Most people die within hours of being taken off life support, though the range stretches from minutes to, in rare cases, weeks or even months. In studies of mechanical ventilation withdrawal in intensive care units, the median time to death falls just under one hour, but roughly a quarter of patients survive past the first day. The wide spread reflects differences in the type of support being withdrawn, the patient’s underlying condition, and how many organs have already begun to fail. Families preparing for this moment deserve honest expectations, and the science behind those expectations is more detailed than most people realize.
The Typical Timeline After Ventilator Withdrawal
When doctors withdraw mechanical ventilation from a critically ill patient, death usually follows quickly. A large study of ICU patients found the median time to death was about 56 minutes, with half of all deaths occurring between roughly 15 minutes and five and a half hours after the ventilator was removed.1PubMed Central. Predictors of time to death after terminal withdrawal of mechanical ventilation in the ICU A separate study reported a nearly identical median of 0.9 hours, though its full range was far wider, stretching from immediate death to 165 hours in some patients.2PubMed Central. Factors Associated with Palliative Withdrawal of Mechanical Ventilation and Time to Death after Withdrawal An Australian study of 80 patients found that about 45 percent died within the first 60 minutes, while the mean time to death was closer to six and a half hours because some patients survived substantially longer, with one living 31 days.3Critical Care and Resuscitation. Prediction of death after withdrawal of life-sustaining treatments
A study that included patients during the COVID-19 pandemic era found that about 54 percent of patients died within one hour of terminal extubation, while 25 percent survived beyond 24 hours.4PubMed Central. Prediction of survival time after terminal extubation: the balance between critical care unit utilization and hospice medicine in the COVID-19 pandemic era That one-in-four figure often surprises families who expect the process to be over within minutes. In reality, the hour-by-hour experience varies enormously depending on the patient’s physiology at the moment of withdrawal.
What Determines Whether It Takes Minutes or Days
Several measurable factors help clinicians estimate how quickly death will follow. A systematic review that pulled together findings from multiple studies identified five consistent predictors of a shorter time to death: whether the patient was on controlled (rather than assisted) ventilation, how poorly their blood was being oxygenated, whether they required drugs to maintain blood pressure, their level of consciousness, and whether basic brain-stem reflexes were still present.5PubMed. Predicting time to death after withdrawal of life-sustaining therapy In plain terms, a patient who cannot breathe at all without the machine, whose blood pressure is propped up by medications, and who shows no neurological responsiveness will generally die very quickly once support is removed.
Research on the specifics bears this out. Patients receiving vasopressors (drugs that raise blood pressure) had a substantially higher chance of dying sooner, while older patients and women tended to survive slightly longer on average after withdrawal.1PubMed Central. Predictors of time to death after terminal withdrawal of mechanical ventilation in the ICU The number of failing organs also mattered: each additional organ system in failure shortened the expected survival time. Conversely, patients who had recently been able to trigger some breaths on their own, who needed less support from the ventilator, and whose oxygen levels were adequate were more likely to survive past the first hour and sometimes past the first day.4PubMed Central. Prediction of survival time after terminal extubation: the balance between critical care unit utilization and hospice medicine in the COVID-19 pandemic era
None of these factors provides a precise countdown. They describe tendencies across large groups. For an individual patient, the ICU team can usually say whether death is likely within the first hour or whether it could take much longer, but they cannot give an exact time.
When Feeding Tubes and Fluids Are Withdrawn Instead
Not every life-support withdrawal involves a ventilator. For patients in prolonged states of severely reduced consciousness, the decision may involve stopping artificial nutrition and hydration through a feeding tube. This is a fundamentally different timeline. Without food and water, the body takes considerably longer to shut down than it does without oxygen.
Interviews with families of patients who died after feeding-tube withdrawal reported that most deaths occurred between 9 and 14 days afterward.6PubMed Central. Deaths after feeding-tube withdrawal from patients in vegetative and minimally conscious states: A qualitative study of family experience Some families described being distressed by the gradual physical changes their loved one went through over those days. The process is slower and, for many families, harder to witness than the relatively rapid death that follows ventilator removal. Palliative care teams typically manage symptoms such as dry mouth and restlessness with comfort medications during this period, but the extended timeline is something families often feel underprepared for.
