The Slow, Troubling Death of the Autopsy
Why you should get an autopsy if it’s the last thing you do
Disclaimer: The images in this story may be disturbing for some people.
For most people, the mere word “autopsy” summons up visceral images. Bodies lying cold and blue in a lab, a Y-shaped incision on their naked chest. Sardonic doctors hovering over them with medieval tools. Grim viridian-tinted light, banks of glowing screens. Bullets clinking into steel pans. There is tenebrous music, probably.
The first thing to know about how a real autopsy lab works is that everything TV taught you is wrong. For one thing, there is no blue lighting anywhere — this is the dubious logic of CSI, in which autopsies are conducted in atmospherically dim rooms. In reality, the lighting is dazzlingly white and stark. The floors are brushed cement, the walls are white, and the tables made of stainless steel.
There are also no glowing screens, no projected holograms, no computers that can instantly spit out a list of every foreign substance in a person’s system. Instead, there are basic tools: scalpels, aging Dell computers, endless cotton towels, long-handled pruning shears that retail for $39.99 at Home Depot.
I had no idea what to expect when I was allowed to spend two weeks observing clinical autopsies at the University of Pittsburgh Medical Center¹. I’d been warned that fainting was common, so as a physical coward who has not watched a horror movie since 1996, I brought a box of chewable antacids, as if that would help. When I walked into the room, swathed in blue PPE from head to toe, two naked men lay supine on two stainless-steel dissection tables. They looked like people rising suddenly from nightmares, with their chests arched and their heads thrown back and mouths agape. Except that they had been cut open from collarbone to pelvis.
This was the moment where fainting tends to happen. I reached for the antacids — and then I found I didn’t need them. Somehow, I understood. It made sense. These men were dead, and this is what we look like when we are dead.
By the end of the second day, I wasn’t just observing anymore, I was standing alongside a body, gripping its skin to hold it open as a pathology resident ducked in to carefully free the liver. By the end, I was covered in blood and fluid as I entered data, prodded heart ventricles, and weighed organs. Correctly, I was prohibited from handling sharp objects.
Through a microscope, I peered at a sliver of lung taken from a young person who died of a drug overdose. At first the lung just seemed like a solid mass of tissue, which is very bad — healthy lungs have a lot of small, open alveolar spaces, because healthy lungs are mostly air. Then the chief pathologist, Jeffrey Nine, MD (who is now the medical examiner for Yavapai County in Arizona), flipped the polarity of the microscope. Suddenly the tissue turned night-dark, and a constellation of tiny silver shards glowed.
“That,” he said, “is why you never leave an intravenous drug user alone in a hospital room with an IV drip.” The shards, he explained, were grains of pills that the person had ground up, stirred with some water, and shot into their veins. But pills don’t dissolve in water — not enough to inject into a blood vessel.
I was looking straight into a cause of death. Those little glowing shards stopped a heart and erased a life. It was, coldly, beautiful.
How to look inside a person
When you die, your physician has to fill out the U.S. Standard Certificate of Death. Box 37 will ask them to describe the “manner of death.” There are six options:
- Pending investigation
- Could not be determined
Of these six, only the first, natural death—that is, from disease or old age—will send your body to the kind of lab I was in. Anything potentially caused by external, unnatural forces, like being found dead at home, getting struck by a falling air conditioner, or overdosing on drugs², delivers your body to the medical examiner for a forensic autopsy, a highly specialized form of autopsy³. In these cases, a trained forensic pathologist seeks to establish the cause of and circumstances surrounding the death — and often testify to them in court⁴. If you’re imagining bullets clinking and estimated times of death and Jerry Orbach or Mark Harmon serving up jaded one-liners, you’re thinking of the forensic autopsy (which is having a crisis of its own).
