Why physicians hate to admit their errors, even to themselves

By Danielle Ofri
Tuesday, August 3, 2010

Precisely two weeks after completing my medical internship, I proceeded to nearly kill a patient. July marked the start of my second year of residency at New York City's Bellevue Hospital, and it was my first time being fully in charge of a patient. He arrived in fully developed diabetic ketoacidosis (DKA), a life-threatening condition in which lack of insulin causes a metabolic cataclysm. It was a classic Bellevue DKA patient story: arrested during a small-time drug deal, tossed into a holding cell, unable to access insulin. The patient sat in the cell as his sugar soared to stratospheric levels. When he began to vomit and his speech slurred, the police brought him to our emergency room.

While a new intern looked to me for guidance -- me with a scant single year more experience -- I placed the patient on an intravenous insulin drip.

DKA is one of those rare, gratifying conditions in which a patient arrives in extremis and, with deft handling of insulin, can be readily "cured." I felt a surge of pride as we watched over the course of eight hours as our patient gained consciousness, got cranky, demanded double portions of food.

We were in a cramped, dingy corner of the ER, stuffed next to a narrow desk, while paramedics wheeled in survivors of motor vehicle crashes and patients with shotgun wounds.

After our DKA patient's glucose returned to normal, I handed a triumphant "d/c insulin drip" order to discontinue the intravenous insulin. I was officially declaring our patient cured.

The nurse took the order from me. "Do you want to give an injection of long-acting insulin before stopping the insulin drip?" she asked, as a clerk pressed two more charts in her direction.

I thought for a moment. Why would I want to use the sledgehammer of long-acting insulin after eight hours of our meticulous adjustments with the insulin drip? "No," I said, turning to my intern, capitalizing on the teaching moment. "If we push him overboard with long-acting insulin, it'll be stuck in his system for hours, and his sugar could bottom out. Let's just keep checking his glucose hourly and give him short-acting insulin as needed."

The nurse raised her eyebrows ever so slightly, then shrugged and went back to her work.

My logic was obvious. It was also wrong. Right-out-of-the-textbook wrong. The very thing you are supposed to do in DKA is administer long-acting insulin just before stopping the insulin drip. Otherwise a patient will turn right around and plunge back into DKA.

When blood tests revealed dangerously rising levels of potassium and acid in my patient, I panicked and paged the senior medical resident for help.

'What were you thinking?'

My intern and I stood nervously while the senior resident scrutinized the numbers. She shot me a withering look. "Didn't you give him long-acting insulin before you turned off the insulin drip?" she demanded. "A little longer like this and he'll be comatose! Next thing you know we'll be calling a code for his cardiac arrest."

I tried to describe how logic would dictate -- wouldn't it? -- that we shouldn't muck up a tenuous situation with long-acting medications, that we wouldn't want to harm the patient by pushing his sugar too low, that we . . . . My words began to run up against one another, progressively garbling under the weight of her granite stare.

"What were you thinking?" the senior resident asked, her voice now like a drill sergeant's.

I stood there stone still as my brain cells slowly dissolved into muck.

"What were you thinking?" she repeated, her voice thundering through the ER.

What had I been thinking? Had I simply forgotten the part about the long-acting insulin? Had I misread the textbook? Was I simply not smart enough to be a doctor?

In the presence of my intern the humiliation was unbearable.

The senior resident wrenched the pen from my hand and furiously wrote orders to restart the insulin drip immediately and to administer long-acting insulin along with calcium and bicarbonate to avoid cardiac arrest from dangerously skewed potassium and acid levels.

When she finally departed, all I wanted to do was to crawl under a rock and weep. But there was an intern waiting for guidance and a patient who needed medical care.

"Let's, uh, check the patient's fluid status," I stammered, "and then draw another set of labs."

