Pronouncing a patient dead in a hospital seems relatively simple: palpate for lack of pulse, determine that the patient's neurological function is absent, then disclose, out loud, his or her full name and the time of death. Except it isn’t that simple at all—in the inpatient setting.
The difference between dead and alive can be straightforward to delineate in the right context. There are professions for which this skill is absolutely paramount. Think of soldiers in the middle of a battlefield, war photojournalists or field-trauma EMTs at the site of a multiple-vehicle crash: each has to make quick, decisive pronouncements about a person’s viability to do his or her job effectively, and the verdict is based on a mix of available objective data and personal experience. The inpatient hospital setting is a much more controlled environment, so you’d expect the protocols for death pronouncement to be that much more standardized. But they are hardly uniform.
Here’s my problem with this situation: when I pronounce a patient dead, neither I nor that person’s family (nor, frankly, the patient himself or herself) should be faced with uncertainties. I should know, for instance, what the definition of death is. I should also know how to medically determine when the patient is no longer alive. I’ll even settle for being certain that a person passed away at the actual time I pronounce it, but physicians and their patients aren’t even afforded that luxury.
Let’s unpack my concerns one by one. First, what is the hospital’s definition of death? You might be surprised to know that the answer depends on whether you were asking before or after 1968. Before that year, death in most hospitals was defined as the cessation of all vital functions, including respiration and a heartbeat, sometimes called the “heart and lung” definition. As the technological ability to keep a patient’s hearts and lungs going advanced, determining who was alive became much more medically murky.
It wasn’t until 1968, when an ad hoc committee at Harvard Medical School developed the so-called Harvard criteria, that “brain death” was introduced into commonly used hospital definitions of death. In an attempt to further unify the inpatient death definition in 1981, the newly formed President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research (PCSEPMBBR, as it’s referred to by no one on this planet) recommended that all states adopt the Uniform Determination of Death Act (UDDA).
The UDDA asserted that “an individual who has sustained either (1) irreversible cessation of circulatory and respiratory functions, or (2) irreversible cessation of all functions of the entire brain, including the brain stem, is dead. A determination of death must be made in accordance with accepted medical standards.” The act aspired to add this new “total brain death” statute, in addition to the boilerplate heart-and-lung definition, to all 50 states’ laws, with mixed success. I’d give it a B-minus for effort.
Most states adopted some form of the UDDA, but the differences among state laws are startling. North Carolina, for instance, doesn’t have a heart-and-lung provision. Louisiana and Texas completely eschew the total-brain-death clause from their hospital definitions of death. And the phrase “in accordance with accepted medical standards” is absent from Georgia law and appears in equally vague forms in Minnesota (“generally accepted medical standards”), Maryland (“ordinary standards of medical practice”) and Florida (“in accordance with currently accepted medical standards”). I defy you to find 10 doctors who can agree on what constitutes “generally accepted” or “ordinary” standards of medical practice, much less decree whether “cessation of all functions of the entire brain, including the brain stem” is absolutely necessary to define death.
It’s not impossible to get hospitals, states and even countries to agree on protocols or medical definitions. Think about systemic inflammatory response syndrome (SIRS) and its cousin sepsis. SIRS is a clinically measurable, systemic state of inflammation that can affect multiple organ systems in the body. And when that life-threatening organ dysfunction is caused by a dysregulated response to infection, it is called sepsis. SIRS is defined as having at least two of the following four criteria: a white blood cell count less than 4,000 per cubic millimeter or greater than 12,000/mm3; a temperature below 36 degrees Celsius or above 38 degrees C; a heart rate greater than 90 beats per minute; and a respiratory rate greater than 20 respirations per minute.
This definition for SIRS is not only agreed on by every hospital in the U.S. but most Western European countries and a number of Asian countries—in 2016 there was even a task force that published “The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3).” We’ve done it for SIRS and sepsis. Why can’t we come up with a universal inpatient definition for death?
There isn’t one answer, of course. I suspect the situation is linked to a difference in outcomes. When a patient has SIRS or sepsis, the result is significantly increased morbidity and mortality and not much else. The consequences and implications of a universal definition of death are much wider-reaching—ethically, legally, financially, politically and medically. That doesn’t, however, make it any less urgent.
If the universal inpatient definition of death is too elusive, the process of pronouncing someone dead should be much more tenable. Most hospitals’ procedures follow roughly the same outline. When I was in medical residency training at a hospital in north–central Massachusetts, our inpatient facility had specific, deliberate steps for the pronouncement of death along these lines:
1. Before entering a patient’s room, discuss the following with the covering nurse (away from the patient and family):
a. When did cardiac activity cease?
b. When did the patient stop breathing?
c. Is the patient an organ donor?
d. Is an autopsy required?
