Intensive Care Unit (ICU)
Reaching critical care within four hours of the decision cut ninety day mortality by 16.2 percent across 48 hospitals, and how often that happened depended on how full the unit already was. This page covers the three levels of care, the named units and which patient each takes, what nurse ratios and out of hours cover actually change, and the six questions that separate a good unit from an average one.
About This Department
The equipment looks the same in every unit. How fast you reach it, and who stands next to it, is what differs.
Across 48 hospitals and 12,380 deteriorating ward patients, reaching critical care within four hours of the decision cut ninety day mortality by 16.2 percent, and how often that happened depended on how full the unit already was.
The four hours nobody photographs
Hospital pages on intensive care show the equipment. Families arriving for the first time look at the equipment too, and reasonably so, because a ventilator and eight infusion pumps are the most alarming objects in the building. The variable that moves survival hardest sits somewhere else entirely, in the gap between the moment a doctor decides a patient needs critical care and the moment that patient is actually in the bed. Researchers measured that gap across 48 hospitals. Twelve thousand three hundred and eighty deteriorating ward patients were assessed by a critical care team, and among those recommended for admission, getting there within four hours cut ninety day mortality by 16.2 percent. The number underneath that one is the uncomfortable part. As the unit filled up, prompt admission fell from 22 percent of patients to 9 percent, so the strongest predictor of whether somebody got the timely admission that improved their survival was how many beds happened to be free that afternoon.
No brochure mentions this.
It explains something families notice and cannot name, which is that two hospitals with identical equipment produce different outcomes. Capacity, staffing and the speed of the escalation decide more than the machinery does, and all three are things a family can ask before committing to anything.
What counts as intensive care
Three different levels of support hide inside the phrase, and the letter that arrives with a transferred patient normally names one of them. Knowing which is which turns an alarming document into a readable one.
This table scrolls sideways on a phone. Drag it across to reach the last column.
| Level | What it provides | Where it happens |
|---|---|---|
| Level 1 | A ward bed with extra observation, oxygen and more frequent checks, for a patient at risk of deteriorating | A general ward, sometimes a monitored bay inside it |
| Level 2 | Single organ support, invasive monitoring and one nurse to two patients, for a patient who needs more than a ward and not a ventilator | High dependency unit, sometimes called HDU or step-down |
| Level 3 | Advanced respiratory support or two or more organs supported at once, with one nurse to one patient | Intensive care unit |
Two practical consequences follow. A patient described as needing level 2 care does not need a ventilator, which is usually the reassurance the family is looking for and rarely the sentence they are given. And a transfer from level 3 to level 2 is progress, even though it feels like abandonment, because the nursing ratio halves on the day it happens.
Which unit, and why the name matters
Hospitals split intensive care by the kind of problem, so the name on the door tells you which specialty leads the decisions. It says nothing about how sick the patient is.
Patients move between them during one admission, and that move is a clinical decision instead of an administrative one. Somebody admitted to medical intensive care with pneumonia who then needs an operation for a perforated bowel is transferred because the surgeon needs to lead the plan, and the family should be told that in those words instead of discovering it from a change of ward number.
Children and newborns
Pediatric and neonatal intensive care are separate specialties with their own units, their own equipment sizes and their own staff, and an adult unit does not substitute for either. The differences run deeper than equipment. Drug doses are calculated by weight and change as the child grows through the admission. Ventilator settings for a two kilogram newborn have almost nothing in common with the settings for an adult. Recognizing deterioration in a small child depends on different signs, since children hold their blood pressure until very late and then fall quickly, and a parent stays at the bedside as a matter of course rather than as a concession, because a frightened child settles for their mother in a way no drug replaces. Any hospital offering treatment to a child should be able to say which unit would receive them and who leads it, before anything is booked. Ratios are tighter as well, since the smallest patients are nursed one to one regardless of how stable they look on the monitor, and a pediatric unit that quotes an adult ratio has told you it is not a pediatric unit.
What happens on the ward first
Almost nobody arrives in intensive care from nowhere. Most patients deteriorate on a ward for hours or days beforehand, and what the hospital does during that window decides how many of them need the bed at all.