“Natural deaths,” however, are given a clinical (or academic) autopsy, which is the kind I was observing. This version exists to investigate disease and to educate doctors. (Medical examiners, it should be said, also sometimes teach.) It takes place in a hospital or private lab, and here, the pathologist is after the exact sequence that led to death, from underlying causes (say, metastatic cancer) to immediate cause (pneumonia). Sometimes there’s no mystery: “Decapitation is easy,” explains Sanjay Mukhopadhyay, MD, director of pulmonary pathology at the Cleveland Clinic. But, he adds, the majority of autopsies reveal how much of a puzzle death can be, even to experienced doctors. “You have a dozen abnormalities, and you have to discover what killed them.”⁵
To begin, the doctor (or their lab tech or pathologist’s assistant) traces a Y-shaped incision down the torso with a scalpel. The bowel is unwound and removed first to prevent bacteria from infiltrating the rest of the chest cavity, and then the ribcage is removed (enter the pruning shears). Once the chest cavity is open, the rest of the organs come out en bloc, from the tongue to the pelvic organs. They are then transferred to a table with running jets of water, where they are washed and carefully dissected on white plastic cutting boards. Tiny samples are taken from each organ and placed in small plastic cassettes in a formalin solution. These cassettes are sent to a lab, where they are fixed in wax and slivered to micron thinness, then placed on a slide, stained with hematoxin and eosin. Under a microscope, they appear as an oddly beautiful pointillist whorl of pink and blue. In some cases, the brain and eyes are removed as well.
Everything is photographed, weighed, felt, and analyzed. The autopsist’s tools are simple, because most of their work is done with the eyes and hands. There is no diagnostic replacement for touch — or smell. Typically, the preliminary results, including a detailed cause of death and description of visible pathologies, are sent to families and doctors within a few days, and the full report (which includes the results of any testing — especially genetic) is sent to the doctor and family after about 30 days. Pathologists also go over the report in detail with the family and answer any questions they may have.
A common fear with autopsies is mutilation: the idea that one cannot have an open-casket funeral afterward. This is false. Dissectors take care to preserve the carotid arteries, which allows morticians to flush the face with formalin, restoring its shape. In the case of the eyes, the vitreous (the gel-like substance that makes up most of the eyeball’s mass) collapses postmortem anyhow, so all morticians will place a hard cap over the socket regardless of whether the eyes are still in place. The face, hands, and extremities are left untouched, unless there is a specific reason to dissect them. Not a single notch of this work is visible in an open-casket funeral.
The long, slow death of the autopsy
In the late 1960s, the autopsy rate in U.S. hospitals was nearly 60%. Today, that number is 4.3%. This decline is not limited to the United States ; in the U.K., the rate is 0.69%. It’s easy to assume this plunge has to do with improvements in medical technology — that with MRIs and PET scans and laparoscopy, there’s nothing doctors cannot see. That is flatly not true, and if you want proof, look no further than the farce that is death certification.
In the United States, our lives are framed by two documents: our birth certificate and our death certificate. Paperwork accompanies us into the world and guides us out again. And just as we need our birth certificates to get a Social Security card or driver’s license, we cannot be buried, nor can our estates be settled, until our death certificates are finalized. Yet, while errors on birth certificates are rare, death certificates are rife with them.
In 2016, Elizabeth Novick, MD, had just started her residency in internal medicine at Georgetown when a patient in the ICU died. Novick was on call. It was her first death, but she did not feel uncomfortable. She had been well trained in how to break the news to the family. “I’m so sorry,” she told them, using the words she’d chosen and practiced, “but she has passed.” Novick asked the family’s permission to perform a short exam to confirm the death and pressed her stethoscope to the patient’s heart and lungs, hearing only silence, and then shone a light into her eyes to make sure the pupils were unresponsive. At this point, Novick pronounced her patient dead and withdrew to allow the family to have their final moments with her.
In the late 1960s, the autopsy rate in U.S. hospitals was nearly 60%. Today, that number is 4.3%.
“But then a nurse literally thrust some paperwork into my hand,” she remembers, “and said, ‘You have to fill that out now.’” This was the “death packet.” Novick had never seen one before; Georgetown had never trained her for this. She felt “woefully unprepared” for both the death certificate and the autopsy request form. The woman had metastatic cancer and pneumonia, but Novick wasn’t sure what exactly had caused her to die. Uncertain, she asked for help, and a senior resident told her, “Just write arrhythmia.” Two days later, Novick got a call from the medical records department telling her she couldn’t write that, and they had to revise the cause of death. This time, it was listed as “hypoxemic respiratory failure due to pneumonia in setting of metastatic cancer.”
