Stroke Unit (Stroke Center)

The strongest treatment in stroke medicine is not a drug. A stroke unit outperforms every one of them.

Pooling 29 randomized trials and 5,902 patients, organized stroke unit care cut the odds of death or dependency by a quarter against a general ward, which works out at six more people in every hundred living independently a year later.

The ward that works better than the medicine

Which treatment for stroke carries the strongest evidence behind it and most people name the clot dissolving drug, or lately the procedure that pulls the clot out with a catheter. Both work and both stay narrow, because each helps a minority of patients who arrive inside a particular window with a particular kind of stroke, while one intervention helps almost everybody who has a stroke, has been tested in randomized trials since the 1980s, and costs nothing beyond organizing the people who are already there.

Being admitted to a stroke unit.

This table scrolls sideways on a phone. Drag it across to reach the last column.

Read the last row twice. No scanner and no catheter produce that effect. It comes from putting the same patients in a defined ward with nurses, doctors, physiotherapists, speech therapists and dietitians who work on stroke every day and meet over each patient every week, Organization does it. The effect comes from putting the same patients in a defined ward and from the unglamorous things that follow, namely the swallow screen before the first drink, the blood sugar and temperature kept in range, the early detection of a chest infection, the mobilization plan, and the fact that somebody owns each of those tasks instead of assuming another team has it.

What the clock actually costs

Time is brain gets repeated so often that the phrase has stopped carrying information. Here it is.

Slide this table sideways on a small screen to reach the last column.

Why an hour in the corridor matters

The three rows above describe a treatment whose value decays continuously, with no deadline to expire at, so a patient treated at 90 minutes does better than the same patient treated at 150 minutes, and both do better than the same patient at four hours. The drug itself does not change across that interval. What changes is how much brain remains salvageable on arrival, and that explains why hospitals measure door to scan and door to needle instead of anything about the medicine itself. A unit that cannot tell you its median door to needle time has not been measuring it.

Recognizing it, and what to do in the next two minutes

Most of the delay in stroke happens before anybody reaches a hospital. Most of that delay comes from the person beside the patient deciding to wait and see.

1. Face. Ask them to smile. One side of the mouth drooping, or an eyelid sitting lower, counts.
2. Arms. Ask them to raise both arms and hold them. One arm drifting down counts, even a little.
3. Speech. Ask them to repeat a sentence. Slurring, wrong words, or a blank look while they search for a word all count.
4. Balance and eyes. Sudden loss of balance, sudden double vision or a part of the visual field going missing count too, and these are the ones people miss because they do not look like the stroke on television.
5. Time. Note the exact minute the symptoms started, or the last minute the person was seen well, and call an ambulance. That single number decides which treatments are still open, so it matters more than any other thing you can supply.

Do not drive them yourself if an ambulance is available. Ambulance crews can alert the hospital before arrival, which starts the scanner and the stroke team moving while the patient is still on the road, and arriving unannounced at a reception desk gives up that head start.

Do not wait to see whether it passes. Symptoms that resolve within an hour are a transient ischemic attack, which is a warning that a stroke is coming and an emergency in its own right.

The first hour inside the doors

Two questions run the first hour in a stroke unit. The order never varies.

Is it a clot or a bleed

A stroke caused by a blocked artery and a stroke caused by a burst one produce the same symptoms and need opposite treatments, so nothing can be given until a scan has separated them. A plain CT of the head takes minutes and answers it. Around 85 percent of strokes turn out to be a blockage and the rest are bleeding, and a drug that dissolves clot would be catastrophic in the second group, which is the entire reason the scan comes first and why no ambulance crew and no doctor will give anything before it.

Where is the blockage

Once a blockage is confirmed, a second question decides whether the patient needs a procedure as well as a drug. A CT angiogram follows the arteries from the neck into the brain and shows whether the blockage sits in one of the large vessels at the base of the brain, and large vessel blockages cause the most severe strokes and respond to mechanical removal in a way they do not respond to drugs alone. Some units add a perfusion scan, which separates brain tissue that has already died from tissue that is still starving and could recover, and that distinction is what opens the late treatment window described further down this page. Two scans, two questions, one decision. Everything else waits.

The drug that dissolves the clot

Intravenous thrombolysis has been the backbone of acute stroke treatment for thirty years and it is frequently explained badly in both directions, since the drug goes into a vein and travels everywhere, dissolving the clot blocking the brain artery and, unavoidably, weakening clot everywhere else in the body. That trade sits at the center of it. Roughly one patient in every sixteen treated within three hours has a better outcome than they would have had, and a small number, in the region of two to three percent, bleed into the brain because of it. Those two facts belong in the same sentence whenever the drug is discussed, because a consent conversation that mentions only the first one is not a consent conversation.