Early warning scores and the team they summon
Nurses on general wards record pulse, blood pressure, breathing rate, oxygen level, temperature and level of consciousness, and a scoring system converts those six numbers into a single figure. Above a threshold, the score triggers a call to a rapid response team, sometimes called a medical emergency team, which is a group of critical care staff who come to the ward rather than waiting for the patient to reach them. Breathing rate is the most useful of the six measurements and the one most often recorded carelessly, so a unit that takes it seriously reveals how the rest of its monitoring runs.
Check whether the hospital runs one, and who answers the call at three in the morning. A team that exists on paper and is staffed by whoever is free is a different thing from a team with a dedicated rota.
One nurse, how many beds
Four words make the most informative question a family can put to a unit. How many nurses tonight.
Level 3 intensive care assumes one nurse for one patient and level 2 assumes one nurse for two, and those ratios exist because a ventilated patient with four infusions needs somebody in the room rather than nearby. A national study covering every adult intensive care admission in Korean general and tertiary hospitals across five years measured what happens when the ratio slips. The risk of dying in hospital rose as the number of intensive care beds assigned to each nurse increased, with an odds ratio of 1.17, and fell as the number of experienced nurses in the unit increased, with an odds ratio of 0.97. Both findings survived adjustment. The second one matters as much as the first, because experience is not interchangeable with headcount, and a unit that has replaced six senior nurses with six new graduates has the same rota and a different capability. Ask the number. A unit proud of it answers immediately, and a unit that deflects into a sentence on teamwork has answered anyway.
Nights, weekends and the hour of arrival
Patients do not choose when they deteriorate, so the question of whether the hour of admission changes the outcome has been asked repeatedly and answered carefully.
What a national cohort found
Danish researchers followed 44,797 intensive care patients and compared admissions inside normal working hours against those outside them. Ninety day mortality came out at 44.2 percent for out of hours admissions against 41.0 percent for in hours admissions, and after adjustment the odds ratio reached 1.07 for out of hours admission and 1.10 for weekend admission. Those are small differences and they are real, and the explanation is almost certainly not that the night staff are worse. Fewer senior doctors are physically present, fewer diagnostic services run, a scan or an interventional procedure may wait until morning, and the patients arriving at four in the morning are sicker to begin with.
Do not worry over the clock. The practical use of that finding is a specific question put to a specific hospital, namely whether an intensive care specialist is physically in the building overnight and at weekends or reachable from home. The answer separates units far more sharply than any equipment list does.
How many beds a country has
Critical care capacity reads like a statistic for policy makers and it reaches individual patients directly, through the finding in the first section. A survey of critical care capacity across Europe found an average of 11.5 critical care beds per 100,000 people, with the range running from 29.2 in Germany down to 4.2 in Portugal. A sevenfold difference separates neighbors on the same continent with comparable medicine. Put it beside the bed strain finding and the consequence is direct. In a system where units run close to full, the deteriorating patient on the ward waits longer, and waiting past four hours was worth 16.2 percent of ninety day survival in the study that measured it.
For somebody weighing treatment in another country, this is the question behind the question. Not how modern the unit looks in the photographs, but whether the hospital has enough critical care capacity that a complication after your operation gets a bed the same afternoon instead of the following morning. Ask how many intensive care beds the hospital runs and how often it operates at full occupancy. A hospital that has never been asked will still have the number.
Two completely different situations produce the same sentence, and families deserve to know which one they are in.
When a bed is not offered, a clinical judgment
The first is a decision that intensive care would not help. Somebody whose organs are failing because an untreatable disease has reached its end does not benefit from a ventilator, and admitting them converts the last days into a procedure. Saying so is difficult and it is the right medicine, and it should come with an explanation of what will be done instead, meaning symptom control, a quiet room, and people present. Nobody is being abandoned in that conversation. A different plan is being made.
When a bed is not offered, a shortage
The second is a bed that does not exist tonight. That is not a clinical judgment and it should never be presented as one. The bed strain study showed exactly how this works, with prompt admission falling from 22 percent to 9 percent as occupancy rose, so scarcity silently changes who gets in, and a family has the right to ask which of the two they are hearing, and the truthful answer to the second is that the hospital is trying to find a bed elsewhere.
Ask the question in those words. The wording forces clarity.
