by Atul Gawande
Johns hunched over the patient, trying intently to insert the tube through her vocal cords. When her sat once again dropped into the 60s, he stopped and put the mask back on. We stared at the monitor. The numbers weren’t coming up. Her lips were still blue. Johns squeezed the bellows harder to blow more oxygen in.
“I’m getting resistance,” he said.
The realization crept over me: this was a disaster. “Damn it, we’ve lost her airway,” I said. “Trache kit! Light! Somebody call down to OR 25 and get Ball up here!”
People were suddenly scurrying everywhere. I tried to proceed deliberately, and not let panic take hold. I told the surgical intern to get a sterile gown and gloves on. I took an antiseptic solution off a shelf and dumped a whole bottle of yellow-brown liquid on the patient’s neck. A nurse unwrapped the tracheostomy kit—a sterilized set of drapes and instruments. I pulled on a gown and a new pair of gloves while trying to think through the steps. This is simple, really, I tried to tell myself. At the base of the thyroid cartilage, the Adam’s apple, is a little gap in which you find a thin, fibrous covering called the cricothyroid membrane. Cut through that and—voilà! You’re in the trachea. You slip through the hole a four-inch plastic tube shaped like a plumber’s elbow joint, hook it up to oxygen and a ventilator, and she’s all set. Anyway, that was the theory.
I threw some drapes over her body, leaving the neck exposed. It looked as thick as a tree. I felt for the bony prominence of the thyroid cartilage. But I couldn’t feel anything through the layers of fat. I was beset by uncertainty—where should I cut? should I make a horizontal or a vertical incision?—and I hated myself for it. Surgeons never dithered, and I was dithering.
“I need better light,” I said.
Someone was sent out to look for one.
“Did anyone get Ball?” I asked. It wasn’t exactly an inspiring question.
“He’s on his way,” a nurse said.
There was no time to wait. Four minutes without oxygen would lead to permanent brain damage, if not death. Finally, I took the scalpel and cut. I just cut. I made a three-inch left-to-right swipe across the middle of the neck, following the procedure I’d learned for elective cases. Dissecting down with scissors while the intern held the wound open with retractors, I hit a vein. It didn’t let loose a lot of blood, but there was enough to fill the wound: I couldn’t see anything. The intern put a finger on the bleeder. I called for suction. But the suction wasn’t working; the tube was clogged with clot from the intubation efforts.
“Somebody get some new tubing,” I said. “And where’s the light?”
Finally, an orderly wheeled in a tall overhead light, plugged it in, and flipped on the switch. It was still too dim; I could have done better with a flashlight.
I wiped up the blood with gauze, then felt around in the wound with my fingertips. This time, I thought I could feel the hard ridges of the thyroid cartilage and, below it, the slight gap of the cricothyroid membrane, though I couldn’t be sure. I held my place with my left hand.
James O’Connor, a silver-haired, seen-it-all anesthesiologist, came into the room. Johns gave him a quick rundown on the patient and let him take over ventilating her.
Holding the scalpel in my right hand like a pen, I stuck the blade down into the wound at the spot where I thought the thyroid cartilage was. With small, sharp strokes—working blindly, because of the blood and the poor light—I cut down through the overlying fat and tissue until I felt the blade scrape against the almost bony cartilage. I searched with the tip of the knife, walking it along until I felt it reach a gap. I hoped it was the cricothyroid membrane, and pressed down firmly. I felt the tissue suddenly give, and I cut an inch-long opening.
When I put my index finger into it, it felt as if I were prying open the jaws of a stiff clothespin. Inside, I thought I felt open space. But where were the sounds of moving air that I expected? Was this deep enough? Was I even in the right place?
“I think I’m in,” I said, to reassure myself as much as anyone else.
“I hope so,” O’Connor said. “She doesn’t have much longer.”
