Neurosurgical ICU (Neurological Intensive Care Unit)

The monitor bolted into the skull did not beat the bedside examination. The nurse waking you every hour is the monitor.

In the only large randomized comparison, managing severe head injury by an intracranial pressure monitor gave no better survival and no better recovery than managing it by the clinical examination and repeated scans. Six month mortality came out at 39 percent against 41 percent. Everything a neurosurgical intensive care unit does runs off that examination.

What makes this unit different from any other

A neurosurgical intensive care unit treats one organ that behaves unlike every other organ in the body. The brain sits inside a rigid box of fixed volume, so anything taking up extra room inside the skull, whether that is blood from a bleed, swollen tissue around an injury, or fluid that has stopped draining, has to push something else out of the way, and what gets pushed out is blood flow to tissue that dies within minutes of losing it. The whole unit exists to catch that happening before it finishes.

Why the staffing looks the way it does

General intensive care units are built around the heart and the lungs, and their alarms fire on blood pressure, oxygen and rhythm. Here the decisive change stays silent on every machine in the room. A pupil that was equal an hour ago and is now larger on one side, an arm that pushed your hand away this morning and drifts today, a patient who opened his eyes to his name at noon and needs a shout at four. None of that trips an alarm. Not one. A nurse trained to look for them finds them, and the difference between finding it at twenty minutes and finding it at two hours decides how much brain is still there afterwards, which is a sentence worth reading twice before comparing hospitals on anything else. That is the whole argument for a separate unit with separately trained staff, and it also makes the nursing ratio the first number to ask for, ahead of anything on the equipment list.

The examination that outranks the machines

Intracranial pressure monitoring has been standard practice in severe head injury for decades, and in 2012 somebody finally tested it properly.

Chesnut and colleagues, writing in the New England Journal of Medicine, randomized 324 patients with severe traumatic brain injury to management guided by an intraparenchymal pressure monitor or to management guided by the clinical examination and repeated imaging. The composite outcome, built from 21 measures of function and cognition at three and six months, came out at 56 against 53 with no significant difference. Six month mortality was 39 percent in the monitored group and 41 percent in the examination group, and intensive care stay ran 12 days against 9. One clear difference did emerge, and it ran opposite to what most people expect, because the group managed without a monitor received more brain-directed treatment than the monitored group did, at 4.8 treatment days against 3.4. Read that last figure carefully, because it explains what the monitor is actually for. Rather than finding trouble the examination misses, it tells a team when to stop treating, which spares a patient hyperosmolar fluids and spells of hyperventilation nobody needed to give.

Why monitors still go in

The reasons are good ones.

A patient deep under sedation cannot be examined at all, a patient in a scanner is out of reach for half an hour, and a pressure trend running upward overnight shows itself before a pupil does. What the trial settled is the hierarchy. The examination leads and the numbers support it, and a unit that treats the screen instead of the patient has the order backwards.

What the numbers on the screen mean

Families spend hours watching these figures without being told which ones matter, and the screen makes no distinction between a number the team is acting on and a number it is simply recording, which is how a relative ends up frightened by a heart rate while the figure that actually changed the plan scrolled past unremarked an hour earlier. Below is the working set, in the order a neurosurgical team reads it, with what moves each one. Learn these six and the room stops being noise.

Narrow screens scroll this table sideways. Swipe or drag to reach every column.

Single readings mean very little. Direction over hours is what the team acts on, so a ward round here spends longer looking at the trend line than at the number currently on the screen, and a relative who asks what the pressure is right now will get an answer covering the last six hours instead.

Who ends up in a neurosurgical intensive care unit

Five broad groups arrive here, and the reason for admission differs sharply between them, so a single answer to how long the stay lasts does not exist. The first two account for most of the beds.

Those two groups look identical from the corridor while being entirely different situations, so ask your surgeon which one applies to your case, because the answer changes what the first forty eight hours mean and it changes what a quiet night is evidence of. Put the question before the operation.

Two further groups turn up regularly. Patients whose fluid spaces have blocked and who need a drain, and patients admitted after a seizure that will not stop. A fifth group is admitted purely for the monitoring that no ward can provide, usually after an aneurysm bleed, where the danger arrives on a schedule that the first days give no hint of.

Sedation, and the problem it creates

Every drug that keeps a ventilated patient comfortable also switches off the examination the unit depends on, and that tension runs through every single day of a neurosurgical intensive care stay without any clean solution to it, only a compromise that gets managed shift by shift. Nobody pretends otherwise.

Sedation holds, and what they are for

Sedation is kept as light as the situation tolerates, and it is lifted deliberately.