The booked bed after an operation
Surgical packages frequently include a planned night in intensive care, and patients read that as a safety feature they are buying. The evidence complicates it. A nationwide study of 2,011,265 patients across 1,524 hospitals compared planned intensive care admission on the day of surgery against going straight back to a ward, for twelve major elective non cardiovascular operations. No difference in hospital mortality appeared for any of the twelve. The one place a benefit did show up was cardiac surgery, where coronary artery bypass grafting carried a risk difference of 1.0 percent in favor of the planned admission. So the booked bed earns its place in heart surgery and in patients whose own condition justifies it, and for a fit patient having a routine major operation it buys closer monitoring while leaving survival where it was. That is a defensible thing to buy and it is not the thing most patients believe they are buying, which is a lower chance of dying.
What to do with that
Two things. Stop treating a planned intensive care night as proof that a package is safer than one without it, since it is frequently a difference in how the hospital organizes recovery. And ask the more useful question instead, which is what happens if you need an unplanned bed on day three, because that is the admission the evidence shows really matters and it is the one nobody books in advance.
How to tell a good unit from an average one
Every answer below is a fact the hospital already knows, and every one of them separates units more reliably than a photograph of a monitor.
Six questions, six short answers, and a hospital used to international patients will have heard all of them before.
What we will not claim
Pages about intensive care promise things. This one will not, and the reasons are specific.
Where a hospital has handed you a survival figure with no range and no case mix attached, ask which patients produced it and watch what happens next.
After intensive care
Leaving intensive care alive is the outcome hospitals count. It is not the outcome the patient lives with. A study followed 7,656 intensive care patients and matched the 5,215 survivors against hospital control patients who had been ill enough to need admission and never needed critical care, and five year mortality in the intensive care survivors reached 32.3 percent against 22.7 percent in the controls, a hazard ratio of 1.33. Their hospital costs over those five years ran 51 percent higher. The gap does not mean the intensive care harmed them. It means that surviving critical illness leaves a mark that outlasts the discharge letter, in the form of weakness, readmissions and the underlying illness that put them there.
What follows from it is practical. A patient leaving intensive care needs a rehabilitation plan and not just a discharge date, a written list of what was damaged and what is expected to recover, and a named doctor who will see them within weeks rather than months. For a patient flying home to another country, that document is the entire handover, and a team that has written it before knows what belongs in it.
Going home to another country
Discharge from intensive care is a handover, and for a patient flying home it is the only handover their own doctor will ever get. Five things belong in the document that leaves with the patient. A plain statement of what happened and why, written for a doctor who was not there. A separation of the organ damage expected to recover from the damage that is permanent, because a physician at home cannot prescribe safely without knowing which, then the microbiology results, including any resistant organism grown, since that changes what a local hospital would reach for on readmission. A rehabilitation plan with a starting point rather than a vague instruction to build up gradually. Finally a named date for the first review, with somebody responsible for making it happen. A document carrying all five travels better than any verbal handover, in any language, through any number of clinics.
The document that travels
Request all of it in English where that is what your own doctor reads.
Remote follow up then runs through the coordinator who handled the admission. That line stays open on WhatsApp once the patient has gone home, and it is what rescues a local doctor who needs one detail from the stay at short notice. Arrange that before departure. Searching for a number under pressure, in a second language, on the day something goes wrong is the situation this exists to prevent, so see a local doctor the same day for fever, for breathing that is getting harder, for new confusion, for passing much less urine, or for pain and swelling in one leg.
Intensive care when you are not from here
Distance changes little clinically and a great deal practically. International patients get let down on the practical part.
Language, at the moment it counts most
Consent conversations, daily updates and any discussion of limits of treatment are the three moments where an interpreter is the difference between a family understanding what is happening and a family guessing, and seven languages are covered directly here, namely English, Arabic, French, Russian, Serbian, Romanian and Spanish, with anything else arranged on request. One coordinator is assigned from the first message and stays with the case through discharge, which matters more in intensive care than anywhere else, because the same family hears the same update every day and continuity of the person delivering it is what makes the sequence make sense.
Staying close by
Patient rooms carry a companion bed, so once the patient has left intensive care one person sleeps in the room for the rest of the admission. While they are still in the unit, the international patients office finds the family somewhere to stay within reach of the ward, moves them between it and the hospital, and collects them from the airport when they land. Meals are cooked halal, vegetarian or diabetic as the ward asks, and a prayer room stays open in the building. Where a female physician is wanted, the request goes to the department and is met wherever the rota allows, so say it in the first message instead of on the first morning. Radiotherapy and chemotherapy run on this site alongside surgery, which counts for a great deal when a complication changes the plan, because nothing is referred out and no records have to travel. Each unit sets its own visiting pattern around the medical round and the nursing handover, and somebody explains it on the first day, which is the conversation to have before the first visit rather than at the door.
