Why even experts still need short checklists to save lives

1

Expertise outruns memory

Modern failures happen less because professionals lack knowledge than because no human memory can reliably hold and sequence everything a complex job now requires.

Gawande starts with a near-miss in his own OR: the team had the skill, yet in the rush someone skipped a step everyone technically knew, and the patient nearly died. He treats that moment as typical rather than freakish. Medicine now stacks thousands of drugs, tests, and procedures, each with its own failure points.

The same overload appears in aviation, engineering, and finance. The old competence model assumed enough training and experience would prevent error. Gawande calls that model obsolete once task volume and pace exceed what unaided memory can track, especially under stress.

He reframes safety as an engineering problem. Instead of asking only how to train people harder, the question becomes how to build systems that catch the predictable lapses of otherwise capable people. Forgetting and overlooking stop being moral failures and become design targets.

In practice: Research on intensive care units that he cites shows patients experience an average of close to two errors a day, most of them omissions rather than deliberate mistakes.

2

Aviation’s flying-fortress fix

The checklist was invented to compensate for the predictable ways skilled attention breaks down, not to compensate for stupidity.

In 1935 the U.S. Army Air Corps tested Boeing’s Model 299, nicknamed the flying fortress. On a demonstration flight it crashed after takeoff, killing Major Ployer Hill, one of the Army’s most experienced pilots. Investigation found no mechanical fault: he had simply forgotten to release a new locking mechanism on the elevator and rudder controls.

The plane was judged “too much airplane for one man to fly.” Boeing test pilots responded with a short list of essential steps read aloud and confirmed before takeoff, landing, and in flight. The list did not replace piloting skill; it handled the part skill could not reliably handle—remembering the same critical items every time regardless of mood, fatigue, or distraction.

Once the checklist was standard, the B-17 flew 1.8 million miles without a serious accident. Gawande uses the case to separate expertise from memory: elite performers still benefit from external structure because expertise itself consumes the attention memory then lacks.

In practice: The Model 299 crash investigation traced the loss to one forgotten unlock step on unfamiliar new controls, not to a lack of flying skill.

3

Simple, complicated, complex

Checklists fit complicated problems with many known coordinated steps, not simple problems that need no help or complex ones that resist standardization.

Gawande borrows Brenda Zimmerman’s three-way split. Baking a cake from a recipe is simple: follow the steps once and the outcome is reliable. Sending a rocket to the moon is complicated: enormous parts and specialists, yet once solved it can be replicated through careful coordination. Raising a child is complex: each case differs, past success guarantees nothing, and no list substitutes for judgment.

The distinction tells the reader where to expect leverage. Surgery, he argues, is mostly complicated rather than complex: anatomy and procedure are largely known in advance, and many failures come from skipping a known step, not from an unsolvable judgment call.

A checklist for anesthesia induction or central-line placement therefore addresses the complicated layer. It frees attention for the genuinely complex decisions, such as how to respond to an unexpected bleed. Critics who say medicine is too individualized miss that most hospital errors are execution failures on well-understood steps.

In practice: Gawande notes that a checklist cannot raise a child, yet it can standardize the known sequence of central-line insertion so judgment is saved for surprises.

4

Skyscrapers run on distributed lists

Large unpredictable projects succeed less through central command than through a decentralized web of checklists that let specialists coordinate without one controlling authority.

Gawande visits a Boston construction site and talks with structural engineer Joe Salvia about how buildings of extreme complexity rise without collapsing despite hundreds of subcontractors, unpredictable soil, and constant design changes. The industry once suffered collapses at rates comparable to surgical disasters.

The response was not smarter lone engineers but elaborate checklist systems: submittal schedules tracking every specification, and communication protocols spelling out who must sign off before work proceeds. No single person holds all the information; the general contractor does not personally verify every welded joint.

Distributed lists force each specialist to confirm their own piece and flag problems upward. Gawande calls the result a discipline of teamwork rather than a hierarchy of command. The lists are targeted prompts for easily forgotten critical steps, not exhaustive manuals, leaving room for specialist judgment on what cannot be anticipated. Historically, structural failures fell even as buildings grew taller and more complex.

In practice: On the Boston site, Salvia’s world runs on submittal schedules and sign-offs so hundreds of trades stay aligned without one mind holding every detail.

5

Five steps beat trained habit

A short simple checklist can outperform even highly trained specialists because expertise alone does not guarantee consistency, and consistency is where lives are lost.

Peter Pronovost, an anesthesiologist and critical-care specialist at Johns Hopkins, faced bloodstream infections from central-line catheters in the early 2000s. He wrote five steps doctors already knew—hand washing, skin cleaning with antiseptic, and the rest—and asked nurses to observe and stop doctors who skipped any step.