Children and Infants
The timeline tends to be shorter for pediatric patients. A scoping review of deaths after withdrawal of life-sustaining therapies in children found that most occur within one hour.7PubMed Central. Pediatric Death After Withdrawal of Life-Sustaining Therapies: A Scoping Review A more detailed analysis of over 500 pediatric cases put finer numbers on this: roughly 72 percent of children died within 30 minutes of withdrawal, and about 87 percent within 60 minutes.8PubMed. Development of a bedside tool to predict time to death after withdrawal of life-sustaining therapies in infants and children Factors that predicted faster death included being a newborn (one month old or younger), receiving high doses of blood-pressure-supporting drugs, being on a heart-lung bypass machine, and requiring high ventilator pressures.
The higher proportion of rapid deaths in children likely reflects the conditions that lead to withdrawal decisions in pediatric intensive care: severe brain injuries, congenital anomalies, and overwhelming infections in very small bodies with less physiological reserve. For parents, the practical takeaway is that the window of time at the bedside after withdrawal is usually brief, and healthcare teams increasingly try to prepare families for that reality in advance.
Outliers Who Survive Much Longer
While most deaths cluster in the first few hours, outliers can survive far beyond what anyone anticipated. A study in Taiwan documented the full range vividly: all patients with cancer as their underlying condition died within one week of ventilator withdrawal, but among patients with non-cancer diseases, about 7 percent lived longer than one week. The most extreme cases included one patient who died at day 55 and another at day 104 after the ventilator was removed.9Journal of the Formosan Medical Association. Clinical characteristics and survival outcomes of terminally ill patients undergoing withdrawal of mechanical ventilation
These outliers are almost always patients whose lungs retained some ability to function on their own but who were on ventilators for other reasons, such as a devastating neurological injury in someone with otherwise healthy lungs. The person continues breathing but remains profoundly disabled. For families, this possibility raises difficult logistical and emotional questions about where the patient will be cared for during what could turn into weeks of continued survival. Many hospitals will transfer such patients to a hospice or palliative care facility, since ongoing ICU care is no longer the goal.
How Prediction Models Try to Narrow the Uncertainty
The wide range of possible outcomes has prompted researchers to develop scoring tools that estimate how quickly a patient will die after withdrawal. These models typically combine factors like the level of ventilator support, blood pressure medication doses, and neurological status into a single score. In their original development, several of these models achieved classification accuracy between 75 and 91 percent for predicting whether death would occur within a set window.10PubMed Central. Predicting Time to Death After Withdrawal of Life-Sustaining Treatment in Children
When tested on new patient populations, however, performance is uneven. External validation of several prominent models found that the best performer correctly identified patients who would die within 60 minutes with good accuracy, but other models did barely better than chance.11American Journal of Transplantation. External validation of prediction models for time to death in potential donors after circulatory death A pediatric-specific bedside tool showed moderate accuracy for predicting death within 30 minutes (about 66 percent) and better accuracy for predicting death within 60 minutes (about 76 percent).12PubMed Central. Validation of a pediatric bedside tool to predict time to death after withdrawal of life support Newer approaches, including deep-learning algorithms, are being developed in an effort to improve these predictions.13PubMed Central. Deep learning unlocks the true potential of organ donation after circulatory death with accurate prediction of time-to-death
In practice, these tools are used more for planning than for giving families a precise time. They help clinicians decide how to stage the withdrawal, where to position the patient, and whether organ donation might be feasible. They are not reliable enough to promise a family that their loved one will die within a specific hour.
Why Organ Donation Depends on Timing
One of the less-discussed reasons clinicians care about predicting time to death is organ donation. In donation after circulatory death, organs can only be recovered if the heart stops within a certain window after withdrawal. Once blood flow to an organ ceases, warm ischemia, the damage from oxygen deprivation at body temperature, begins accumulating rapidly. The classical time limits are about 30 minutes of warm ischemia for the liver and pancreas, 60 minutes for the lungs, and 120 minutes for the kidneys.14PubMed Central. Potential for organ donation after controlled circulatory death: a retrospective analysis
If a potential donor does not die quickly enough after withdrawal, the organs become unusable and the donation cannot proceed. This is one reason the uncertainty around timing is so consequential. A patient whose family has agreed to organ donation may be taken to an operating room or a nearby suite before withdrawal so that surgical teams can recover organs quickly once death occurs. If the patient survives beyond the ischemia window, the donation is called off, and the patient is moved to a comfort-care setting. The experience can be jarring for families who had expected both death and donation to happen seamlessly. Better prediction models could reduce the number of times families and medical teams go through this process only for donation to fail.