I took the tracheostomy tube and tried to fit it in, but something seemed to be blocking it. I twisted it and turned it, and finally jammed it in. Just then Ball, the surgical attending, arrived. He rushed up to the bed and leaned over for a look. “Did you get it?” he asked. I said that I thought so. The bag mask was plugged onto the open end of the trache tube. But when the bellows were compressed the air just gurgled out of the wound. Ball quickly put on gloves and a gown.
“How long has she been without an airway?” he asked.
“I don’t know. Three minutes.”
Ball’s face hardened as he registered that he had about a minute in which to turn things around. He took my place and summarily pulled out the trache tube. “God, what a mess,” he said. “I can’t see a thing in this wound. I don’t even know if you’re in the right place. Can we get better light and suction?” New suction tubing was found and handed to him. He quickly cleaned up the wound and went to work.
The patient’s sat had dropped so low that the oximeter couldn’t detect it anymore. Her heart rate began slowing down—first to the 60s and then to the 40s. Then she lost her pulse entirely. I put my hands together on her chest, locked my elbows, leaned over her, and started doing chest compressions.
Ball looked up from the patient and turned to O’Connor. “I’m not going to get her an airway in time,” he said. “You’re going to have to try again from above.” Essentially, he was admitting my failure. Trying an oral intubation again was pointless—just something to do instead of watching her die. I was stricken, and concentrated on doing chest compressions, not looking at anyone. It was over, I thought.
And then, amazingly, O’Connor: “I’m in.” He had managed to slip a pediatric-size endotracheal tube through the vocal cords. In thirty seconds, with oxygen being manually ventilated through the tube, her heart was back, racing at a hundred and twenty beats a minute. Her sat registered at 60 and then climbed. Another thirty seconds and it was at 97 percent. All the people in the room exhaled, as if they, too, had been denied their breath. Ball and I said little except to confer about the next steps for her. Then he went back downstairs to finish working on the stab-wound patient still in the OR.
We eventually identified the woman, whom I’ll call Louise Williams; she was thirty-four years old and lived alone in a nearby suburb. Her alcohol level on arrival had been three times the legal limit, and had probably contributed to her unconsciousness. She had a concussion, several lacerations, and significant soft-tissue damage. But X rays and scans revealed no other injuries from the crash. That night, Ball and Hernandez brought her to the OR to fit her with a proper tracheostomy. When Ball came out and talked to family members, he told them of the dire condition she was in when she arrived, the difficulties “we” had had getting access to her airway, the disturbingly long period of time that she had gone without oxygen, and thus his uncertainty about how much brain function she still possessed. They listened without protest; there was nothing for them to do but wait.
Consider some other surgical mishaps. In one, a general surgeon left a large metal instrument in a patient’s abdomen, where it tore through the bowel and the wall of the bladder. In another, a cancer surgeon biopsied the wrong part of a woman’s breast and thereby delayed her diagnosis of cancer for months. A cardiac surgeon skipped a small but key step during a heart valve operation, thereby killing the patient. A general surgeon saw a man racked with abdominal pain in the emergency room and, without taking a CT scan, assumed that the man had a kidney stone; eighteen hours later, a scan showed a rupturing abdominal aortic aneurysm, and the patient died not long afterward.
How could anyone who makes a mistake of that magnitude be allowed to practice medicine? We call such doctors “incompetent,” “unethical,” and “negligent.” We want to see them punished. And so we’ve wound up with the public system we have for dealing with error: m
alpractice lawsuits, media scandal, suspensions, firings.
There is, however, a central truth in medicine that complicates this tidy vision of misdeeds and misdoers: all doctors make terrible mistakes. Consider the cases I’ve just described. I gathered them simply by asking respected surgeons I know—surgeons at top medical schools—to tell me about mistakes they had made just in the past year. Every one of them had a story to tell.
