by Atul Gawande
I had assumed that achieving this kind of teamwork was mostly a matter of luck. I'd certainly experienced it at times--difficult operations in which everyone was nonetheless firing on all cylinders, acting as one. I remember an eighty-year-old patient who required an emergency operation. He had undergone heart surgery the week before and had been recovering nicely. But during the night he'd developed a sudden, sharp, unrelenting pain in his abdomen, and over the course of the morning it had mounted steadily in severity. I was asked to see him from general surgery. I found him lying in bed, prostrate with pain. His heart rate was over one hundred and irregular. His blood pressure was dropping. And wherever I touched his abdomen, the sensation made him almost leap off the bed in agony.
He knew this was trouble. His mind was completely sharp. But he didn't seem scared.
"What do we need to do?" he asked between gritted teeth.
I explained that I believed his body had thrown a clot into his intestine's arterial supply. It was as if he'd had a stroke, only this one had cut off blood flow to his bowel, not his brain. Without blood flow, his bowel would turn gangrenous and rupture. This was not survivable without surgery. But, I also had to tell him, it was often not survivable even with surgery. Perhaps half of the patients in his circumstance make it through. If he was one of them, there would be many complications to worry about. He might need a ventilator or a feeding tube. He'd already been through one major operation. He was weak and not young. I asked him if he wanted to go ahead.
Yes, he said, but he wanted me to speak with his wife and son first. I reached them by phone. They too said to proceed. I called the operating room control desk and explained the situation. I needed an OR and a team right away. I'd take what ever and whoever were available.
We got him to the OR within the hour. And as people assembled and set to work, there was the sense of a genuine team taking form. Jay, the circulating nurse, introduced himself to the patient and briefly explained what everyone was doing. Steve, the scrub nurse, was already gowned and gloved, standing by with the sterile instruments at the ready. Zhi, the senior anesthesiologist, and Thor, his resident, were conferring, making sure they had their plans straight, as they set out their drugs and equipment. Joaquim, the surgery resident, stood by with a Foley catheter, ready to slip it into the patient's bladder as soon as he was asleep.
The clock was ticking. The longer we took, the more bowel would die. The more bowel that died, the sicker the man would become and the lower his chance of survival. Everyone understood this, which by itself was a lot. People don't always get it--really feel the urgency of the patient's condition. But these people did. They were swift, methodical, and in sync. The case was far from easy, but for what ever reason, it seemed like nothing could thwart us.
The patient was a big man with a short neck and not much lung reserve, making it potentially difficult to place a breathing tube when Zhi sent him off to sleep. But Zhi had warned us of the possibility of trouble and everyone was ready with a backup plan and the instruments he and Thor might need. When Joaquim and I opened up the patient, we found that the right colon was black with gangrene--it had died--but it had not ruptured, and the remaining three-fourths of the colon and all the small bowel seemed to be okay. This was actually good news. The problem was limited. As we began removing the right colon, however, it became evident that the rest of the colon was not, in fact, in good shape. Where it should have been healthy pink, we found scattered dime-and quarter-sized patches of purple. The blood clots that had blocked off the main artery to the right colon had also showered into the arterial branches of the left side. We would have to remove the patient's entire colon, all four feet of it, and give him an ostomy--a bag for his excreted wastes. Steve, thinking ahead, asked Jay to grab a retractor we'd need. Joaquim nudged me to make the abdominal incision bigger, and he stayed with me at every step, clamping, cutting, and tying as we proceeded inch by inch through the blood vessels tethering the patient's colon. The patient began oozing blood from every raw surface--toxins from the gangrene were causing him to lose his ability to clot. But Zhi and Thor kept up with the fluid requirements and the patient's blood pressure was actually better halfway through than it had been at the beginning. When I mentioned that I thought the patient would need an ICU, Zhi told me he'd already arranged it and briefed the intensivist.
Because we'd worked as a single unit, not as separate technicians, the man survived. We were done with the operation in little more than two hours; his vital signs were stable; and he would leave the hospital just a few days later. The family gave me the credit, and I wish I could have taken it. But the operation had been symphonic, a thing of orchestral beauty.
Perhaps I could claim that the teamwork itself had been my doing. But its origins were mysterious to me. I'd have said it was just the good fortune of the circumstances--the accidental result of the individuals who happened to be available for the case and their particular chemistry on that particular afternoon. Although I operated with Zhi frequently, I hadn't worked with Jay or Steve in months, Joaquim in even longer. I'd worked with Thor just once. As a group of six, we'd never before done an operation together. Such a situation is not uncommon in hospitals of any significant size. My hospital has forty-two operating rooms, staffed by more than a thousand personnel. We have new nurses, technicians, residents, and physician staff almost constantly. We're virtually always adding strangers to our teams. As a consequence, the level of teamwork--an unspoken but critical component of success in surgery--is unpredictable. Yet somehow, from the moment we six were all dropped together into this particular case, things clicked. It had been almost criminally enjoyable.
