tradeoff.
The takeaway for a stroke coordinator: the formal window is fixed by evidence, but the ability to use it depends on dispatch, triage, transport, imaging, and lab turnaround working in parallel. A center that shaves 30 minutes off its internal process effectively extends its usable window by 30 minutes.
Common Misreadings of the Window, and What They Cost
The single most common mistake is treating 24 hours as a universal green light. A patient who presents at hour 20 with a large core and no mismatch does not qualify, and proceeding anyway exposes them to procedural risk without established benefit. The guidelines say "selected patients"; the selection criteria are doing the work.
A second mistake is assuming the 6-hour rule applies to all strokes. It applies to anterior-circulation large-vessel occlusion. Lacunar strokes, small-vessel disease, and most posterior-circulation events follow different logic, and the 2015 trials did not study them.
A third is confusing symptom onset with the time the patient was found. Emergency medical services documentation should record last known well, and an unclear history should trigger imaging-based selection rather than a guess.
A fourth, subtler error: letting the door-to-groin clock crowd out the door-to-needle clock for thrombolysis. The two run in parallel, and the alteplase decision should not wait on thrombectomy logistics. Hospitals that defer thrombolysis until the angiography team is ready lose the thrombolysis benefit in exchange for nothing.
FAQ
Can thrombectomy be done after 24 hours?
Only in narrow, non-standardized circumstances, mostly posterior circulation and usually within research or registry protocols. The randomized evidence base for anterior-circulation thrombectomy stops at 24 hours, with DAWN providing the late-window data. A clinician considering treatment beyond that point is making a judgment call, not following a guideline.
Does the 24-hour window apply if CT perfusion is unavailable?
No. The late-window recommendation depends on advanced imaging demonstrating a favorable mismatch, and non-contrast CT alone cannot establish that. A center without perfusion or MRI capability should transfer the patient rather than extend the window on clinical grounds. The imaging is the qualifier, not a formality.
Is the 6-hour window measured from symptom onset or from arrival?
From the last known well time, which is usually symptom onset but can be earlier if the patient was found down or woke with symptoms. Door-to-groin puncture time is a separate metric with its own 90-minute target for direct arrivals and 60 minutes for transfers. Confusing the two clocks is a common protocol error.
What is the door-to-groin target for a transferred patient?
Sixty minutes or less, per the 2019 AHA/ASA guideline update. Real-world performance often lags that target, with published series showing median times closer to 90 minutes for transfers. The gap is a legitimate quality-improvement target, and closing it extends the usable window without changing any formal criteria.
Do wake-up strokes still qualify for thrombectomy?
Yes, when imaging supports it. The clock cannot be pinned to a specific hour, so selection relies on mismatch imaging, similar to the DAWN and DEFUSE 3 approach. The WAKE-UP trial validated MRI-based selection for thrombolysis in this scenario, and the same imaging-first logic governs thrombectomy decisions.
How much does each minute of delay actually cost?
A widely cited Stroke analysis estimated each minute of delay to reperfusion costs about 1.9 days of healthy life, and each 15-minute delay roughly one month. Those figures come from pooled trial data and are used for quality improvement, not individual prognosis. They are the strongest available argument for parallel processing in the emergency department.
Why do the 2015 trials say 6 hours while later guidelines say 24?
The 2015 trials enrolled within 6 hours and established efficacy there. DAWN, published in 2018, tested 6 to 24 hours in patients selected by clinical-imaging mismatch and found benefit. The guideline extension reflects the new evidence, not a reinterpretation of the old trials, and the 6-hour standard remains the strongest-evidenced window.
Can a patient receive thrombolysis and thrombectomy together?
Yes, and most eligible patients do. Intravenous alteplase is given within 4.5 hours when criteria are met, and thrombectomy follows if a large-vessel occlusion is confirmed. The treatments are sequenced rather than exclusive, and the alteplase decision should not be delayed while thrombectomy logistics are arranged.
This article is general educational information about stroke care timing and is not a diagnosis, treatment plan, or substitute for clinical judgment. Eligibility for endovascular therapy depends on individual imaging, examination, and history, and those decisions belong to a qualified local stroke team. If you or someone near you has sudden weakness, facial droop, speech difficulty, or vision loss, call local emergency services immediately rather than waiting to see whether symptoms resolve.