What drives the cost
No figure appears on this page, for an arithmetic reason. Billing runs by the day and by what gets used, so any number quoted before anybody knows the length of stay describes nothing at all. Six things move the total. The number of nights, which dwarfs everything else and which nobody can forecast on day one. Ventilation next, since a ventilated bed costs more than a monitored one in every health system on earth, and then kidney support, because dialysis is priced separately almost everywhere, and then which antibiotics and antifungals are used, as the drugs reached for against a resistant organism sit at the expensive end. How much blood and how many platelets are transfused. And whether a procedure becomes necessary partway through, whether that means a tracheostomy, a drain placed in the chest or the abdomen, a radiology guided drainage or a second trip to the operating room. Length of stay dwarfs the other five put together.
This table scrolls sideways on a phone. Drag it across to reach the second column.
| Ask this | Why the answer matters |
|---|---|
| What does a daily intensive care rate cover, and what is billed on top of it | This single question separates a real quotation from a headline one faster than any other |
| Are ventilation, dialysis and blood products inside the daily rate or itemized | These are the three items most often assumed to be included, and the three most often not |
| Does the rate drop when the patient steps down to a ward | A level 2 bed costs less than a level 3 bed, and some quotations carry one rate throughout |
| What happens financially if the stay runs past the estimate | Get this answer while it is theoretical, since nobody negotiates well on day nine |
| Is a planned postoperative night already inside the surgical quotation | It usually is, and an unplanned admission later usually is not |
Published packages from hospitals in this market are consistent on that last point, and confirm it in writing for your own case. A figure that means anything comes only after a doctor has read the current reports, and that review costs nothing.
What we arrange
An intensive care admission strands a family in an unfamiliar city with nothing planned. The practical arrangements are part of the care.
Learn the name of that coordinator before anything is agreed. A hospital that cannot produce one has already told you how the next few weeks will run.
What the daily update should contain
Families are handed numbers and reassurance, and those are the two least useful things available. A useful update answers four questions in order.
- What is the main problem today, named in one sentence, and is it the same problem as yesterday.
- Which organs are being supported, and is that list longer or shorter than yesterday.
- What is the plan for the next 24 hours, meaning the specific thing the team intends to try or stop.
- What would change the plan, in either direction, and what the team is watching for.
Write the answers down. Over a two week admission the same family speaks to several doctors, sometimes in two languages, and a notebook turns four separate conversations into one thread.
A fifth question belongs in that conversation and rarely appears. Put it plainly. What does the team expect the patient to be able to do in three months, because the answer separates a doctor thinking of survival from one thinking of the person.
Questions we are asked, an intensive care unit FAQ
What is the difference between intensive care and high dependency
The nursing ratio and the level of organ support. High dependency, sometimes called level 2 or step-down, gives one nurse to two patients, invasive monitoring and support for a single failing organ. Level 3 intensive care gives one nurse to one patient, and supports advanced breathing or two or more organs at once. A patient described as needing level 2 does not need a ventilator, which is usually the reassurance a family is actually looking for and rarely the sentence they are given. The move in the other direction confuses families just as often. Transfer from level 3 down to level 2 reads as abandonment because the nurse who was in the room all night is now covering the bed next door as well, and it is in fact the first piece of good news in the admission. Ask which level your relative is on, and ask again whenever the ward changes, because the number carries more information than any description of how they are doing.
Does it matter what time of day somebody is admitted
A little, and less than the delay does. In a Danish study of 44,797 patients, ninety day mortality was 44.2 percent for out of hours admissions against 41.0 percent in hours, with an adjusted odds ratio of 1.07 and 1.10 for weekends. Fewer senior staff are present and fewer services run at night. The more useful question is whether an intensive care specialist is physically in the building overnight at the hospital you are considering.
How many nurses should there be
One per patient in level 3 and one per two patients in level 2. A national study covering every adult intensive care admission in Korean general and tertiary hospitals over five years found in hospital death rose as beds per nurse increased, with an odds ratio of 1.17.