The suggestion Novick received was far from unusual. According to Lundberg, when adults die suddenly and unexpectedly from natural causes, the physician who pronounces them dead frequently writes “arrhythmia” or some variant, like “cardiac arrest.” “Cardiac arrest” is particularly infuriating to pathologists; as one put it to me, “We all die of cardiac arrest. Your heart stops. It means nothing.” This happens not merely because doctors don’t (or can’t) take the time to actually work through the complex causes of death, but also because medical students are virtually never trained to fill out a death certificate. In my time researching this article, I heard of only one school, Johns Hopkins, that covered it as part of death training.
As a result, a vast percentage of death certificates — as much as 85% in some studies — have an error in the Cause of Death section, and about half have multiple errors. But despite their wild unreliability, death certificates are still a major source of data. Hospitals use them to compile mortality and morbidity numbers and send them to the National Center for Health Statistics and other agencies that allocate funding and resources. Death certificates are also the main source of state and national data on drug overdose deaths. (A 2018 study found that in the United States, the rate of opioid overdoses was 20% to 35% higher than reported figures.⁶)
As bad as the death certification issue sounds, autopsies also reveal something even more dire: As one study put it, “autopsies detect antemortem diagnostic errors at a frequent and enduringly consistent rate.” In over 10% of cases (some studies have estimated the number to be triple that), pathologists find that doctors missed something that, had it been caught, would have allowed the patient to walk out of the hospital alive.
“That’s 10% of deaths in America!” says Elizabeth Dole, MD, a former pathologist at Johns Hopkins who now runs a private autopsy practice in Dallas. “It’s the equivalent of a jumbo jet going down every day. But because it happens one by one, we don’t really pay attention. I’ve been a crusader for this for 20 years, but the rates keep going down. The process for autopsies is completely broken.”
The causes of the decline in autopsies — they’re not dead yet — are thorny and complicated. The most frequently cited problem is financial: In 1971, the Joint Committee for Accreditation of Health Organization (JCAHO) elected to eliminate a requirement that hospitals perform autopsies on 20% of deaths to maintain their accreditation, which can affect their ability to participate in Medicaid and Medicare. It’s worth noting, however, that this requirement was in place for only about six years, and there were problems with its implementation. Then, in 1986, Medicare decided that autopsies were not part of patient care and thus could not be funded by the government, and the rate plunged nationwide as hospitals eliminated their autopsy suites and stopped covering the cost for families. Today, Dole says, many private hospitals have given up their autopsy suites, and the vast majority are now done in teaching hospitals with pathology residency programs.
Death, it turns out, is hard to monetize⁷. Hospitals, as author and physician Siddartha Mukherjee has pointed out, have essentially become giant billing machines. But unlike other hospital procedures, an autopsy has no billing code. If something cannot be billed for, it exists in an administrative twilight zone. According to Harold Sanchez, MD, an assistant professor of pathology at Yale University School of Medicine, “No financial incentives currently exist for either the hospital or the pathologist to perform autopsies, because reforms in health care coverage and reimbursement have essentially eliminated direct funding for autopsies altogether.” Instead, hospitals get a lump sum from Medicare Part A to fund a “variety of services,” but because that money is fungible, Sanchez argues that it incentivizes hospitals to save the money for things that contribute more to their bottom line — in short, hospitals can make money by not doing autopsies. Lundberg is more blunt: “They don’t like it because you’re spending money on dead people.”
The result is that whether or not you can afford an autopsy depends in part on where you die. If you expire in a big teaching hospital, the service will likely be offered to you free of charge—that is, if the doctor is aware of the policy, which they probably are not. But otherwise, you’re likely to be referred to a private practice, like Dole’s, which can run more than $10,000, depending on the breadth of examination, and it will not be covered by insurance⁸.
The funding problem begets another, more intractable problem: a serious loss of expertise. To be board certified, pathologists must complete 50 autopsies. It used to be 100, but the Board of American Pathology was forced to reduce it because pathology residents simply were not able to hit those numbers. This inexperience matters because autopsy dissection, far from being blunt or brutal, is an exquisitely precise art. It is nothing short of full-body surgery, but without almost any of the technology that facilitates surgery on the living. Practitioner skill is paramount. When I was in his lab, Jeffrey Nine told me, “I recently saw a battlefield autopsy kit from the Civil War, and I thought, I could get the job done with that.” But Nine — who at the time I met him was also a Presbyterian pastor — has done more than 4,000 autopsies.