Sedation holds work like this. The drugs get stopped or reduced at a planned moment, in the morning with the full team present, so the patient can be examined properly. The team wants to see eyes opening, a hand squeezing on request, both sides of the body moving equally, and the ventilator being fought rather than accepted. Then the sedation goes back on before pressure or agitation becomes a problem. Where the pressure inside the skull is unstable, the hold gets postponed, because rousing a patient raises that pressure for several minutes and the price is not worth paying that morning.

Families watch this moment and read it as the patient waking up, when it is a test of the examination and nothing more, so ask the team directly which of the two they believe they are looking at. They will say.

When the pressure will not come down

Treatment for raised pressure runs as a staircase. Each step carries more risk than the step below it, so nobody skips ahead, and most patients never leave the first two. The upper steps are where the decisions get hard, and where a family needs to understand what is being traded before anyone asks them to agree to it.

1. Simple measures first. Head of the bed raised, the neck straight so the veins draining the head are not kinked, fever treated aggressively because every degree raises the brain demand for blood, and pain and coughing controlled.
2. Deeper sedation, and drainage of cerebrospinal fluid through a drain if one is already in place. Removing a few milliliters of fluid from a rigid box drops the pressure immediately.
3. Hyperosmolar treatment, meaning concentrated salt solution or mannitol given through a central line, which pulls water out of brain tissue for a period of hours and buys time for something definitive.
4. Brief controlled hyperventilation, which shrinks the blood vessels and the volume of blood inside the skull within minutes. It is a rescue measure for a deteriorating patient on the way to a scanner, because holding it for long starves the tissue it is meant to protect.
5. Barbiturate coma, which shuts brain activity down to its lowest metabolic setting. The price is steep, because it drops blood pressure, suppresses immunity and removes the examination completely for days on end.
6. Removing part of the skull. Swelling tissue then has somewhere to go.

That last step needs its own paragraph. The evidence behind it is the most sobering in this specialty.

Hutchinson and colleagues randomized 408 patients with head injury and pressure above 25 mmHg that had resisted everything else, reporting in the New England Journal of Medicine in 2016. At twelve months, death had occurred in 30.4 percent of the surgical group against 52.0 percent of the medical group, so the operation saved roughly one life in five. The survivors did not arrive evenly across the outcome scale. Vegetative state ran at 6.2 percent against 1.7 percent, and severe disability with dependence on others for care ran at 18.0 percent against 14.0 percent. Good recovery came out at 9.8 percent against 8.4 percent, which is the same within chance.

Removing part of the skull converts deaths into survivors, and most of those survivors live with severe disability. Surgeons put that in front of families before the operation rather than after it.

Two treatments that sounded right and were not

Neurocritical care has a particular history of ideas that made perfect physiological sense and then failed when somebody tested them. Both examples below changed practice by being negative.

Cooling the brain

Lowering body temperature reduces the brain demand for oxygen and reliably brings intracranial pressure down, so cooling looked like an obvious treatment for years. Andrews and colleagues enrolled 387 patients across 47 centers in 18 countries, randomizing those whose pressure stayed above 20 mmHg despite basic measures to cooling plus standard care or to standard care alone. Recruitment was suspended on safety grounds before the planned number was reached. Cooling did what it was supposed to do, reducing the need for the harsher later steps from 54 percent to 44 percent, and the patients did worse anyway. A favorable outcome at six months occurred in 26 percent of the cooled group against 37 percent of the others.

Why that result mattered beyond cooling

Cooling hit its physiological target and missed the only target that counts. Controlling a number and helping a patient turned out to be two separate achievements, and the trial is now the standard example used to explain why those two get confused.

Driving the blood sugar down

Tight glucose control was adopted widely in intensive care before the neurological evidence existed. A 2018 meta-analysis in Critical Care pooled ten randomized trials covering 1,066 head injury patients and found no mortality difference between tight and conventional targets, a borderline advantage for tight control on neurological outcome, and a fivefold higher rate of severe hypoglycemia with it. An injured brain tolerates a low sugar worse than almost any other tissue, so most units now run a moderate target and check the level several times a day. Chasing a number into a range the brain cannot survive is the same mistake the cooling trial made, in a different unit of measurement.

Neither result should make you distrust intensive care. Both are reasons to ask what a unit does and why it does it, because a unit still cooling head injuries to bring a pressure number down in 2026 is working from a 2009 understanding of the evidence and has not read the trial that closed the question.