2. Enter the room.
3. Verbally identify yourself.
4. Verbally identify the patient.
5. Ensure the patient isn’t hypothermic.
6. Note the general appearance of the patient and any spontaneous movement.
7. Note a lack of reaction to verbal or tactile stimulation (check tactile stimulation with a strong sternal rub or by pinching the patient in the medial brachial area, immediately distal to the axilla).
8. Note no pupillary light reflex (pupils should be fixed and dilated).
9. Ensure no breath or lung sounds, for 30 to 60 seconds. Waiting a whole minute for lung sounds may seem excessive until you’ve seen a living patient with Cheyne-Stokes respirations take 90 seconds in between breaths. Fun party trick.
10. Ensure no heart beat or pulse for 30 to 60 seconds.
11. Pronounce the patient dead out loud and identify the time.
Not all hospitals agree with my former one. The Stanford University School of Medicine’s palliative care department believes that checking for a pupillary light reflex and performing sternal rubs are completely unnecessary and can cause needless trauma to the family members bearing witness. In Stanford’s eyes, according to its Web site, the pronouncement of death is more a “solemn ritual, the importance of which transcends the business of certification,” and it “may formally give permission for loved ones to grieve.” The Department of Veteran Affairs’ Birmingham/Atlanta Geriatric Research Education and Clinical Center in Alabama informs physicians they should spend much more time determining the individual relationships of every family member to the patient they are preparing to pronounce and should confer very little time on the pronouncement itself.
Contrast those approaches with the original protocol developed by the Harvard committee, which, as described in a 2017 paper by neurologist Ariane Lewis and her colleagues, included ensuring “(1) unreceptiveness or unresponsiveness to any external stimulation; (2) absence of movement or breathing (defined as absence of movement in response to pain, touch, sound, or light over the course of one hour and total absence of spontaneous breathing after discontinuation of the ventilator for three minutes); and (3) absence of reflexes (fixed and dilated pupils, no blinking or movement of the eyes to head turning or irrigation of the ears with ice water, no posturing, no corneal or pharyngeal reflexes, no swallowing or yawning or vocalization, no muscle contraction in response to tapping of tendons, no plantar response).” Perhaps the most stringent criterion of that protocol mandated the measurement of a flat, isoelectric reading on an electroencephalogram (EEG) to ensure complete brain death, both at that time and 24 hours later. I’ve pronounced eight people dead in my medical career thus far; I can’t recall ever having seen an EEG being used in a death pronouncement.
The pronouncement process can run the gamut from grotesquely technical to completely ceremonial. I don’t think there are any wrong answers for how this should be done. I merely find it fascinating that there is so much stylistic variation in a procedure that most people assume is rigidly systematic.
My most philosophical bone to pick with the death pronouncement procedure is, admittedly, a personal one that may not bother other pronouncers in the slightest. I have a problem with the clerical concept of “time of death.”
When physicians pronounce a patient dead, those words are legal language that create legal consequences. Think of the phrases “I now pronounce you husband and wife” or “You are under arrest” and how they transform the lives of the subjects of those pronouncements. When the doctor declares someone deceased at a specific time, only then are they legally dead—even if, in reality, they might have died 20 minutes prior.
That gap in time between when a patient actually passed away and the time recorded on a death certificate has always haunted me. I understand the administrative commitments of a hospital and don’t begrudge any health care workers for how or when they choose to accomplish this somber task. But I’ve consistently found it strange that a death pronouncement can be delayed by how many other patients the on-call physician happens to be juggling that night, by whether the electronic health-record system requires documentation before the pronouncement or simply if the doctor decided to take the elevator versus the stairs to the patient’s room.
These are the absurdly mundane realities of a pronouncement that go unsaid and remain tacitly understood by health care providers alone. Mundane enough for me to see my grandmother’s time of death listed as “July 3, 2015, 2150 hours” and wonder just how busy the attending physician was that evening.
So let’s imagine you’re a young resident physician covering the inpatient floors one Wednesday evening. You’ve been paged to the medical-surgical unit to pronounce a patient dead. You’ve been informed that the family is already in the room, waiting for you. You’ve got a lot to consider before you proceed. Is it before or after the year 1968? Are you in the state of North Carolina? Is an EEG machine available?
And are you thinking about taking the elevator or the stairs?"