"Got it," the intern said. Matter-of-factly he began tearing open gauze pads, labeling test tubes. It was the normalcy of his actions that allowed me to breathe again. That normalcy was an act of compassion that I've never forgotten. We returned to our patient. In two days he was back with the cops.

The senior resident graduated and went off to another job. The intern became a fine doctor in private practice. I continued an academic career at Bellevue. Since that day, I have never failed to inject long-acting insulin before stopping an insulin drip.

Lesson learned. Doctor reeducated.

Near miss caught in time by the system of having more-experienced doctors supervise less-experienced doctors.

Case closed. Or should it have been?

Shattering effects

Certainly in those days, about 20 years ago, that's how such near misses and errors were treated. If this happened today, there might be a different ending to the story. The patient would be approached by the medical team and possibly by risk-management personnel, informed of a medical error that had been potentially life-threatening, given an apology and told that the hospital and physician accepted full responsibility.

It seems entirely obvious: Doctors need to apologize for their errors, even if the patient didn't suffer irreparable harm. But in the real world of medicine, acknowledging responsibility is a dicey proposition. To most physicians it's tantamount to handing your head to a lawyer on a surgical tray. This fear of lawsuits is so potent that even the most ethical physicians want to clam up when issues of medical error arise.

Apologies can often be considered evidence of fault. Depending on the state, there are some protections for apologies that express sympathy, but there is variation as to whether this extends to apologies that invoke responsibility for an error. Typically, the most that doctors can muster is, "I'm so sorry this happened." Such pseudo-apologies are taken to task by Aaron Lazare in his fascinating book "On Apology" (Oxford University Press, 2004). They are meaningless, he writes, because they lack the key ingredient of acknowledging responsibility.

Lazare notes three emotions that influence the decision to apologize: empathy, guilt and shame. But then he distinguishes between guilt and shame. Guilt is usually associated with a particular incident, but shame reflects a failure of one's entire being. Although guilt often prods a person to make amends, shame induces a desire to hide.

Shame, Lazare writes, is "an emotional reaction to the experience of failing to live up to one's image of oneself." Here, I believe, he puts his finger on the precise fiber of resistance in doctors.

When I think back to that moment in the ER, it was shame that overpowered me. Of course I felt guilty -- that was the easy part. But the shame was paralyzing. It was the shame of realizing that I wasn't who I thought I was. Up until that moment, I'd thought I was a competent, even excellent, doctor. In one crashing moment, that persona shattered to bits.

Addressing shame

One could argue that this is a self-centered way of viewing the episode: how the doctor felt. But it's precisely the doctor's emotions that stand as the major impediment to the full-disclosure policies that are increasingly demanded.

One has to wonder why we doctors feel our entire sense of self at risk when we admit error. Perhaps the culture of perfection in medicine fosters a strictly binary analysis: You're either an excellent doctor or a failure. In most other aspects of life, we seem to be able to accept the notion of "good enough." But there is no room for the good-enough doctor.

Shame worms its way into the heart and is remembered like few other things. It seems so out of the realm of medical education, fodder only for those on the couch with their analyst. But it is the elephant in the room.

No doctor will easily confess to error when a core sense of self is at risk. It's difficult to develop policy that addresses such a murky and uncomfortable issue. But it wouldn't hurt for the senior faculty to talk publicly to trainees about their own errors and to address how they dealt with the shame. The very fact of these doctors' continuing to be doctors, despite their errors and the attendant assaults on their egos, would itself be a potent lesson.

Policy efforts have been directed toward proactive settlements with patients, and there is evidence that lawsuits are declining. But even if these programs can coax physicians to come forward, the gut instinct to hide an error and defend the inner self will always be the first lynx to pounce on the heart.

Ofri (http://www.danielleofri.com) is an attending physician at Bellevue Hospital and the editor in chief of the Bellevue Literary Review. This essay, excerpted from August's edition of Health Affairs, is available at http://content.healthaffairs.org/cgi/content/full/29/8/1549.

© 2010 The Washington Post Company