In over 10% of cases (some studies have estimated the number to be triple that), pathologists find that doctors missed something that, had it been caught, would have allowed the patient to walk out of the hospital alive.
“People don’t know how to do them anymore,” Dole says. “If you only did 50 autopsies when you were training, you’re not that proficient. You’re not going to know how to handle a complex case.”
All of this adds up to a dire shortage of forensic pathologists, not only in the United States, but worldwide. Troublingly, the average age for forensic pathologists is 52 — young doctors are not going into the field. And from a career perspective, it’s not hard to see why. Autopsy pathology is one of the least remunerative, least glamorous fields in medicine. Autopsy suites are invariably located deep in the subterranean bowels of the hospital. Working with the dead is also psychologically difficult — burnout rates among forensic pathologists in particular are exceptionally high. There’s no real way to numb oneself against having to dissect a child.
But even if all these things were suddenly solved — if Medicare started covering autopsies; if hospitals grasped their incomparable value as quality control; if, overnight, autopsy pathology became a dead-sexy⁹ field — one intractable bottleneck would still persist: Autopsies require the family’s consent, and autopsies freak people out.
Autopsies freak people out
On the one hand, autopsies do owe something to forensic procedurals like Bones: Many of the young people now entering the field were first hooked by seeing autopsies on TV (even when they are hilariously wrong). On the other hand, these shows cement the notion that autopsies are reserved for murder victims or patients who die of a mysterious disease. They also utterly misrepresent the science, Dole says, in a way that does real damage. “Those shows are frustrating. In an hour’s time, they’ve looked into every drug that could possibly go into a person’s body, and it doesn’t work that way. The expectations on the part of the families are too high, and there’s no one who explains the process to them.”
Yet of the myriad ways film and TV screw up the truth about autopsies, maybe the worst is the way many of them depict pathologists themselves: cold and emotionless eccentrics who have trouble functioning among the living, so they retreat to the house of the dead. (Grey’s Anatomy is maybe the worst offender, but competition is stiff.) The message we get is that the dead body is so taboo, so violently disagreeable, that you have to be at least a little deranged to tolerate it. These characters don’t just insult autopsists and their assistants—they actively discourage families from entrusting their loved ones’ bodies to them, imagining that anyone who would do this for a living must be irredeemably ghoulish.
Over the course of seven autopsies, I learned that however messy human bodies are in life, they are infinitely more so in death. I learned that a bad heart causes the liver to develop a mesmerizing “nutmeg” pattern. I learned that when I die, I intend every last piece of my body to go to science. Because I learned that to be an autopsist is to understand the weight of the life that has been lost. It is to hold that weight in your hands and let it speak. More than half of Americans die in hospitals or nursing homes, hidden behind nylon curtains and euphemism. The assumption that those who choose to work with the dead must be somehow damaged says far more about the way we in this era have deliberately insulated ourselves from death than it does about the people who have chosen not to.
Medicine is a stressful job, and all professionals kvetch about their clients — except autopsists. Snark is forbidden in an autopsy lab. In Pittsburgh, I was struck by the tone in the room. There was no punning, no grim but weary humor, no forced irreverence. Nor was it forbidding or sepulchral: Many pathologists and their assistants play music as they work, and they have conversations and joke around like anyone else. But they were as careful with their language as they were with their scalpels. They — and every pathologist I have spoken to — talked about the body as if it were a gift, because that is exactly what it is. Dr. Nine is clear: “Autopsy is the best way to honor somebody.”
But all these misconceptions have, with considerable administrative help, brought the autopsy — the foundational technique of modern medicine — to the brink of extinction. And then the virus showed up.
You can’t cure what you can’t see
This spring, a father and son were admitted to the ICU at Rush Copley Hospital outside Chicago. Both were sick with Covid-19. At 34, the son was healthier than his seventysomething father. The father survived. His son died.
For Hesham Hassaballa, MD, the pulmonologist in the Rush Copley ICU who cared for the father and son, their case remains one of the most agonizing examples of what makes the novel coronavirus so crushingly frustrating for doctors: “It’s the same virus, and it acts completely differently in everybody,” he explained to me over the phone. “What works for one patient may not work for the next,” even when the patients are as closely related as a father and son, Hassaballa says ruefully. In the 16 years he has practiced clinical care, he has never seen anyone so sick as his Covid-19 patients. This virus, he says, “causes destruction of lung tissue that happens in 30 years in 30 days.”
