Armghan Ans: The 2026 Stroke Guidelines Are Excellent, Now the Hard Part

Armghan Ans, Asst. Professor and Director of Stroke at UPMC Washington, Founder of MDAdopt™, shared on LinkedIn:

”The 2026 AHA/ASA stroke guidelines dropped in January.

Within days, a neurologist on Reddit wrote:

‘These guidelines were written like they were crafted by people recently gifted shares in imaging companies.’

That’s not entirely fair.

But it points to something real.

Not about the guidelines. About the gap.

119 pages. 7 major changes. Each one has a version of this problem.

For example:

• A 4.5–9 hour thrombolysis window that requires imaging most hospitals don’t have
• DOAC ‘relative contraindications’ that turn a 10-second rule into a real-time judgment call
• EVT expansion into ASPECTS 3–5, the least reliable scoring range

The science is not the problem.

Implementation is.

I broke down what actually changed, and what it demands from the front lines:

What changed, what it demands, and what nobody is talking about yet.

The 2026 AHA/ASA Acute Ischemic Stroke Guidelines dropped on January 26th. 119 pages.

The most significant update since 2019. Within days, a neurologist on Reddit summarized his reaction this way: ‘My takeaway is that these guidelines were written like they were crafted exclusively by people recently gifted shares in imaging companies.’

That line isn’t entirely fair. But it tells you something important, not about the guidelines themselves, but about the gap between what they recommend and what most hospitals can actually do.

This newsletter is about that gap. Not to criticize the guidelines.

They are genuinely excellent. They reflect a decade of landmark trial evidence and represent real progress for patients.

But excellent guidelines and implemented guidelines are two different things. And the distance between them is where stroke care actually lives.

What Actually Changed

The 2026 guidelines replace the 2018 document and its 2019 update. The changes that matter most, before and after, in plain language:

Tenecteplase (TNK) moved from a narrow Class IIb recommendation to COR 1, LOE A, now effectively co-equal with alteplase for all eligible patients within 4.5 hours.

Single bolus. No infusion pump.

• Post-EVT blood pressure:

Targeting SBP <140 mmHg after successful recanalization is now COR 3: Harm.

The previous practice of aggressive BP lowering post-EVT does not improve outcomes and may cause harm.

• Large core EVT:

Selected patients with ASPECTS 3-5 meeting specified clinical and imaging criteria are now COR 1 candidates for thrombectomy, based on SELECT2 and ANGEL-ASPECT trials.

EVT eligibility just expanded into territory it had never entered before.

• Basilar artery occlusion:

EVT within 24 hours is now strongly recommended for selected patients who meet specified criteria (baseline mRS 0-1, NIHSS ≥10, and PC-ASPECTS ≥6), based on the ATTENTION and BAOCHE trials.

• Mobile stroke units:

COR 1, LOE A for the first time. Where available, MSUs are now guideline-recommended over conventional EMS for thrombolytic-eligible patients.

• Extended thrombolysis window:

IVT from 4.5 to 9 hours is now supported in selected patients using advanced imaging to confirm salvageable tissue.

• Pediatric stroke:

First-ever recommendations for IVT and EVT in children, from 28 days to 18 years of age.

Each one of these is a genuine advance. Each one also carries a workflow cost that the guideline document cannot pay for you.

What This Feels Like From the Floor

This is the part of the guidelines conversation that rarely happens in academic circles. Seven changes. Seven implementation problems at the hospitals where most stroke patients actually present first.

1. The Extended Window You Can’t See Through

The 2026 guidelines now support thrombolysis from 4.5 to 9 hours using CT perfusion or MRI DWI-FLAIR mismatch to identify salvageable tissue.

A wake-up stroke patient who would have been sent home in 2019 may now be a treatment candidate. That is transformative — in theory.

In practice, the guidelines state that centers receiving stroke patients must now have the capability to perform and interpret CT or MR perfusion imaging, or arrange rapid transfer to centers that can.

A statewide Texas Medicare study found that only 3% of acute stroke patients actually received CT perfusion imaging.

Only 14% of hospitals were even capable of performing it.

CT perfusion availability is concentrated in urban centers, with many non-urban regions lacking immediate access.

The extended window exists on paper. For a significant portion of American hospitals, there is no imaging to open it.

The treatment opportunity is real.

The infrastructure to access it is not. And this is precisely what the Reddit neurologist was pointing at.

The 4.5-to-9-hour window is only accessible to hospitals that have the technology to see inside it. That is not most hospitals.

2. DOAC Reclassification –  One Rule Became Fifty Judgment Calls

Before 2026, DOAC use within 48 hours was simple. Hard stop. No thrombolysis. Every physician, every nurse, every protocol treated it the same way.

The 2026 guidelines replaced that single rule with a color-coded risk gradient.

Recent DOAC exposure is now a relative contraindication (orange on Table 8, not red).