The politics were awkward: a nurse was authorized to interrupt a physician mid-procedure. After a year, the ten-day line-infection rate had dropped from eleven percent to zero. Pronovost then took the list statewide in Michigan through the Keystone Initiative, covering more than one hundred intensive care units.

Within eighteen months the initiative was credited with preventing an estimated 1,500 deaths and saving roughly 175 million dollars in costs tied to preventable infections. The list taught nothing new and needed no new technology; it only made a known standard happen every time. Many physicians first resisted it as an insult to training—the recurring obstacle of professional pride.

In practice: At Hopkins the ten-day central-line infection rate fell from eleven percent to zero once nurses could halt a skipped step.

6

Nineteen items, eight countries

A nineteen-item surgical checklist produced complication and death reductions comparable to a major new drug, but only where teams genuinely adopted it rather than merely tolerated it.

Building on Pronovost, the WHO recruited Gawande to lead Safe Surgery Saves Lives and design a list usable from a rural Tanzania clinic to a Toronto teaching hospital. The team drafted, tested, and revised nineteen items across three pause points: before anesthesia, before incision, and before the patient leaves the OR—confirming identity, antibiotic timing, instrument counts, and more.

Piloted in eight hospitals across eight countries in 2007 and 2008, the results published in the New England Journal of Medicine in 2009 showed major complications falling from about 11 percent to 7 percent and deaths roughly halving from about 1.5 percent to 0.8 percent across more than 7,600 patients.

Almost every item was something teams already claimed to know. The gain came from consistent execution and from forcing a brief conversation among surgeon, anesthesiologist, and nurses before cutting—often the first time some had spoken by name. Strongest gains tracked genuine embrace, not mere posting of the form.

In practice: Across the eight-country pilot, death rates fell from about 1.5 percent to 0.8 percent while the list also made teams introduce themselves by name and role.

7

Brevity is the design rule

A checklist’s power comes from ruthless brevity and precise design, not comprehensiveness; a poorly designed list can be worse than none.

Gawande draws on Boeing’s process and aviation safety experts for concrete rules. Good checklists run five to nine items, aimed only at steps most likely to be forgotten or most catastrophic if skipped. They use plain language, fit on one page, and ideally take under ninety seconds.

Two formats matter: READ-DO lists followed step by step like a recipe, and DO-CONFIRM lists that let a team work from memory then pause to verify. The surgical list used the latter. Killer checklists, by contrast, are bloated multipage documents that try to cover every scenario; staff learn to skim them, producing empty box-ticking.

Boeing tests drafts on real pilots in simulators and revises wherever people stumble or skip, rather than designing in a conference room and imposing from above. Gawande treats good checklist writing as hard specialized work closer to a short poem than a manual—field-tested and willing to cut items that only seem important.

In practice: Boeing’s method rejects conference-room lists in favor of repeated simulator trials that expose where pilots actually stumble.

8

Silence defeats the list

A checklist fails without a culture that lets the most junior person interrupt the most senior, because most errors first surface as small unspoken doubts.

Even a well-designed list is only as strong as the willingness to use it. That willingness requires flattening hierarchy enough that a nurse can question a surgeon, or a co-pilot a captain, without fear. Gawande points to aviation crashes traced to excessive cockpit authority gradient.

The 1978 United Airlines flight that ran out of fuel over Portland while the captain, fixed on a landing-gear problem, ignored crew fuel warnings became a turning point. Crew Resource Management training followed industry-wide, teaching juniors to speak up and seniors to invite it.

Operating rooms showed the same pattern: nurses noticed missing consent forms or wrong-site marks yet stayed silent out of deference. The WHO list’s opening briefing—each person states name and role—exists to break that silence and give everyone standing. Hospitals with the largest trial gains were those where surgeons visibly took the list seriously and modeled speaking up, not those that only mandated paperwork.

In practice: After the 1978 Portland fuel-exhaustion crash, aviation made junior challenge and senior invitation standard training across crews.

9

The tool leaves the hospital

The checklist is a general-purpose tool for extending human capability under complexity, not a medical technology.

Closing chapters widen the lens. At the Cheesecake Factory, standardized preparation checklists let relatively inexperienced kitchen staff produce a large complex menu consistently across hundreds of locations—reliability from procedure rather than culinary genius.

Investors and fund managers describe personal checklists that guard against predictable psychological errors, such as chasing a rising stock out of excitement instead of analysis. The list defends them against their own overconfidence rather than supplying new information.

Disaster response after Hurricane Katrina in 2005, Gawande notes, was hampered less by missing resources than by absent coordination protocols—who communicates what to whom. The pattern matches construction’s earlier lesson. He closes by urging readers in any field to build and test a checklist for a frequent task, treating continued refusal as avoidable professional arrogance.

In practice: Cheesecake Factory kitchens use prep checklists so inexperienced staff can still hit consistent quality across a long menu and many sites.