The Checklist Manifesto: How to Get Things Right

by Atul Gawande

The capability of individuals is not proving to be our primary difficulty, whether in medicine or elsewhere. Far from it. Training in most fields is longer and more intense than ever. People spend years of sixty-, seventy-, eighty-hour weeks building their base of knowledge and experience before going out into practice on their own—whether they are doctors or professors or lawyers or engineers. They have sought to perfect themselves. It is not clear how we could produce substantially more expertise than we already have. Yet our failures remain frequent. They persist despite remarkable individual ability.

Maybe we should be asking if 60-80 hour weeks are the best way to learn, practice, and retain knowledge and if those conditions are optimal for thinking and making decisions?

I asked the people in Harvard Vanguard’s medical records department if they would query the electronic system for how many different kinds of patient problems the average doctor there sees annually. The answer that came back flabbergasted me. Over the course of a year of office practice—which, by definition, excludes the patients seen in the hospital—physicians each evaluated an average of 250 different primary diseases and conditions. Their patients had more than nine hundred other active medical problems that had to be taken into account. The doctors each prescribed some three hundred medications, ordered more than a hundred different types of laboratory tests, and performed an average of forty different kinds of office procedures—from vaccinations to setting fractures. Even considering just the office work, the statistics still didn’t catch all the diseases and conditions. One of the most common diagnoses, it turned out, was “Other.”

Maybe physicians need research assistants for non emergency cases that appear complex.

A further difficulty, just as insidious, is that people can lull themselves into skipping steps even when they remember them. In complex processes, after all, certain steps don’t always matter. Perhaps the elevator controls on airplanes are usually unlocked and a check is pointless most of the time. Perhaps measuring all four vital signs uncovers a worrisome issue in only one out of fifty patients. “This has never been a problem before,” people say. Until one day it is. Checklists seem to provide protection against such failures. They remind us of the minimum necessary steps and make them explicit. They not only offer the possibility of verification but also instill a kind of discipline of higher performance.

In 2001, though, a critical care specialist at Johns Hopkins Hospital named Peter Pronovost decided to give a doctor checklist a try. He didn’t attempt to make the checklist encompass everything ICU teams might need to do in a day. He designed it to tackle just one of their hundreds of potential tasks, the one that nearly killed Anthony DeFilippo: central line infections. On a sheet of plain paper, he plotted out the steps to take in order to avoid infections when putting in a central line. Doctors are supposed to (1) wash their hands with soap, (2) clean the patient’s skin with chlorhexidine antiseptic, (3) put sterile drapes over the entire patient, (4) wear a mask, hat, sterile gown, and gloves, and (5) put a sterile dressing over the insertion site once the line is in. Check, check, check, check, check. These steps are no-brainers; they have been known and taught for years. So it seemed silly to make a checklist for something so obvious. Still, Pronovost asked the nurses in his ICU to observe the doctors for a month as they put lines into patients and record how often they carried out each step. In more than a third of patients, they skipped at least one. The next month, he and his team persuaded the Johns Hopkins Hospital administration to authorize nurses to stop doctors if they saw them skipping a step on the checklist; nurses were also to ask the doctors each day whether any lines ought to be removed, so as not to leave them in longer than necessary. This was revol...

But here was a situation that hadn’t been anticipated on the construction checklist: the tilting of the upper floors. At a minimum, a water cleanup would be needed and the schedule adjusted for it. That alone could throw the builders’ tidy plans off track. Furthermore, the people involved had to somehow determine whether the tilting indicated a serious construction defect. I was curious to know how they handled this question, for there was inevitable uncertainty. How could they know that the problem was just ordinary settling, that loading the steel frame would in fact level out the floor? As Rouillard acknowledged, “variances can occur.” This was a situation of true complexity.

skyscraper builders must run into thousands like it—difficulties they could never have predicted or addressed in a checklist designed in advance. The medical way of dealing with such problems—with the inevitable nuances of an individual patient case—is to leave them to the expert’s individual judgment. You give the specialist autonomy. In this instance, Rouillard was the specialist. Had the building site been a hospital ward, his personal judgment would hold sway. This approach has a flaw, however, O’Sullivan pointed out. Like a patient, a building involves multiple specialists—the sixteen trades. In the absence of a true Master Builder—a supreme, all-knowing expert with command of all existing knowledge—autonomy is a disaster. It produces only a cacophony of incompatible decisions and overlooked errors. You get a building that doesn’t stand up straight. This sounded to me like medicine at its worst. So what do you do? I asked. That was when O’Sullivan showed me a different piece of paper hanging in his conference room. Pinned to the left-hand wall opposite the construction schedule was another butcher-block-size sheet almost identical in form, except this one, O’Sullivan said, was called a “submittal schedule.” It was also a checklist, but it didn’t specify construction tasks; it specified communication tasks. For the way the project managers dealt with the unexpected and the uncertain was by making sure the experts spoke to one another—on X date regarding Y process. The ...

He showed me an application called Clash Detective that ferreted out every instance in which the different specs conflicted with one another or with building regulations. “If a structural beam is going where a lighting fixture is supposed to hang, the Clash Detective turns that beam a different color on-screen,” he said. “You can turn up hundreds of clashes. I once found two thousand.” But it’s not enough to show the clash on the screen, he explained. You have to resolve it, and to do that you have to make sure the critical people talk. So the computer also flags the issue for the submittal schedule printout and sends an e-mail to each of the parties who have to resolve it. There’s yet another program, called ProjectCenter, that allows anyone who has found a problem—even a frontline worker—to e-mail all the relevant parties, track progress, and make sure a check is added to the schedule to confirm that everyone has talked and resolved the matter. When we were back at the McNamara/Salvia offices, Bernie Rouillard showed me one such e-mail he’d gotten that week. A worker had attached a digital photo of a twelve-foot steel I beam he was bolting in. It hadn’t lined up properly and only two of the four bolts could fit. Was that all right, the worker wanted to know? No, Rouillard wrote back. They worked out a solution together: to weld the beam into place. The e-mail was also automatically sent to the main contractor and anyone else who might potentially be required to sign off....

There is a particularly tantalizing aspect to the building industry’s strategy for getting things right in complex situations: it’s that it gives people power. In response to risk, most authorities tend to centralize power and decision making. That’s usually what checklists are about—dictating instructions to the workers below to ensure they do things the way we want. Indeed, the first building checklist I saw, the construction schedule on the right-hand wall of O’Sullivan’s conference room, was exactly that. It spelled out to the tiniest detail every critical step the tradesmen were expected to follow and when—which is logical if you’re confronted with simple and routine problems; you want the forcing function. But the list on O’Sullivan’s other wall revealed an entirely different philosophy about power and what should happen to it when you’re confronted with complex, nonroutine problems—such as what to do when a difficult, potentially dangerous, and unanticipated anomaly suddenly appears on the fourteenth floor of a thirty-two-story skyscraper under construction. The philosophy is that you push the power of decision making out to the periphery and away from the center. You give people the room to adapt, based on their experience and expertise. All you ask is that they talk to one another and take responsibility. That is what works. The strategy is unexpectedly democratic, and it has become standard nowadays,