The decision now depends on which agent, what dose, when the last dose was taken, the patient’s renal function, and a real-time risk-benefit assessment.

This affects an estimated one in six stroke patients otherwise eligible for thrombolysis. The science supports the change.

A study of nearly 49,000 patients with recent DOAC use found that thrombolysis was associated with improved functional outcomes, with a symptomatic intracranial hemorrhage rate of 3.5%, suggesting acceptable safety in this registry analysis.

The evidence is there. The treatment benefit is real.

The operational problem is time. By the time an ED physician has confirmed the last DOAC dose, checked renal function, located Table 8, and made a defensible real-time decision, minutes have passed.

In a stroke, those minutes are neurons.

Every hospital that ran a binary DOAC checklist now needs a nuanced risk stratification protocol, trained physicians, and documentation infrastructure that most haven’t built yet.

3. ASPECTS 3-5 — Expanding Eligibility Into the Hardest Zone to Score

The guidelines now recommend EVT for selected patients with large core infarcts (ASPECTS 3-5) based on compelling trial evidence.

More patients eligible for a potentially life-changing treatment. That is unambiguously good.

Here is the problem nobody is discussing: ASPECTS scoring is least reliable exactly in the 3-6 range.

Research shows that even among experienced readers, agreement on ASPECTS in the mid-range category is weaker than at extreme values, the precise zone where the new EVT eligibility boundary now sits.

Even expert neurologists and neuroradiologists benefit meaningfully from AI-assisted decision support tools when scoring in this range.

A multi-reader study found that e-ASPECTS software significantly improved ASPECTS accuracy even among board-certified US specialists.

Without that support, at a community hospital at 2 am, the margin for error on an ASPECTS 4 versus an ASPECTS 6 is real. And the clinical consequence of that error is now larger than it has ever been.

4. Transfer Decisions –  The Judgment Call Nobody Can Make for You

The 2026 guidelines updated transport destination recommendations: when local systems lack rapid interhospital transfer capability, direct transport to an EVT-capable center should be considered. Sensible guidance. Impossibly context-dependent in practice.

The data on why this matters: in a national study of 108,913 stroke transfers, the median door-in-door-out time was 174 minutes.

The Joint Commission benchmark is 120 minutes. In one large hub-and-spoke network, only 18% of transferred patients achieved even a 90-minute DIDO.

DIDO time is a significant independent predictor of 90-day functional outcome. Every minute at the wrong hospital costs neurons.

But bypass is not automatically better. The RACECAT trial, the only high-quality randomized data on this question, found no overall 90-day outcome benefit of direct transport to a thrombectomy-capable center in the studied stroke system. A secondary analysis showed worse functional outcomes in hemorrhagic stroke patients who were bypassed.

The guideline’s answer is nuanced: it depends on the local system.

The ED physician’s reality is that they have to make that call at the front door, in real time, without a clear definition of what ‘well-functioning’ means for their system, with a patient deteriorating in front of them.

5. TNK – The Dose Correction Most Hospitals Haven’t Made

Most forward-leaning stroke centers have already switched to tenecteplase. The bolus is simpler, the evidence is strong, and the 2026 guidelines formalized what many had already adopted.

But there is a specific dose correction the guidelines made that has not received enough attention: tenecteplase 0.4 mg/kg is now explicitly not recommended due to a harm signal.

This dose was previously used at some centers for minor stroke presentations. Order sets at hospitals using TNK (including those that switched before the 2026 guidelines) need to be audited for this specific change. Not all of them have been.

For hospitals that have not yet switched, the institutional machinery remains real: formulary committee review, pharmacy training, nursing protocol updates, order set rewrites.

The bedside simplicity of a single bolus does not automatically translate to institutional simplicity.

6. The BP Reversal — Fighting the System to Do the Right Thing

Post-EVT blood pressure management was one of the most significant reversals in the entire guideline.

Targeting SBP <140 mmHg after successful recanalization is now classified as potentially harmful — not merely unhelpful.

The clinical moment this creates: a patient returns from the interventional suite with SBP 155.

The nurse reaches for the antihypertensive. That is what three years of protocol training says to do. That is what the standing order says.

That is what the EMR alert is configured to prompt. The physician has to intervene and say – Don’t.

And they have to do this against the entire weight of the institutional environment, which has not yet been updated.

This is not a controversial scientific question. The evidence is clear.

The problem is that the evidence changed before the systems did.

Protocols, nursing training, and clinical decision support tools built on the old paradigm are still in place at most hospitals. Until they are updated, the physician who knows the new guideline has to fight the old infrastructure every time.

7. Pediatric Stroke — The Recommendation Without a Protocol

Save this one for last, because it deserves to land.

The 2026 guidelines include the first-ever recommendations for IVT and EVT in children with acute ischemic stroke.

Alteplase is now COR 2b for ages 28 days to 18 years within 4.5 hours of disabling deficits. EVT is COR 2a for children 6 and older within 6 to 24 hours with salvageable tissue.