In 1991, the New England Journal of Medicine published a series of landmark papers from a project known as the Harvard Medical Practice Study—a review of more than thirty thousand hospital admissions in New York State. The study found that nearly 4 percent of hospital patients suffered complications from treatment which either prolonged their hospital stay or resulted in disability or death, and that two-thirds of such complications were due to errors in care. One in four, or 1 percent of admissions, involved actual negligence. It was estimated that, nationwide, upward of forty-four thousand patients die each year at least partly as a result of errors in care. And subsequent investigations around the country have confirmed the ubiquity of error. In one small study of how clinicians perform when patients have a sudden cardiac arrest, twenty-seven of thirty clinicians made an error in using the defibrillator—charging it incorrectly or losing too much time trying to figure out how to work a particular model. According to a 1995 study, mistakes in administering drugs—giving the wrong drug or the wrong dose, say—occur, on average, about once every hospital admission, mostly without ill effects, but 1 percent of the time with serious consequences.
If error were due to a subset of dangerous doctors, you might expect malpractice cases to be concentrated among a small group, but in fact they follow a uniform, bell-shaped distribution. Most surgeons are sued at least once in the course of their careers. Studies of specific types of error, too, have found that repeat offenders are not the problem. The fact is that virtually everyone who cares for hospital patients will make serious mistakes, and even commit acts of negligence, every year. For this reason, doctors are seldom outraged when the press reports yet another medical horror story. They usually have a different reaction: That could be me. The important question isn’t how to keep bad physicians from harming patients; it’s how to keep good physicians from harming patients.
Medical malpractice suits are a remarkably ineffective remedy. Troyen Brennan, a Harvard professor of law and public health, points out that research has consistently failed to find evidence that litigation reduces medical error rates. In part, this may be because the weapon is so imprecise. Brennan led several studies following up on the patients in the Harvard Medical Practice Study. He found that fewer than 2 percent of the patients who had received substandard care ever filed suit. Conversely, only a small minority among the patients who did sue had in fact been the victims of negligent care. And a patient’s likelihood of winning a suit depended primarily on how poor his or her outcome was, regardless of whether that outcome was caused by disease or unavoidable risks of care.
The deeper problem with medical malpractice suits is that by demonizing errors they prevent doctors from acknowledging and discussing them publicly. The tort system makes adversaries of patient and physician, and pushes each to offer a heavily slanted version of events. When things go wrong, it’s almost impossible for a physician to talk to a patient honestly about mistakes. Hospital lawyers warn doctors that, although they must, of course, tell patients about injuries that occur, they are never to intimate that they were at fault, lest the “confession” wind up in court as damning evidence in a black-and-white morality tale. At most, a doctor might say, “I’m sorry that things didn’t go as well as we had hoped.”
There is one place, however, where doctors can talk candidly about their mistakes, if not with patients, then at least with one another. It is called the Morbidity and Mortality Conference—or, more simply, M & M—and it takes place, usually once a week, at nearly every academic hospital in the country. This institution survives because laws protecting its proceedings from legal discovery have stayed on the books in most states, despite frequent challenges. Surgeons, in particular, take the M & M seriously. Here they can gather behind closed doors to review the mistakes, untoward events, and deaths that occurred on their watch, determine responsibility, and figure out what to do differently next time.
At my hospital, we convene every Tuesday at five o’clock in a steep, plush amphitheater lined with oil portraits of the great doctors whose achievements we’re meant to live up to. All surgeons are expected to attend, from the interns to the chairman of surgery; we’re also joined by medical students doing their surgery “rotation.” An M & M can include almost a hundred people. We file in, pick up a photocopied list of cases to be discussed, and take our seats. The front row is occupied by the most senior surgeons: terse, serious men, now out of their scrubs and in dark suits, lined up like a panel of senators at a hearing. The chairman is a leonine presence in the seat closest to the plain wooden podium from which each case is presented. In the next few rows are the remaining surgical attendings; these tend to be younger, and several of them are women. The chief residents have put on long white coats and usually sit in the side rows. I join the mass of other residents, all of us in short white coats and green scrub pants, occupying the back rows.