This seemed like luck, as I say. But suppose it wasn't. That's what the checklists from Toronto and Hopkins and Kaiser raised as a possibility. Their insistence that people talk to one another about each case, at least just for a minute before starting, was basically a strategy to foster teamwork--a kind of team huddle, as it were. So was another step that these checklists employed, one that was quite unusual in my experience: surgical staff members were expected to stop and make sure that everyone knew one another's names.
The Johns Hopkins checklist spelled this out most explicitly. Before starting an operation with a new team, there was a check to ensure everyone introduced themselves by name and role: "I'm Atul Gawande, the attending surgeon"; "I'm Jay Powers, the circulating nurse"; "I'm Zhi Xiong, the anesthesiologist"--that sort of thing.
It felt kind of hokey to me, and I wondered how much difference this step could really make. But it turned out to have been carefully devised. There have been psychology studies in various fields backing up what should have been self-evident--people who don't know one another's names don't work together nearly as well as those who do. And Brian Sexton, the Johns Hopkins psychologist, had done studies showing the same in operating rooms. In one, he and his research team buttonholed surgical staff members outside their operating rooms and asked them two questions: how would they rate the level of communications during the operation they had just finished and what were the names of the other staff members on the team? The researchers learned that about half the time the staff did not know one another's names. When they did, however, the communications ratings jumped significantly.
The investigators at Johns Hopkins and elsewhere had also observed that when nurses were given a chance to say their names and mention concerns at the beginning of a case, they were more likely to note problems and offer solutions. The researchers called it an "activation phenomenon." Giving people a chance to say something at the start seemed to activate their sense of participation and responsibility and their willingness to speak up.
These were limited studies and hardly definitive. But the initial results were enticing. Nothing had ever been shown to improve the ability of surgeons to broadly reduce harm to patients aside from experience and specialized training. Yet here, in three separate cities, teams had tried out these unusual checklists, and each had found a positive effe
ct.
At Johns Hopkins, researchers specifically measured their checklist's effect on teamwork. Eleven surgeons had agreed to try it in their cases--seven general surgeons, two plastic surgeons, and two neurosurgeons. After three months, the number of team members in their operations reporting that they "functioned as a well-coordinated team" leapt from 68 percent to 92 percent.
At the Kaiser hospitals in Southern California, researchers had tested their checklist for six months in thirty-five hundred operations. During that time, they found that their staff 's average rating of the teamwork climate improved from "good" to "outstanding." Employee satisfaction rose 19 percent. The rate of OR nurse turnover--the proportion leaving their jobs each year--dropped from 23 percent to 7 percent. And the checklist appeared to have caught numerous near errors. In one instance, the preoperative briefing led the team to recognize that a vial of potassium chloride had been switched with an antibiotic vial--a potentially lethal mix-up. In another, the checklist led the staff to catch a paperwork error that had them planning for a thoracotomy, an open-chest procedure with a huge front-to-back wound, when what the patient had come in for was actually a thoracoscopy, a videoscope procedure done through a quarter-inch incision.
At Toronto, the researchers physically observed operations for specific evidence of impact. They watched their checklist in use in only eighteen operations. But in ten of those eighteen, they found that it had revealed significant problems or ambiguities--in more than one case, a failure to give antibiotics, for example; in another, uncertainty about whether blood was available; and in several, the kinds of unique and individual patient problems that I would not have expected a checklist to help catch.
They reported one case, for example, involving an abdominal operation under a spinal anesthetic. In such procedures, we need the patient to report if he or she begins to feel even a slight twinge of pain, indicating the anesthetic might be wearing off and require supplementation. But this particular patient had a severe neurological condition that had left him unable to communicate verbally. Instead, he communicated through hand signals. Normally, we restrain the arms and hands of patients to keep them from inadvertently reaching around the sterile drapes and touching the surgeons or the operative field. In this instance, however, the regular routine would have caused a serious problem, but this was not clearly recognized by the team until just before the incision was made. That was when the surgeon walked in, pulled on his gown and gloves, and stepped up to the operating table. Because of the checklist, instead of taking the knife, he paused and conferred with everyone about the plans for the operation. The Toronto report included a transcript of the discussion.
"Are there any special anesthetic considerations?" the surgeon asked.
"Just his dysarthria," the anesthesiologist said, referring to the patient's inability to speak.
The surgeon thought for a moment. "It may be difficult to gauge his neurological function because we have these issues," he said.
The anesthesiologist agreed. "I've worked out a system of hand signals with him."
"His arm will [need to] be accessible then--not tucked," the surgeon said. The anesthesiologist nodded, and the team then worked out a way to leave the patient's arms free but protected from reaching around or beneath the drapes.
"My other concern is the number of people in the room," the anesthesiologist went on, "because noise and movement may interfere with our ability to communicate with the patient."
"We can request silence," the surgeon said. Problem solved.
None of these studies was complete enough to prove that a surgical checklist could produce what WHO was ultimately looking for--a measurable, inexpensive, and substantial reduction in overall complications from surgery. But by the end of the Geneva conference, we had agreed that a safe surgery checklist was worth testing on a larger scale.