A newer drug, tenecteplase, has largely replaced the older alteplase in many countries and the reason is practical rather than dramatic. In a pragmatic randomized trial of 1,577 patients, 36.9 percent treated with tenecteplase reached a near normal outcome against 34.8 percent with alteplase, a difference of 2.1 percent whose confidence interval crosses zero, with symptomatic bleeding into the brain in 3.4 percent against 3.2 percent. The outcomes match. What tenecteplase changes is the delivery, since it goes in as a single injection over seconds rather than an hour long infusion, which saves real minutes and makes transfer between hospitals far simpler.

Ask which agent the unit uses and what its median door to needle time was last year. Both answers are short.

Pulling the clot out

Mechanical thrombectomy changed severe stroke more than any development in the previous two decades, and it applies to a minority of patients. A catheter goes in through the artery at the groin or the wrist, travels up into the brain under x-ray guidance, and the clot is either grabbed with a stent shaped device or sucked out through a wide tube. Speed still decides the result. The procedure takes under an hour in practiced hands and it only works for blockages in the large arteries at the base of the brain, which is roughly one ischemic stroke in five, and those are the strokes that otherwise leave people unable to speak or to move one side.

Five randomized trials were pooled at the individual patient level, covering 1,287 patients. Thrombectomy reduced disability at 90 days with an adjusted common odds ratio of 2.49 and a confidence interval of 1.76 to 3.53. The number needed to treat to improve one patient by at least one level of disability was 2.6. Numbers like that are rare anywhere in medicine, and they explain why a hospital with a thrombectomy service and a hospital without one are not offering the same thing to a patient with a large vessel blockage.

The late window, six to twenty four hours

For years, arriving late meant nothing could be done. Imaging changed that, because a scan can show which patients still have brain tissue that is starving and salvageable rather than already dead, and those patients benefit long after the clock says they should not. Patients treated between 6 and 24 hours after they were last seen well, selected on exactly that mismatch, reached functional independence in 49 percent of cases at 90 days against 13 percent with standard care alone. Thirty three percentage points separate those two groups, and the second group would previously have been sent to a ward to recover as best they could. Anybody told that a relative arrived too late should ask whether a perfusion or diffusion scan was done before that conclusion was reached, because the late window is exactly where hospitals differ. Running it requires the scanner, the software, an interventional team available at three in the morning and a radiologist willing to make the call, and a unit that has all four will answer the question in one sentence.

When the stroke is a bleed

One stroke in six turns out to be bleeding into the brain and not a blockage, and almost everything on this page so far runs backwards for those patients. No clot dissolving drug, no thrombectomy. The work goes into stopping the bleeding growing, which means bringing a high blood pressure down quickly and carefully, reversing any blood thinning medication the patient takes, and repeating the scan a few hours later to see whether the bleed has expanded. Blood thinners give families something to do immediately, because knowing the exact name and the time of the last dose changes what gets given in the next ten minutes. Bring the boxes if you are bringing anything.

Surgery has a narrow and dramatic place here. In the worst kind of blockage stroke, where a whole middle cerebral artery territory swells and presses on the rest of the brain, removing a panel of skull to let the swelling expand outward saves lives. Pooling three randomized trials covering 93 patients, surgery within 48 hours raised survival from 29 percent to 78 percent, and the proportion reaching a moderate level of disability or better rose from 24 percent to 75 percent. It is offered mainly to younger patients and decided within a day or two. The conversation has to name both what the operation saves and what it leaves undone.

The screen before the first sip of water

Around half of patients cannot swallow safely in the first days after a stroke, and a large share of them have no idea, because the muscles that protect the airway fail silently. Pneumonia follows, caused by food and fluid going into the lungs, with no infection caught on the ward involved, and it is one of the commonest reasons a patient who survived the stroke itself does not survive the admission. Screen first. Prevention comes down to a nurse doing a short, formal swallow screen before anything passes the lips, including water and including tablets, and across 2,532 acute stroke admissions, hospitals with a formal screening protocol had a pneumonia rate of 2.4 percent against 5.4 percent at hospitals without one. A checklist halved it, and it is exactly the kind of thing organized stroke unit care does that a general ward does not.

Ask whether the screen has been done before you offer a relative a glass of water. Families cause this complication with kindness more often than anybody likes to say. Where the screen fails, nobody goes hungry. A speech and language therapist assesses properly, thickened fluids and modified textures follow, and a feeding tube through the nose covers the gap where it takes longer, all of which is ordinary and temporary in most patients, and all of which depends on somebody having done the screen in the first place.

