The presumption of human error

Naturally, we respect and admire doctors. We believe that health care is scientific. We think of hospitals as places of safety. For all these reasons, it comes as something of a shock to realise that errors still play such a significant role in whether we leave a hospital better or worse, alive or dead.

The National Audit Office estimates that there may be 34,000 deaths annually as a result of patient safety incidents. When he was medical director, Liam Donaldson warned that the chances of dying as a result of a clinical error in hospital are 33,000 times higher than dying in an air crash. This isn’t a problem peculiar to our health-care system. In the United States, errors are estimated to be the third most common cause of deaths in health care, after cancer and heart disease. Globally, there is a one-in-ten chance that, owing to preventable mistakes or oversights, a patient will leave a hospital in a worse state than when she entered it.

There are other industries where mistakes carry grave consequences, but the mistakes of doctors carry a particular moral charge because their job is to make us better, and we place infinite trust in the expectation they
will do so. When you think about it, it’s extraordinary we’re prepared to give a virtual stranger permission to cut us open with a knife and rearrange our insides as we sleep.

Perhaps because of the almost superstitious faith we need to place in surgeons, we hate to think of them as fallible; to think that they perform worse when they are tired, or that some are much better at the job than others, or that hands can slip because of nerves, or that bad decisions get taken because of overconfidence, or stress, or poor communication. But all of these things happen, because doctors are human.

What the medical profession can learn from the airline industry about how to protect against human error. A riveting story about how we must build around the assumption of inevitable human error.

The passage above mirrors my own journey towards realizing that not all doctors are infallible. It may seem silly now, but as a child, I had a learned reverence of the medical profession. All the training, all the accreditation, the requirement to address them by a title all their own—“Doctor”—was blinding.

In 1997 I was back in the Bay Area and went to play pickup basketball with some old Stanford classmates at a local gym. A scrimmage game against a group of older Greeks turned heated, as such games are wont to be with the physical release of so much pent up testosterone. On one drive to the basket, I took a hard shove and went flying sideways. I landed on my left food and my left knee flew sideways an opponent's leg that was planted on the ground.

I felt a searing pain immediately and collapsed. Some teammates carried me to the sideline, and my knee immediately started swelling. I'd never felt anything like this before. Something had happened, but I didn't know what.

I crutched my way back to Seattle, stayed on crutches for a few days, and eventually got in to see an ortho. He laid me down, tugged on my leg a bit this way and that, moved my leg around, and gave me a comforting diagnosis. It was a mild sprain, I could resume light physical activity after the swelling subsided.

At the gym, on an elliptical trainer, something didn't feel right. But the doctor had done some tests, who was I to question him? The web existed, but it was much sparser than it is now. WebMD and sites like that didn't exist.

Back then, Amazon's unofficial official company sport was broomball. The popular stereotypes of technology companies being populated with a bunch of meek, gaunt, sun-deprived software developers neglect the army of MBA's with their world-conquering ambitions, the ex college jocks in business development, the crazy endurance athletes whose motor played on the field or in the office. We played at company functions, and the games felt like some form of trial by combat.

My knee still felt off, but I wasn't about to miss out on our team's broomball contest. We played on a muddy field, it was like some form of field hockey minus pads. You just had to accept that you'd leave battered, your shins a mess of bruises. We taped tennis balls to the ends of our broomsticks so as not to take out anyone's eyes.

From the start I couldn't move that well, so I hung back to play defense. And then an opponent broke loose, a herd of people chasing him, and I moved to intercept the ball. At best, with all the momentum he'd built up, I hoped to deflect the ball horizontally to give the rest of my team a chance to catch up and reset.

As I moved diagonally to meet the path of the ball, he tried to make a sharp cut, but on the muddy field, he couldn't turn enough, and both ball and opponent came sideways and collided with me.

My left leg experienced what the doctor would later call a pivot shift, where the top and lower leg came out of alignment. I fell to the ground screaming. My day was over, and I don't remember now how I drove myself home considering it was a manual transmission.

I found myself back in that same ortho's office a day later, and I told him something wasn't right, to check me again. This time, he consented to perform an MRI.

When the results came back, he was almost sheepish in sharing the news. Though he'd performed the standard Lachman Test and some other tests the last time I'd come in, in fact I did have a torn ACL. I'd been running around for weeks without my left ACL.

Needless to say, I didn't let that ortho perform my ACL reconstruction.

The first time I visited, was the ortho hesitant to order an MRI because of the expense, because I was on an HMO? Or did he just not perform the Lachman test properly? It still haunts me, but the lasting consequence was the shattering of my belief in the infallibility of doctors. I still have deep respect for the medical profession, my brother and his wife are both doctors whom I turn to again and again for advice, but nothing about medical training magically removes human error from day to day life.


If the severity of Elaine’s condition in those crucial minutes wasn’t registered by the doctors, it was noticed by others in the room. The nurses saw Elaine’s erratic breathing; the blueness of her face; the swings in her blood pressure; the lowness of her oxygen levels and the convulsions of her body. They later said that they had been surprised when the doctors didn’t attempt to gain access to the trachea, but felt unable to broach the subject. Not directly, anyway: one nurse located a tracheotomy set and presented it to the doctors, who didn’t even acknowledge her. Another nurse phoned the intensive-care unit and told them to prepare a bed immediately. When she informed the doctors of her action they looked at her, she said later, as if she was overreacting.

Reading this, you may be incredulous and angry that the doctors could have been so stupid, or so careless. But when the person closest to this event, Martin Bromiley, read Harmer’s report, he responded very differently. His main sensation wasn’t shock, or fury. It was recognition.

RELATED: Atul Gawande's great book A Checklist Manifesto.