For each case, the chief resident from the relevant service—cardiac, vascular, trauma, and so on—gathers the information, takes the podium, and tells the story. Here’s a partial list of cases from a typical week (with a few changes to protect confidentiality): a sixty-eight-year-old man who bled to death after heart valve surgery; a forty-seven-year-old woman who had to have a reoperation because of infection following an arterial bypass done in her left leg; a forty-four-year-old woman who had to have bile drained from her abdomen after gallbladder surgery; three patients who had to have reoperations for bleeding following surgery; a sixty-three-year-old man who had a cardiac arrest following heart bypass surgery; a sixty-six-year-old woman whose sutures suddenly gave way in an abdominal wound and nearly allowed her intestines to spill out. Ms. Williams’s case, my failed tracheostomy, was just one case on a list like this. David Hernandez, the chief trauma resident, had subsequently reviewed the records and spoken to me and others involved. When the time came, it was he who stood up front and described what had happened.
Hernandez is a tall, rollicking, good old boy who can tell a yarn, but M & M presentations are bloodless and compact. He said something like: “This was a thirty-four-year-old female unrestrained driver in a high-speed rollover. The patient apparently had stable vitals at the scene but was unresponsive, and was brought in by ambulance unintubated. She was GCS 7 on arrival.” GCS stands for the Glasgow Coma Scale, which rates the severity of head injuries, from three to fifteen. GCS 7 is in the comatose range. “Attempts to intubate were made without success in the ER and may have contributed to airway closure. A cricothyroidotomy was attempted without success.”
These presentations can be awkward. The chief residents, not the attendings, determine which cases to report. That keeps the attendings honest—no one can cover up mistakes—but it puts the chief residents, who are, after all, underlings, in a delicate position. The successful M & M presentation inevitably involves a certain elision of detail and a lot of passive verbs. No one screws up a cricothyroidotomy. Instead, “a cricothyroidotomy was attempted without success.” The message, however, was not lost on anyone.
Hernandez continued, “The patient arrested and required cardiac compressions. Anesthesia was then able to place a pediatric ET tube and the patient recovered stable vitals. The tracheostomy was then completed in the OR.”
So Louise Williams had been deprived of oxygen long enough to go into cardiac arrest, and everyone knew that meant she could easily have suffered a disabling stroke or worse. Hernandez concluded with the fortunate aftermath: “Her workup was negative for permanent cerebral damage or other major injuries. The tracheostomy tube was removed on Day 2. She was discharged to home in g
ood condition on Day 3.” To the family’s great relief, and mine, she had woken up in the morning a bit woozy but hungry, alert, and mentally intact. In a few weeks, the episode would heal to a scar.
But not before someone was called to account. A front-row voice immediately thundered, “What do you mean, ‘a cricothyroidotomy was attempted without success’?” I sank into my seat, my face hot.
“This was my case,” Dr. Ball volunteered from the front row. It is how every attending begins, and that little phrase contains a world of surgical culture. For all the talk in business schools and in corporate America about the virtues of “flat organizations,” surgeons maintain an old-fashioned sense of hierarchy. When things go wrong, the attending is expected to take full responsibility. It makes no difference whether it was the resident’s hand that slipped and lacerated an aorta; it doesn’t matter whether the attending was at home in bed when a nurse gave a wrong dose of medication. At the M & M, the burden of responsibility falls on the attending.
Ball went on to describe the emergency attending’s failure to intubate Williams and his own failure to be at her bedside when things got out of control. He described the bad lighting and her extremely thick neck, and was careful to make those sound not like excuses but merely like complicating factors. Some attendings shook their heads in sympathy. A couple of them asked questions to clarify certain details. Throughout, Ball’s tone was objective, detached. He had the air of a CNN newscaster describing unrest in Kuala Lumpur.