Getting people moving, and the trial that surprised everybody

Early mobilization sounds like something nobody could argue with, and the largest trial of it found the opposite of what it set out to prove.

Two thousand one hundred and four patients were randomized to very early and intensive mobilization, starting within 24 hours of the stroke, or to usual care. Fewer patients in the very early intensive group reached a favorable outcome at three months, 46 percent against 50 percent, with an adjusted odds ratio of 0.73 and a p value of 0.004. Pushing harder and earlier made things worse. What the same data supported was shorter, more frequent sessions started early, and not long intensive ones, and that is what good units now do.

Gentler and more often wins early. Keep this one in mind when a rehabilitation program sells intensity as the product. More is not automatically better in the first days, the evidence says so explicitly, and a unit that quotes hours of therapy per day without mentioning how those hours are distributed is quoting the wrong number.

Later on the picture reverses, and intensity does matter.

Finding out why it happened

Treating the stroke and preventing the next one are different jobs, and the second one is where a stroke unit earns the rest of its reputation. The work runs as a sequence of questions. Are the arteries in the neck narrowed. Is the heart throwing clots, which means looking for an irregular rhythm and looking at the heart with ultrasound. Is the blood pressure, the cholesterol or the diabetes the driver. Is there something less common behind it, which matters most in patients under 55, where a tear in an artery wall or a hole between the heart chambers turns up often enough to be worth hunting for. Order matters here. Each answer changes the medication the patient leaves on, and leaving on the wrong one is how people come back. The order matters as much as the list, because a narrowed neck artery found in week one can be opened before the second stroke, and the same finding three months later has usually arrived too late to matter.

The rhythm that hides

Atrial fibrillation matters more than anything else on that list, because it changes treatment from an antiplatelet drug to a proper anticoagulant and cuts the risk of a second stroke substantially. It comes and goes, so a standard tracing on the ward misses most of it. Among 441 patients whose stroke had no obvious cause, an implanted monitor found atrial fibrillation in 12.4 percent within twelve months against 2.0 percent on conventional follow up, a hazard ratio of 7.3. Six times as many. Anticoagulation changes the arithmetic so completely in that group that finding it late is one of the more expensive misses in the whole of stroke medicine, so if a discharge summary says the cause was not found, the next question is how long the heart rhythm was actually monitored, because 24 hours is not an answer. Ambulatory monitors worn for one to four weeks sit between a single tracing and an implanted device, and they find a useful share of it without surgery.

The months that decide the rest of it

Recovery after a stroke is not a single curve. It runs fastest in the first three months, continues more slowly for a year and beyond, and depends heavily on what happens during that first stretch.

One question separates rehabilitation services. Who sets the goals. A program built around what the patient wants to do again, named specifically, outperforms one built around a standard timetable of sessions, every time.

How long it lasts, and what happens once you are home

Two timetables run side by side after a stroke. Families are usually told neither.

Length of stay

The acute admission runs from a few days to a couple of weeks, depending on the size of the stroke, whether a procedure was needed and how quickly swallowing and walking come back. Inpatient rehabilitation is a separate decision made at the end of that, and it runs on a different scale, commonly three to six weeks for somebody with meaningful weakness or speech loss and longer where the deficit is severe. Dates come after the assessment, never before it. Any quoted length before an assessment is an estimate, and a service that quotes a fixed number of weeks to somebody it has never examined is quoting a package rather than a plan. Fitness to fly turns on the stroke being stable, on blood pressure being controlled and on whether the patient can manage the journey with the help available, and the usual guidance for airlines sits at a few weeks after the event and never at days, and it needs checking before anybody books a return flight on the assumption that discharge and departure fall on the same day.

Follow up once you are home

What travels home decides the next two years. Four documents, and no more. A discharge summary naming the stroke type, the artery involved, what treatment was given and what the cause investigation found or failed to find, then the medication list with the reason for each drug, since an anticoagulant and an antiplatelet are prescribed for different findings and swapping them is how second strokes happen. The therapy handover with the current level of function and the goals still open, written so a therapist at home can continue instead of restarting, and finally a named date for the first review, with blood pressure, cholesterol and heart rhythm the three things being checked.

Request all of it in English where that is what your own doctor reads.

The coordinator stays reachable on WhatsApp once the patient has gone home, which is what rescues a local doctor who needs one detail from the admission. See a doctor the same day for any new weakness, new difficulty speaking, a sudden severe headache, a drooping face or loss of vision, because a second stroke announces itself the same way the first one did.