A working group took the different checklists that had been tried and condensed them into a single one. It had three "pause points," as they are called in aviation--three points at which the team must stop to run through a set of checks before proceeding. There was a pause right before the patient is given anesthesia, one after the patient is anesthetized but before the incision is made, and one at the end of the operation, before the patient is wheeled out of the operating room. The working group members divvied up the myriad checks for allergies, antibiotics, anesthesia equipment, and so on among the different pause points. They added any other checks they could think of that might make a difference in care. And they incorporated the communication checks in which everyone in the operating room ensures that they know one another's names and has a chance to weigh in on critical plans and concerns.
We made a decision to set up a proper pilot study of our safe surgery checklist in a range of hospitals around the world, for which WHO committed to providing the funds. I was thrilled and optimistic. When I returned home to Boston, I jumped to give the checklist a try myself. I printed it out and took it to the operating room. I told the nurses and anesthesiologists what I'd learned in Geneva.
"So how about we try this awesome checklist?" I said. It detailed steps for everything from equipment inspection to antibiotic administration to the discussions we should have. The rest of the team eyed me skeptically, but they went along. "Sure, what ever you say." This was not the first time I'd cooked up some cockamamie idea.
I gave the checklist to Dee, the circulating nurse, and asked her to run through the first section with us at the right time. Fifteen minutes later, we were about to put the patient to sleep under general anesthesia, and I had to say, Wait, what about the checklist?
"I already did it," Dee said. She showed me the sheet. All the boxes were checked off.
No, no, no, I said. It's supposed to be a verbal checklist, a team checklist.
"Where does it say that?" she asked. I looked again. She was right. It didn't say that anywhere.
Just try it verbally anyway, I said.
Dee shrugged and started going through the list. But some of the checks were ambiguous. Was she supposed to confirm that everyone knew the patient's allergies or actually state the allergies? she asked. And after a few minutes of puzzling our way through the list, everyone was becoming exasperated. Even the patient started shifting around on the table.
"Is everything okay?" she asked.
Oh yes, I told her. We're only going through our checklist. Don't worry.
But I was getting impatient, too. The checklist was too long. It was unclear. And past a certain point, it was starting to feel like a distraction from the person we had on the table.
By the end of the day, we had stopped using the checklist. Forget making this work around the world. It wasn't even working in one operating room.
6. THE CHECKLIST FACTORY
Some time after that first miserable try, I did what I should have done to begin with. I went to the library and pulled out a few articles on how flight checklists are made. As great as the construction-world checklists seemed to be, they were employed in projects that routinely take months to complete. In surgery, minutes matter. The problem of time seemed a serious limitation. But aviation had this challenge, too, and somehow pilots' checklists met it.
Among the articles I found was one by Daniel Boorman from the Boeing Company in Seattle, Washington. I gave him a call. He proved to be a veteran pilot who'd spent the last two decades developing checklists and flight deck controls for Boeing aircraft from the 747-400 forward. He'd most recently been one of the technical leaders behind the flight deck design for the new 787 Dreamliner, including its pilot controls, displays, and system of checklists. He is among the keepers of what could be called Boeing's "flight philosophy." When you get on a Boeing aircraft, there is a theory that governs the way your cockpit crew flies that plane--what their routines are, what they do manually, what they leave to computers, and how they should react when the unexpected occurs. Few have had more experience translating the theory into practice than Dan Boorman. He is the lineal de
scendant of the pilots who came up with that first checklist for the B-17 bomber three-quarters of a century ago. He has studied thousands of crashes and near crashes over the years, and he has made a science of averting human error.
I had a trip to Seattle coming up, and he was kind enough to agree to a visit. So one fall day, I drove a rental car down a long flat road on the city's outskirts to Boeing's headquarters. They appeared rather ordinary--a warren of low, rectangular, institutional-looking buildings that would not be out of place on the campus of an underfunded state college, except for the tarmac and hangar of airplanes behind them. Boorman came out to meet me at security. He was fifty-one, pilot-trim, in slacks and an open-collared oxford shirt--more like an engineering professor than a company man. He took me along a path of covered concrete sidewalks to Building 3-800, which was as plain and functional as it sounds. A dusty display case with yellowing pictures of guys in silver flight suits appeared not to have been touched since the 1960s. The flight test division was a fluorescent-lit space filled with dun-colored cubicles. We sat down in a windowless conference room in their midst. Piles of checklist handbooks from US Airways, Delta, United, and other airlines lay stacked against a wall.
Boorman showed me one of the handbooks. It was spiral bound, about two hundred pages long, with numerous yellow tabs. The aviation checklist had clearly evolved since the days of a single card for taxi, takeoff, and landing, and I wondered how anyone could actually use this hefty volume. As he walked me through it, however, I realized the handbook was comprised not of one checklist but of scores of them. Each one was remarkably brief, usually just a few lines on a page in big, easy-to-read type. And each applied to a different situation. Taken together, they covered a vast range of flight scenarios.