Autopsies aren’t just the gold standard for diagnosis; they are the best — and sometimes only — way of seeing the damage a disease does to a body.
This virus attacks the body in mysterious and baffling ways. “The public needs to understand that Covid-19 doesn’t just affect the lungs and cause a flulike illness,” says Judy Melinek, MD, author and former San Francisco–area pathologist. “There can be long-term consequences of the viral infection, including infection of the heart with heart rupture, and chronic damage to other organs, including the kidneys. In some patients, blood clots get so bad that their limbs need to be amputated.”
Since the pandemic began, doctors have been scrambling to understand how Covid-19 is transmitted, how it invades the body, how it sickens and kills. From the hell-for-leather global sprint for a vaccine to the FDA certifying dozens of suspect antibody tests, the medical establishment has been forced to drive itself so fast that the brakes on the adoption of new medicines or approaches haven’t just been ignored—they’ve burst into flame and disintegrated. Alfred Lee, MD, associate professor of hematology at Yale School of Medicine, told me, “In the olden days, we used to get our science from meetings and papers and so forth. Now, preprint” — meaning drafts of scientific articles that have not completed the peer-review process — “has taken over. Entire paradigm shifts take place overnight based entirely on hearsay.”
Autopsies aren’t just the gold standard for diagnosis; they are the best — and sometimes only — way of seeing the damage a disease does to a body. Given the all-hands-on-deck approach that the medical establishment has taken to fighting Covid-19 and the mysteriousness surrounding the virus in its early days, you might naturally assume that many of its victims would have received autopsies.
You would be wrong.
The first death publicly attributed to Covid-19 was in early January. But the autopsy is nearly extinct in China, where the coronavirus began, which is why it wasn’t until the end of February — after more than 1,500 people had already died — that the first autopsy report on a single Chinese Covid-19 patient came out. The first deaths in the United States were recorded around the same time, though it is possible that there were others before then. By mid-March, the WHO declared a global pandemic. But according to Sanjay Mukhopadhyay, the Cleveland Clinic pulmonologist, there was still a “vacuum” of autopsy data through the end of March. In other words, even with thousands already dead worldwide, virtually none of those deaths had been properly investigated. “There was a ton of clinical data and no pathological data,” Mukhopadhyay says. “When doctors asked us what is the underlying pathology of this disease, we couldn’t tell them.”
“Obviously, autopsies should have been done from the very beginning in every country in which Covid made an appearance. It’s as obvious as the nose on your face.”
One of the most critical developments in the understanding of the virus has been the discovery that between 49% and 79% (depending on the study) of patients develop a potentially lethal profusion of blood clots; as a result, the standard of care in all U.S. hospitals is now to start all incoming Covid-19 patients on low doses of blood thinners. But then, when autopsy data finally began trickling out, it revealed an even more common — and deadly — effect of the virus, one that even sophisticated CT scans were unable to pick up: a sudden eruption of hundreds of microscopic clots in the lungs. Recently, Lee and a team from Yale published a study in Lancet Hematology showing that the virus attacks endothelial cells — the “guardrails” that keep blood vessels clean and clot-free — not only in the sickest Covid-19 patients but also in those who had milder symptoms. And these were just patients who wound up in the hospital; right now, no one knows anything about what goes on in people who ride out the virus at home. (Disclosure: I’m one of those people; back in April I spent 28 days isolated from my family with fever, ragged breathing, and Columbo.) It is likely, though, that the damage Lee’s team detected is also happening to people who stay at home, especially those who are being found dead by EMTs.
Attacking endothelial cells allows SARS-CoV-2 to run rampant throughout the body like a conspiracy theory on Facebook. Knowing this gives doctors a clearer sense of how it spreads — and it’s possible this is how it affects so many organs and why it seems to present differently in all of Hesham Hassaballa’s patients. “Autopsy data here was critical,” Lee says. “Covid is an entirely new field. What was gained from autopsies at different points really helped move that field.”