The cheap tablet that genuinely works

After an aneurysm bleeds into the space around the brain, the arteries there can narrow days later and starve territory that survived the original bleed. This delayed injury, rather than the bleed itself, is what takes a good proportion of patients who reach hospital alive.

An oral calcium channel blocker reduces it. A Cochrane review pooled 16 trials covering 3,361 patients and found that oral nimodipine cut the risk of a poor outcome by roughly a third, with a relative risk of 0.67 and a confidence interval running from 0.55 to 0.81. Given by vein it did not show the same benefit, and other drugs in the same class did not either, which leaves an old, inexpensive and thoroughly unglamorous tablet standing as one of the very few treatments in this field with a clear positive trial behind it.

How it is given

Three weeks of it, by mouth or by tube, starting as soon as the diagnosis is made.

The question to ask on day one

Ask whether it has started. No question a family asks on the first day after an aneurysm bleed is more useful.

What families see, and what it actually means

Almost everything alarming in this room has an explanation. Most of them sound far less frightening once somebody says them out loud, and the list below covers what relatives ask first, in roughly the order they notice things.

Columns run past the edge on a small screen. Slide the table across to read them.

Get the nurse to explain anything you have been staring at. Nobody here minds.

Why nobody will give you a number in the first week

Families want a percentage. An early percentage would be worse than no answer at all.

Two reasons, and they compound each other. The first is biological. Swelling has not finished, sedation is still in the system, and the brain has a slow recovery curve that does not begin to show itself for weeks, so an examination on day three describes day three and very little else. The second is a trap the specialty has documented at length. A pessimistic prediction made early tends to change the treatment that follows it, treatment that has been reduced produces a worse outcome, and the worse outcome then appears to confirm the prediction that caused it.

How good units guard against it

Deliberate safeguards exist, and good units write them down.

They set a period during which full treatment continues whatever anyone suspects, they involve more than one senior clinician in any conversation on limiting treatment, and they keep the examination going across enough days for a trend to appear.

What a prognosis conversation should contain

When that conversation does happen, it should reference the imaging, the examination and its trajectory, and any electrical testing that has been done, and it should come with an explicit statement of how certain the team is. Any team that answers a percentage question on day two is guessing.

Leaving the unit, and the road after it

Moving to a neurosurgical ward happens once the examination has held steady for a day or more, pressure control no longer needs drugs, the airway is safe and any drain has come out. Alarms go quiet. One to one nursing stops, and most families find the first ward day harder than they expected it to be, because the constant watching they had learned to read as danger had also become the thing reassuring them, and its removal registers as abandonment before it registers as progress.

Recovery here runs on a different clock. Useful improvement continues for months, and in younger patients for a year or more, which is the encouraging part and the exhausting part at the same time, because a family that expects the curve to finish in six weeks will read a normal recovery as a failure and give up on therapy that was still working. Early gains come quickly and then the curve flattens into slow work, and the things that come back last are attention, memory, word finding and the ability to hold a plan in mind, none of which show up on a scan and all of which decide whether somebody returns to their old life. Physiotherapy, occupational therapy and speech and language therapy carry that stage, and starting them while the patient is still an inpatient matters more than which building they happen in.

Putting the skull back

Where part of the skull was removed, it goes back.

Surgeons store the bone and replace it at a second operation called a cranioplasty, once the swelling has fully settled, which takes weeks to months. Until then a protective helmet is worn for anything active, and your team gives you the timing in writing so the second operation can be planned around travel instead of the other way round.

Planning the trip, the stay and the flight home

Neurosurgical cases divide into two travel situations, and they need different planning.

A planned operation with a booked bed is straightforward to arrange. Send the imaging first, get the written opinion, then travel with the dates fixed. An emergency admission during a trip is a different problem, and the thing to understand in advance is that flying is the constrained part. Air travel after a brain injury or intracranial surgery carries specific issues that have nothing to do with how well the patient looks. Air trapped inside the skull after an operation expands as the cabin pressure drops, the lower oxygen at cabin altitude is tolerated poorly by injured brain, seizure risk is highest in the early weeks, and a patient who still needs oxygen or suction needs an airline that will carry them, and an escort as well. Your surgeon sets the date, it is a separate clearance from discharge, and for intracranial surgery it usually runs several weeks rather than several days.

Book a changeable return ticket. Buy insurance covering an extended stay and a medical repatriation, and read that policy properly before you buy on price, because repatriation for a neurosurgical patient means a stretcher, an escort and sometimes an air ambulance, which is the expensive kind and the kind cheap policies quietly exclude.