Annual childhood arterial ischemic stroke incidence in North America has been reported at approximately 1.3 to 1.72 per 100,000 children.

It is uncommon enough that most community emergency departments may go years without seeing a single case.

Children are more likely to present with seizure, altered mental status, or stuttering deficits, symptoms that delay recognition and require imaging confirmation before diagnosis.

The guideline writing group co-vice chair called these recommendations ‘a remarkable shift that establishes the foundation for future research.’

The guideline itself acknowledges that much work remains to adapt prehospital and hospital stroke protocols for pediatric patients.

That acknowledgment is important. Because what it means in practice is this: the protocols don’t exist yet at most of the hospitals that will see these patients.

The recommendations call for EVT performed by a neurointerventionalist with pediatric experience.

Most community hospitals don’t have a neurointerventionalist at all.

A child arrives in a community ED with acute hemiplegia.

The guidelines now say there is a treatment pathway. The hospital has never activated it.

There is no pediatric stroke protocol on the wall, no pediatric neurology on call, and no established transfer agreement for pediatric EVT.

The recommendation is correct, necessary, and long overdue. The infrastructure to execute it is years away at most centers.

The Equity Problem the Guidelines Named

The 2026 guidelines made an unusual admission.

They acknowledged directly that stroke center certification is associated with urban location, higher income service areas, and higher profit margins.

The hospitals best positioned to implement these recommendations are not the hospitals where the majority of stroke patients first present.

Mobile stroke units are now COR 1, LOE A.

There is no funding mechanism to give one to a rural hospital system operating on thin margins. Extended window imaging requires CT perfusion infrastructure that urban comprehensive centers have and community hospitals don’t.

ASPECTS AI scoring tools exist and are increasingly used at comprehensive centers, often where expert interpretation is already available.

The guidelines did not create this inequity. But they widened the distance between what the standard of care now requires and what most hospitals can currently deliver.

The Work That Remains

The physicians frustrated by these guidelines are not wrong about the evidence. They are right about the gap.

Every one of the changes described above is scientifically justified. Every one of them is also asking something of hospitals, physicians, nurses, pharmacists, and systems that has not yet been fully accounted for.

Protocol rewrites. Order set audits. Imaging infrastructure. Transfer agreements. Pediatric pathways.

DOAC risk stratification tools. BP alert reconfiguration.

Guidelines don’t implement themselves.

The evidence moves faster than institutions do. And the hospitals that feel that gap most acutely are the ones furthest from the academic centers where the guidelines were written.

That gap is not a failure of the guidelines. It is the work that remains.

Next week: What the 2026 stroke guidelines demand of AI stroke tools — and whether those tools are ready.

About the Author

Armghan Ans, MD, is a practicing neurologist and Director of Stroke who helps neuro-AI companies diagnose why their FDA-cleared products are not getting adopted — and fix workflow, trust, and integration barriers before they become renewal problems.

TAIN – The AI Neurologist explores the gap between AI innovation and clinical adoption. Subscribe for insights from the hospital hallways where AI products either thrive or die.

Sources:

• 2026 AHA/ASA Guideline for the Early Management of Patients With Acute Ischemic Stroke. Prabhakaran S et al. Stroke. January 26, 2026. doi:10.1161/STR.0000000000000513
• Kim Y et al. Utilization and Availability of Advanced Imaging in Patients With Acute Ischemic Stroke. Circulation: Cardiovascular Quality and Outcomes. 2021;14(4):e006989.
• Yaghi S et al. Intravenous Thrombolysis in Patients on Direct Oral Anticoagulants: Analysis of the Get With The Guidelines Stroke Registry. JAHA. 2026.
• Kallmes DF et al. Impact of e-ASPECTS software on the performance of physicians compared to a consensus ground truth: a multi-reader, multi-case study. Frontiers in Neurology. 2023. doi:10.3389/fneur.2023.1221255
• Gonzalez NR, Zachrison KS, Prabhakaran S. Assessing changes, additions to 2026 acute ischemic stroke guideline. ISC 2026 Daily Coverage, American Heart Association. February 2026.
• Stamm B et al. Door-in-Door-out Times for Interhospital Transfer of Patients With Stroke. JAMA. 2023;330(7):636-649.
• Ahmed RA et al. Impact and determinants of door in-door out time for stroke thrombectomy transfers in a large hub-and-spoke network. Interventional Neuroradiology. 2024.
• Puetz V et al. Interobserver Reliability of Baseline Noncontrast CT ASPECTS for Intra-Arterial Stroke Treatment Selection. AJNR. 2012;33(6):1046.
• Pérez de la Ossa N et al. RACECAT Trial. JAMA. 2022.
• Ramos-Pachón A et al. Effect of Bypassing the Closest Stroke Center in Patients with Intracerebral Hemorrhage: Secondary Analysis of RACECAT. JAMA Neurology. 2023.”

Stay updated with Hemostasis Today.