Whether to travel for this

Traveling for a stroke costs us inquiries to say plainly, and the answer stands anyway.

Nobody flies during a stroke. Acute stroke belongs to whichever hospital is nearest and capable of scanning and treating, because everything on the first half of this page is measured in minutes and a flight costs hours, so a family abroad watching a relative deteriorate should be calling an ambulance and nothing else. Anybody who tells you otherwise is selling something that will hurt the patient.

What does travel

Two things, and both arrive weeks later. The first covers rehabilitation, where the acute treatment went well, the patient is stable, and the therapy available at home is thin or has run out of funded sessions. The second covers the cause investigation, where somebody was discharged with the cause unexplained, on a drug that may be the wrong one, and wants the workup done properly before the second stroke arrives. Both count as planned journeys with a file that can be read in advance, which is the opposite of an emergency. That difference is the whole point. A planned admission lets the receiving team read the scans, agree the goals and book the therapists before the patient lands, and none of that is available to somebody arriving on an ambulance from an airport. Send the file first. A written opinion costs nothing and it arrives before any money has been committed to flights, hotels or a program length that may turn out to be wrong for this patient.

What drives the cost

No figure appears on this page, because a stroke admission and a rehabilitation program are priced on length and intensity, and neither is knowable before somebody has read the file.

Find out what a daily rehabilitation rate covers and what sits outside it, whether the quoted program length is a minimum or an estimate, and what happens financially if the assessment on arrival recommends a longer or shorter stay than the one quoted. Published packages from hospitals in this market are consistent on what they include and vague on what happens when the plan changes, so confirm that part in writing for your own case.

What we arrange

A rehabilitation admission runs for weeks and the family is in it the whole time, so the practical side is part of the treatment.

Check the speech therapy before booking anything. Therapy delivered in a language the patient does not think in does very little, so confirm which language it runs in and who delivers it.

What to ask a stroke service

Every question below has a short answer, and a long vague reply to any of them is itself the answer.

• Is this a stroke unit, meaning a defined ward with its own trained team, or or do stroke patients go to a general ward with a neurologist visiting.
• What was the median door to needle time last year, and what proportion of eligible patients were treated inside 4.5 hours.
• Is thrombectomy available on site around the clock, and if not, which hospital does it and how long does the transfer take.
• Is a formal swallow screen done before any oral intake, and by whom.
• For a rehabilitation admission, who sets the goals, how many hours of therapy per day, in which language, and does the program run at weekends.

Start there. That first question separates the whole field. Everything the evidence at the top of this page describes depends on the answer being the first option rather than the second.

Questions we are asked, a stroke unit FAQ

References

1. Langhorne P, Ramachandra S. Organised inpatient (stroke unit) care for stroke, network meta-analysis. Cochrane Database Syst Rev. 2020;4(4):CD000197.
2. Emberson J, Lees KR, Lyden P, et al. Effect of treatment delay, age, and stroke severity on the effects of intravenous thrombolysis with alteplase for acute ischaemic stroke, a meta-analysis of individual patient data from randomised trials. Lancet. 2014;384(9958):1929-1935.
3. Menon BK, Buck BH, Singh N, et al. Intravenous tenecteplase compared with alteplase for acute ischaemic stroke in Canada (AcT), a pragmatic, multicentre, open-label, registry-linked, randomised, controlled, non-inferiority trial. Lancet. 2022;400(10347):161-169.
4. Goyal M, Menon BK, van Zwam WH, et al. Endovascular thrombectomy after large-vessel ischaemic stroke, a meta-analysis of individual patient data from five randomised trials. Lancet. 2016;387(10029):1723-1731.
5. Nogueira RG, Jadhav AP, Haussen DC, et al. Thrombectomy 6 to 24 hours after stroke with a mismatch between deficit and infarct. N Engl J Med. 2018;378(1):11-21.
6. AVERT Trial Collaboration group. Efficacy and safety of very early mobilisation within 24 h of stroke onset (AVERT), a randomised controlled trial. Lancet. 2015;386(9988):46-55.
7. Hinchey JA, Shephard T, Furie K, et al. Formal dysphagia screening protocols prevent pneumonia. Stroke. 2005;36(9):1972-1976.
8. Vahedi K, Hofmeijer J, Juettler E, et al. Early decompressive surgery in malignant infarction of the middle cerebral artery, a pooled analysis of three randomised controlled trials. Lancet Neurol. 2007;6(3):215-222.
9. Sanna T, Diener HC, Passman RS, et al. Cryptogenic stroke and underlying atrial fibrillation. N Engl J Med. 2014;370(26):2478-2486.