The revelations go beyond clotting. In the early days of the virus, doctors were swift to intubate patients, leading to a worldwide run on ventilators. But intubation is a high-risk and dangerous procedure — not only to the patient, but also to the doctor, because thrusting a tube into the lungs forces a jet of toxic respiratory droplets in the air. Autopsies, says George Lundberg, MD, a veteran pathologist and former editor of Medscape, showed that Covid-19 lungs are two to three times heavier than normal lungs, filled with fluid, inflammation, and debris — there’s nowhere for forced air to go, so intubation doesn’t help; in many cases, providing passive oxygen is better. Pathologists are also beginning to find that Covid-19 can cause hyperactivity in the cells that produce platelets. They are finding these cells, called megakaryocytes, in places they’ve “never seen” before, like the heart. “Obviously,” Lundberg says, “autopsies should have been done from the very beginning in every country in which Covid made an appearance. It’s as obvious as the nose on your face.”
In other words, when people are dying as never before, it helps to look inside them. And yet, for all these findings, as I write this in September, six crushing months into this pandemic, only a relative handful of studies using autopsies have been published, and of those, only a few have come from the United States. “The Covid era,” Lee says, “has really amplified the need for autopsy data.” And yet, even at a teaching hospital as august as Yale, he has not heard of a single autopsy taking place. (I was unable to confirm this, as the Yale autopsy office did not respond to multiple requests for comment.)
There are a number of reasons for this: First and foremost, in the early days of the virus, there was considerable (and justified) fear that autopsists would be infected. “Autopsies are much more intimate than almost anything,” Mukhopadhyay says. “The fear was that that degree of intimacy might be infectious.” In March, the Occupational Safety and Health Administration also issued a confusing — and since-retracted — set of guidelines advising against doing any Covid-19 autopsies without a compelling reason. Although dead bodies are still able to transmit infection, with negative-pressure rooms and proper PPE, the process is safe: No autopsy practitioner has been shown to have gotten the virus from a body. (One incident in Thailand was reported, but it’s far from certain, Mukhopadhyay says, that the doctor was infected from the body.)
Yet now that autopsies have begun, however haltingly, political considerations have seeped in. One of the most difficult things for clinicians to sort out is whether a patient died from Covid-19 or with Covid-19. The difference matters. For one thing, the White House has dedicated itself to undermining or outright hiding statistics, which makes it all the more essential that clinicians get the numbers right. Even absent political cynicism, without an autopsy, the distinction remains so difficult that the CDC’s much-invoked number of coronavirus deaths (about 200,000 as of press time) is actually just a provisional death count. These are deaths attributed solely to Covid-19. But on its website, next to that tally, the CDC maintains two other columns: deaths due to pneumonia and Covid-19, and deaths due to pneumonia, influenza, and Covid-19. If you find this confusing, you’re not alone.
The struggle to autopsy Covid-19 patients has cost doctors and patients valuable time. But Covid-19 is just the spark that reveals — far too late — the presence of a gas leak in the broader world of medicine.
Autopsies are medicine
Autopsies are the reason humans know anything about medicine at all. As far back as the time of Charlemagne, they were used in legal cases, but it was not until the 13th century in Italy that the medical establishment began to embrace human dissection. In the mid-1700s, anatomists cemented the connection between pathological and clinical symptoms — putting to bed Galen’s theory of humors, which had ruled Western medicine for nearly 1,800 years.
Even so, for centuries, autopsies were the province of doctors who often paid grave robbers to exhume fresh corpses. The backlash to this practice was so extreme that, according to medical historian Michael Sappol, in the United States between 1765 and 1854, there were at least 17 “medical school riots,” in which torch-wielding mobs descended on nascent medical institutions, furious at the desecrations committed by the anatomists within. In 1788, 20 people died in a melee at Columbia College in New York, and then-Secretary of Foreign Affairs John Jay had his skull cracked with a rock. (It is worth noting that the Columbia faculty had long been stealing bodies with impunity from what was known as the “Negro Burial Ground” on Wall Street, despite efforts on the part of the city’s free and enslaved blacks to stop it; the rioting began only once it was reported that the doctors had exhumed a white woman’s body.)
“I don’t want this to be the last thing I say to them… You know the closure that the family is seeking… It almost feels like you’re delaying that and saying, ‘It’s not over yet. Let’s do some digging.’”