Is a planned night in intensive care after surgery a sign of a safer hospital
Not on its own. A study of 2,011,265 patients across 1,524 hospitals found no hospital mortality difference from planned intensive care admission after twelve major elective non cardiovascular operations, and the benefit appeared only in cardiac surgery, where bypass grafting carried a risk difference of 1.0 percent. Put the more useful question as well. What happens if an unplanned bed is needed on day three.
How long do people stay
No average means anything here, because a planned overnight stay after major surgery is one night while respiratory failure needing a ventilator runs into a week or more. Length of stay is also the single largest driver of what an admission costs. Plan around the next review date.
Can we stay with our relative
Intensive care units set their own visiting pattern, built around the medical round and the nursing handover, and somebody explains it on the first day, and once the patient steps down to a ward the companion bed in the room means one person stays overnight for the rest of the admission. Accommodation for the rest of the family is arranged by the international patients office.
What should we ask every day
Four things. What the main problem is today and whether it has changed. Which organs are being supported and whether that list is growing or shrinking. What the plan is for the next 24 hours. And what would change it in either direction. Write the answers down, because over a long admission you will speak to several doctors and a notebook turns those conversations into one thread.
References
- Harris S, Singer M, Sanderson C, et al. Impact on mortality of prompt admission to critical care for deteriorating ward patients, an instrumental variable analysis using critical care bed strain. Intensive Care Med. 2018;44(5):606-615.
- Han KT, Kim S. Impact of nursing policy on securing intensive care unit nurse staffing and in-hospital mortality. Intensive Crit Care Nurs. 2024;86:103807.
- Buck DL, Christiansen CF, Christensen S, et al. Out-of-hours intensive care unit admission and 90-day mortality, a Danish nationwide cohort study. Acta Anaesthesiol Scand. 2018;62(7):974-982.
- Rhodes A, Ferdinande P, Flaatten H, et al. The variability of critical care bed numbers in Europe. Intensive Care Med. 2012;38(10):1647-1653.
- Ohbe H, Matsui H, Kumazawa R, Yasunaga H. Postoperative ICU admission following major elective surgery, a nationwide inpatient database study. Eur J Anaesthesiol. 2022;39(5):436-444.
- Lee S, Kim G. Rapid response systems in hospital settings, an umbrella review. Intensive Crit Care Nurs. 2026;95:104463.
- Lone NI, Gillies MA, Haddow C, et al. Five-year mortality and hospital costs associated with surviving intensive care. Am J Respir Crit Care Med. 2016;194(2):198-208.
Editor's note
Written by the Biruni Hospital medical editorial team. Reviewed by Assoc. Prof. Dr. Suna KOÇ, Anesthesia and Reanimation.
Medically reviewed by
Assoc. Prof. Dr. Suna KOÇ
Anesthesia and Reanimation
Related Treatments
View All
Bone Marrow Transplant Unit (BMT Unit)
Between the mid 1990s and the mid 2000s, death within 200 days of an allogeneic transplant that was not caused by relapse fell by 60 percent, and almost none of that came from a new drug. This page is about the part nobody photographs, meaning the filtered room, the rules that have evidence behind them and the ones that do not, the laboratory chain behind the cells, and what a family can actually do.
Burn ICU (Burn Intensive Care Unit)
Age, burn size and smoke inhalation are the three numbers that set the outlook after a major burn. This page walks through fluids, early surgery and grafting, the long healing phase, and when a burn patient can be flown abroad.
Burn Unit (Burn Center)
Burn care runs on one figure, the percentage of the body surface burned, and a systematic review of 28 studies found that referring hospitals overestimate it in half or more of cases. Fluid volume, the transfer decision and the survival estimate are all calculated from that number, which is why this page starts with how it is measured and who should be measuring it.
Cancer Pain Management Unit
The Cancer Pain Management Unit at Biruni Hospital in Istanbul treats pain caused by cancer and by its treatment. Two in three people with advanced cancer report pain, and reviews show nearly one in three receive medication too weak for it. The unit uses stepwise medicines, radiotherapy for bone pain, nerve blocks and spinal pumps, and plans follow-up for patients who return home abroad. The pain team reviews reports and current medicines free of charge.
Cancer Patient ICU Care
Cancer once counted as a reason not to offer an intensive care bed at all. Pooled data from 7,354 critically ill cancer patients retired that rule, and the argument now is about which patient, for what, and how long full support runs before anybody formally reassesses it.