What Pain Medication Does and Does Not Do
A common concern among families is whether the medications given during withdrawal are actually hastening death. Opioids like morphine are routinely administered during the process to manage visible distress, labored breathing, and any pain the patient might experience. Research on this question has produced a counterintuitive finding: higher morphine doses during the final hour of life were actually associated with a slight delay in death, not acceleration. Each one-milligram-per-hour increase in morphine infusion during the last hour was linked to death arriving about eight minutes later.15PubMed. The infusion of opioids during terminal withdrawal of mechanical ventilation in the medical intensive care unit
This finding makes physiological sense. Morphine can reduce the sensation of air hunger and suppress the stress response, which may actually allow the body to hold on slightly longer rather than being overwhelmed by the distress of suffocation. For families worried that comfort medications are a form of euthanasia, this evidence is reassuring: the drugs are treating suffering, not causing death. The patient’s underlying condition is what proves fatal once the machine stops doing the work their body cannot.
The Lazarus Phenomenon
Rarely, a patient’s heart spontaneously restarts after it has apparently stopped. This event, sometimes called the Lazarus phenomenon or autoresuscitation, is documented but uncommon. A prospective study of out-of-hospital cardiac arrests in which resuscitation was terminated in the field found that delayed return of heart activity occurred about 6 times per 1,000 cases, with the heart restarting between 3 and 8 minutes after efforts had ceased.16PubMed. Delayed return of spontaneous circulation (the Lazarus phenomenon) after cessation of out-of-hospital cardiopulmonary resuscitation
A broader review spanning over three decades of published cases identified 65 patients worldwide in whom the heart restarted after resuscitation was stopped. Of those, about 28 percent eventually made a full recovery.17PubMed Central. Autoresuscitation (Lazarus phenomenon) after termination of cardiopulmonary resuscitation – a scoping review The phenomenon is rare enough that it does not meaningfully change the expectations around planned withdrawal of life support in the ICU, where the clinical picture is usually far more dire than an isolated cardiac arrest. But it is one reason that most protocols for donation after circulatory death require a mandatory observation period of several minutes after the heart stops before death is formally declared. The standard wait in many countries is five minutes, though practices vary.
What Dying Patients May Still Perceive
One of the most emotionally charged questions families ask is whether their loved one can still hear them in those final hours. Research using brain-wave monitoring in actively dying hospice patients suggests the answer may be yes. Unresponsive patients who were hours from death still showed brain responses to changes in sounds that were similar to the responses seen in healthy, alert young adults. Some patients even showed higher-level brain activity associated with the conscious processing of what they were hearing.18PubMed Central. Electrophysiological evidence of preserved hearing at the end of life
The findings do not prove that dying patients understand speech or experience what they hear in the way a conscious person would. Brain-wave responses can reflect automatic processing rather than deliberate awareness. But the evidence does suggest that the auditory system keeps functioning much closer to the moment of death than most people assume. Palliative care nurses have long encouraged families to keep talking to their loved ones and to be mindful of what is said in the room, and this research supports that instinct.
How Culture and Geography Shape These Decisions
The question of how long someone lives after being taken off life support also depends on where in the world the decision is made, because practices around withdrawal vary enormously by country and culture. A large study across ICUs in Asia found that rates of withholding and withdrawing treatment differed substantially from region to region, influenced by economic conditions, cultural values, religious beliefs, legal frameworks, and the personal attitudes of physicians.19JAMA Internal Medicine. Withholding and Withdrawal of Life-Sustaining Treatments in Intensive Care Units in Asia In some settings, withdrawal of mechanical ventilation is legally or culturally prohibited, meaning the question of survival time after withdrawal simply does not arise. Patients instead remain on support until cardiac death occurs on its own, which can take days to weeks.
In countries where withdrawal is practiced, local norms still shape the specifics. Some institutions favor a gradual reduction in ventilator settings over hours, allowing the body to adjust slowly and giving families more time at the bedside. Others perform what is sometimes called immediate extubation, removing the breathing tube in a single step. Both approaches aim for comfort, but the pace of dying can differ. The gradual approach sometimes extends the time to death, while immediate extubation tends to produce a faster trajectory, especially in patients who were fully dependent on the ventilator. Families who are navigating these decisions in real time benefit from asking the care team directly which method will be used and what the expected timeline looks like in their specific situation.