Even today, the clinical autopsy remains (behind, perhaps, electroshock therapy and abortion) one of medicine’s most polarizing — and misunderstood — procedures. Religious objections are common, especially among Jews and Muslims; Judaism, for instance, requires that the body be buried intact but allows autopsies if the cause of death is unclear, if the autopsy might help save lives, or to bring the family knowledge that might guide their grieving.
“The fear the family has that their loved one’s body will be desecrated in some way is very real,” says Karen Dahlman, PhD, who knows well the difficulty of asking families to make anatomical gifts: For more than six years, until 2011, she served as family liaison for the Alzheimer’s Disease Brain Bank at Mount Sinai in New York. “They’re afraid their loved one will suffer somehow, and it will look like they’ve suffered.” This fear is echoed even among doctors — including Hesham Hassaballa, who despite being confounded by the death of the 34-year-old Covid-19 patient whose father survived the disease, was not comfortable recommending to the family that he be autopsied. To Hassaballa, who knows the agony of losing a child, the prospect of a loss to science paled in comparison to the family’s need to have the body restored to them so they could at least begin the process of getting closure.
Hassaballa is far from alone in his aversion. Even though 68% of families in one study (other studies went even higher) gave permission if a doctor “very strongly” recommended it, doctors rarely even ask. This happens in large part because virtually all hospitals, when they ask for consent, do so at the worst possible time: immediately after death. Alfred Lee, whose own father is a pathologist, expressed the clinician’s dilemma when it comes to broaching the subject: “I don’t want this to be the last thing I say to them… You know the closure that the family is seeking and that, as a provider, you’re trying to bring. It almost feels like you’re delaying that and saying, ‘It’s not over yet. Let’s do some digging.’” His colleague George Goshua, MD, a Yale fellow in hematology and oncology and co-lead author on the endothelial cell study, puts it bluntly: “When you’re in a room and literally people are screaming with grief, that is not the time.”
That’s why, Dole says, the way the person asking for consent approaches the family makes all the difference; that person must be able to assuage those concerns and convey the value of the procedure. But in most hospitals, that person is often a young intern or resident, like Novick: overworked and underslept, with precious little coaching in autopsy process or policy and talking to a family member in what is likely the worst moment of their life.
In an ideal world where hospitals take autopsies seriously, a trained person from a decedent’s affairs office would contact the family independently and lay out their options, explaining the hospital’s autopsy protocol in detail, as well as its costs and benefits. Many doctors cannot have this conversation, because many doctors have never seen an autopsy — which also means they don’t fully grasp their value.
By catching hospitals’ mistakes, bringing comfort to families, and potentially catching new forms of disease before they spread, autopsies are a critical part of this continuum of care.
An autopsy makes a doctor better. It can show a doctor exactly how a treatment affected their patient’s body, which is crucial for clinical trials, and especially crucial in assessing the impact of Covid-19 treatment. It’s a powerful educational tool. It’s an essential metric for public health, and it is the only way to preserve tissue for research. “If you don’t look for your errors, you will never learn from them,” Judy Melinek argues. “An autopsy is the last chance a doctor has to see if the diagnoses they made in the living patient were accurate and understand why treatments failed.”
The strange truth about being a doctor today is that unless you are a surgeon, once you have completed first-year anatomy, you may never, not once, see the organ or system you specialize in with your own eyes. A pulmonologist may never hold a lung bubbled with emphysema — or leaden with Covid-19 — in their hand. An oncologist may never feel how a friable tumor crumbles when rubbed between their fingers. And a radiologist, whose job it is to see under patients’ skin, may never know what the glowing masses their images reveal look like in living color.
More important, perhaps, an autopsy can provide lasting closure and comfort for families.
In April 2017, Ted Trautman’s father died at a small hospital in St. Louis Park, Minnesota, a suburb of Minneapolis. Trautman’s father was quadriplegic, but his death was sudden. Complicated family dynamics led to confusion about the chain of custody, because one of the strangest legal elements of death is that when we die, we cease to be people in the eyes of the law. We become property. Legal power of attorney dissolves, and the chain of custody varies from state to state. (In some states, like Maryland, patients cannot consent to their own autopsy prior to death, a law pathologists are lobbying to change.) The cause of death was listed as “complications from quadriplegia.”