Send what you already have before you commit to anything. For a brain or spine problem that means the CT and MRI images themselves and not only the report, any angiogram, the operation note if surgery has already happened, the current medicine list with doses, and a plain description of what the patient can do today compared with a week ago. A neurosurgeon and an intensive care physician read them together and tell you what the plan would look like, whether a reserved bed forms part of it, and where they would advise against moving the patient at all. That review is free. It commits you to nothing. For the paperwork, our international patients office puts your appointment confirmation together with an invitation letter carrying the hospital name and the name of the treating doctor, which is the document most consulates ask for with a medical visa application, and it goes out ten days or so ahead of travel.

After you get back home

Follow up for a neurosurgical patient happens mostly where you live, and it gets arranged before you leave, never improvised afterwards. You go home with the discharge summary, the operation note, the pathology result where there is one, and the imaging on disc, which together let a neurologist or neurosurgeon at home take over without requesting records from another country. Your coordinator stays reachable on the same WhatsApp number, so a question about a wound, a drug dose or a scan result reaches somebody who already knows the case. Remote check-ins follow intervals that match the operation, and you send the local scan reports in as they are done so the surgeon here looks at the same pictures your own doctor has. Two things send you straight to a local emergency department first, and they are a seizure, and any new or worsening weakness, confusion or severe headache, none of which should wait for a message to reach another country before somebody puts a scanner over the patient. Tell us afterwards. Tell us early.

Conversations here are harder than on almost any other page of this site, so language matters more. Seven languages are covered directly by our international patients team, namely English, Arabic, French, Russian, Serbian, Romanian and Spanish, with interpreting in anything else arranged on request. One coordinator takes your case from the first message, sits in on the conversations that matter, follows it through discharge and keeps the same WhatsApp number live once you have flown home. Visiting inside the unit follows the unit rules, which your coordinator sends you in writing before you arrive, and once the patient moves to a normal room one person stays overnight because the rooms have a companion bed. Halal, vegetarian and diabetic diets come from the hospital kitchen, there is a prayer room on site, and a request for a female physician goes to the department and is accommodated wherever the rota allows.

What moves the cost of a neurosurgical intensive care stay

No published figure will apply to your case, because the cost of this stay follows what the stay turns into. Learn which items move it instead.

Drag the table sideways where a phone cuts off the last column.

Health the patient arrives with moves the figure before anything happens. Age, diabetes, heart and lung disease, blood thinning medication and a previous operation at the same site all raise the chance of an extra night and an extra intervention.

Packages published by hospitals and medical travel agencies in this market typically cover the surgeon and anesthesia fees, the planned ward nights, pre-operative tests, any implant named in the quote, an interpreter, transfers and a set number of hotel nights. Flights, travel insurance, extra hospital nights and the treatment of a complication sit outside almost every one of them, and so does intensive care beyond the nights the quote assumed. On a neurosurgical case that last exclusion is the one that matters.

Five questions to put in writing before you commit

How many intensive care nights does this figure assume, and what happens financially if the count goes up. Does it include monitoring hardware and any tracheostomy. Would a return to theater be covered, or billed separately afterwards as a fresh admission with its own nights attached. Is the cranioplasty quoted at all, and if it is, when does that quote expire and what happens to the price if the second operation has to wait six months for the swelling to settle.

Any number that means something comes from a clinician who has read the imaging. That reading costs you nothing.

Neurosurgical ICU FAQ

References

• Chesnut RM, Temkin N, Carney N, Dikmen S, Rondina C, Videtta W, et al. A trial of intracranial-pressure monitoring in traumatic brain injury. New England Journal of Medicine. 2012;367(26):2471-2481.
• Hutchinson PJ, Kolias AG, Timofeev IS, Corteen EA, Czosnyka M, Timothy J, et al. Trial of decompressive craniectomy for traumatic intracranial hypertension. New England Journal of Medicine. 2016;375(12):1119-1130.
• Andrews PJD, Sinclair HL, Rodriguez A, Harris BA, Battison CG, Rhodes JKJ, et al. Hypothermia for intracranial hypertension after traumatic brain injury. New England Journal of Medicine. 2015;373(25):2403-2412.
• Dorhout Mees SM, Rinkel GJE, Feigin VL, Algra A, van den Bergh WM, Vermeulen M, et al. Calcium antagonists for aneurysmal subarachnoid haemorrhage. Cochrane Database of Systematic Reviews. 2007;2007(3):CD000277.
• Hermanides J, Plummer MP, Finnis M, Deane AM, Coles JP, Menon DK. Glycaemic control targets after traumatic brain injury. A systematic review and meta-analysis. Critical Care. 2018;22(1):11.