Asiana Flight 214

Patrick Smith of Ask the Pilot  fame provided an informed critique of many of the hasty judgments people grasped at wildly after Asiana Flight 214 crashed at SFO. Among those:

Lastly, we're hearing murmurs already about the fact that Asiana Airlines hails from Korea, a country with a checkered past when it comes to air safety. Let's nip this storyline in the bud. In the 1980s and 1990s, that country's largest carrier, Korean Air, suffered a spate of fatal accidents, culminating with the crash of Flight 801 in Guam in 1997. The airline was faulted for poor training standards and a rigid, authoritarian cockpit culture. The carrier was ostracized by many in the global aviation community, including its airline code-share partners. But Korean aviation is very different today, following a systemic and very expensive overhaul of the nation’s civil aviation system. A 2008 assessment by ICAO, the civil aviation branch of the United Nations, ranked Korea's aviation safety standards, including its pilot training standards, as nothing less than the highest in the world, beating out more than 100 other countries. As they should be, Koreans are immensely proud of this turnaround, and Asiana Airlines, the nation's No. 2 carrier, had maintained an impeccable record of both customer satisfaction and safety.

Whatever happened on final approach into SFO, I highly doubt that it was anything related to the culture of Korean air safety in 2013. Plane crashes are increasingly rare the world over. But they will continue to happen from time to time, and no airline or country is 100 percent immune. 

Malcolm Gladwell arguably did more than anyone to popularize the theory that Korean culture lay at the root of Korean Air's poor safety record in the 1980's and 1990's. In this interview with CNN Money, he summarized his theory from his book Outliers:

F: You share a fascinating story about culture and airline safety.

G: Korean Air had more plane crashes than almost any other airline in the world for a period at the end of the 1990s. When we think of airline crashes, we think, Oh, they must have had old planes. They must have had badly trained pilots. No. What they were struggling with was a cultural legacy, that Korean culture is hierarchical. You are obliged to be deferential toward your elders and superiors in a way that would be unimaginable in the U.S.

But Boeing (BA, Fortune 500) and Airbus design modern, complex airplanes to be flown by two equals. That works beautifully in low-power-distance cultures [like the U.S., where hierarchies aren't as relevant]. But in cultures that have high power distance, it's very difficult.

I use the case study of a very famous plane crash in Guam of Korean Air. They're flying along, and they run into a little bit of trouble, the weather's bad. The pilot makes an error, and the co-pilot doesn't correct him. But once Korean Air figured out that their problem was cultural, they fixed it.

A fairly thorough rebuttal to Gladwell's theory was posted at the blog Ask a Korean:

First, the way in which Gladwell quoted the transcript is severely misleading. This is the full transcript, which goes from pp. 185 to 187 of the NTSB report:

CAPTAIN: 어... 정말로... 졸려서... (불분명) [eh... really... sleepy... (unintelligible words)]
FIRST OFFICER: 그럼요 [Of course]
FIRST OFFICER: 괌이 안 좋네요 기장님 [Captain, Guam condition is no good]
FIRST OFFICER: Two nine eighty-six
CAPTAIN: 야! 비가 많이 온다 [Uh, it rains a lot]
CAPTAIN: (unintelligible words)
CAPTAIN: 가다가 이쯤에서 한 20 마일 요청해 [Request twenty miles deviation later on]
CAPTAIN: ... 내려가면서 좌측으로 [... to the left as we are descending]
(UNCLEAR SPEAKER): (chuckling, unintelligible words)
FIRST OFFICER: 더 오는 것같죠? 이 안에. [Don't you think it rains more? In this area, here?]

(emphases mine)

Note the difference between the full transcript, and the way Gladwell presented the transcript. Gladwell only quoted the first two lines and the last line of this sequence, omitting many critical lines in the process. In doing so, Gladwell wants to create an impression that the first officer underwent some period of silent contemplation, and decided to warn the captain of the poor weather conditions in an indirect, suggestive manner. 

The full transcript reveals that this is clearly not the case. The first officer spoke up directly, clearly, and unmistakably:  "Captain, Guam condition is no good." It is difficult to imagine how a person could be more direct about the poor weather condition. Further, there was no silent contemplation by the first officer. Nearly three minutes elapse during this sequence, during the captain and the first officer chatted constantly. And it is the captain who first brings up the fact that it is raining a great deal: "Uh, it rains a lot." In this context, it is clear that the first officer is engaged in some friendly banter about the rain, not some indirect, ominous warning about the flight conditions.

To be fair to Gladwell, when asked about whether he thought his theory from Outliers came into play in the case of Asiana Flight 214, he did not bite

We asked Malcolm Gladwell for his thoughts on the use of his essay in the particular context of the Asiana crash. "I can understand why my Outliers chapter has been of interest, given how central cockpit communication issues are in plane crashes," Gladwell told The Atlantic Wire in an email, adding, "My sense is that we should wait for the full report on the crash before drawing any conclusions about its cause." As for the applicability of his work to the recent Asiana crash, Gladwell noted that his essay was specific to the problems (and solutions) of one airline — Korean Air, "which I think did an extraordinary job of addressing the cultural issues involved in pilot communication. This was a crash involving a completely different airline," he said. 

The NTSB has yet to issue any formal assessment of what happened that day. I happened to arrive at SFO for a flight to Paris just a few hours after the crash occurred, and we could see the wreckage in the distance from our gate. Our flight was delayed by 9 hours, and we finally took off at around 9pm that night.

Our plane turned onto the runway and accelerated towards takeoff. The moment just before our wheels left the ground, I saw, just out the window to the left, the wreckage of Asiana Flight 214, sitting just off our runway, illuminated by some giant spotlights, like a giant burned out metallic skeleton that had just been dug out of the earth.