“I absolutely wish there had been an autopsy,” Trautman told me by phone, but the hospital did not have autopsy facilities. “I wasn’t there when he died, and I just wanted to know what happened. It was surprising to me that no one asked. People ask me what my dad died of, and I can’t really answer. I would like to be able to say, ‘My dad had a heart attack.’ But I can’t. There’s a blank spot. It feels like you should be able to say conclusively why someone died.”
Autopsists are deeply sensitive to their role in providing this comfort to grieving families. Jeffrey Nine believes autopsy should be the default for all deaths. “Not just for the patient’s sake,” he says, “but for the family’s. Without autopsy, you can’t tell if you’re harming patients.”
Autopsies aren’t supplements to medicine—they are medicine. Far from being remote or chilly, the way autopsy allows a doctor the rare chance to see the patient’s entire body and how their various pathologies are interlinked makes it a far more intimate form of medicine.
But the medicine autopsies provide isn’t fast-acting; it takes time for the knowledge they reveal to soak into the medical establishment. And if a country hemorrhages talent, erases facilities, and devalues the discipline, it can’t simply start up again. It takes years to train pathologists, and even then, it’s not like a single autopsy is going to completely change the face of medicine. That’s not its utility. Autopsies are part of what Atul Gawande has called the heroism of incremental care: the kind of steady, unflashy treatment practiced by primary care doctors who, rather than save a single life with a stroke of a scalpel, can save dozens through preventive and supportive care, like helping patients manage diabetes. By catching hospitals’ mistakes, bringing comfort to families, and potentially catching new forms of disease before they spread, autopsies are a critical part of this continuum of care.
Money follows demand. Right now, it’s easy for hospitals to claim they can’t afford to offer autopsies. As long as the autopsy is thought of as unspeakable, as mutilating, as beside the point — as long as we fail to understand the immense importance of planning for our own death and disposal — there is no pressure on hospitals or Medicare to cover them. This pandemic should be an opportunity to change that. It should be an opportunity for this nation to finally learn to square up to death a little, to bite the bullet and sit down to have the unpleasant conversations about end-of-life care and resuscitation and what to do with our bodies when we are gone, and write the answers down and hug and say “I love you,” and then go pick up something for dinner and maybe realize that conversation wasn’t nearly as bad as we’d braced ourselves for. These conversations don’t hasten death. They let us control it, at least a little. They let us get our minds around it. And that’s what autopsies are about, in the end. We can’t choose the way we die, but still, it’s our death. It’s the last thing that will ever belong to us. We deserve to let our loved ones know how it happened.
So, get an autopsy. It won’t hurt.
¹ I did this for the same reason I have done a number of demented things recently: I was doing research for a novel. It’s worth noting here that observations of autopsies are extremely rare; you can’t just walk in off the street. My visit was for research and journalistic purposes, and had been carefully arranged between my graduate program at the time and UPMC.
² The slide that Dr. Nine showed me of the lung ravaged by an overdose was not from an ongoing clinical autopsy, since an overdose would not have been in his purview as an academic pathologist; it was most likely a slide used for teaching purposes.
³ Medical examiners, by the way, are different from coroners, and in the US, different states use different systems. A medical examiner is a specialist in forensic pathology. The coroner can be anyone. In 1,300 counties in America, the coroner is an elected politician, not necessarily a doctor, which is exactly as bad an idea as it sounds.
⁴ One pathologist I spoke to put the value of a forensic pathologist simply: “When the legal system gets involved, it benefits everyone to have someone familiar with how bullets move in a body and can convey that to a jury.”
⁵ To make the distinction clear: Forensic and clinical autopsies involve similar techniques, and there’s abundant knowledge-sharing between them — some pathologists, like Dr. Nine, can do both. But their expertises and clienteles are different. Clinical autopsists serve hospitals and families, and can be done by any board-certified pathologist; forensic pathologists are specialists who serve the state (or work in private practice) and assist with law enforcement investigations.
⁷ At least, it is for hospitals.
⁸ Private services also can provide forensic services in criminal cases. In May of this year, the family of George Floyd, understandably mistrustful of the Minneapolis authorities, engaged a private pathologist to conduct an independent autopsy.
⁹ I’m sorry.