Everybody makes mistakes. But far too few people take the opportunity to learn from them. We'd all be better people if we did. And for doctors, acknowledging errors could mean the difference between a patient's life or death.
In a study where doctors were faced with a simulated medical emergency and had to choose from uncertain treatment options, a scenario requiring a certain amount of trial and error, doctors who paid more attention to their mistakes fared much better than those who focused on their treatment successes.
This is a message everybody could benefit from.
The doctors were presented with a situation where a patient had suffered a heart attack and given a simplified six-factor patient history. They then had to choose from two emergency drug treatments.
The key to choosing the right drug was whether the patient had diabetes. One drug had a 75 percent success rate in patients with diabetes and a 25 percent success rate in all other patients. The second drug gave completely opposite results. But the doctors didn't know this. They had to figure it out from seeing which of their virtual patients lived and which died.
They were given 64 computer simulated trials to learn from, with 10 seconds to select a treatment option for each patient. Results of their treatment were described as either success or failure. Then the same exercise was done as a test: 64 new patients, with no information given to the doctors on whether the treatment succeeded or failed.
Their success rates, their own written accounts, and brain scans taken while the doctors were going through this exercise all provided some very interesting results.
About one-quarter of the doctors eventually found the proper treatment pattern, choosing correctly between 77 and 98 percent of the time. The other doctors did much worse. Based on fMRI scans, which measure changes in blood flow to different parts of the brain, doctors who performed well tended to show high activity in their frontal lobe when treatments failed, while the doctors who performed poorly tended to activate the frontal lobe when a treatment was successful.
The researchers interpret this as meaning that the better performing doctors were focusing more on failed treatments and gradually learning from their mistakes. The poorly performing doctors focused more on successful treatments, with each success confirming what the doctors (falsely) thought they knew about which treatment was better. The researchers term this sort of behavior "success chasing." It's what happens when people ignore their mistakes.
The doctors' own written accounts give other clues about what was going through their heads as they tried to figure out why these drugs sometimes worked and other times didn't. Choosing the right drug was a puzzle. As they treated more patients, doctors altered their treatment strategies based on the results they were seeing. For example, at one point in the trial, one doctor was prescribing drug A for females over 55 and men under 55, and drug B for all other patients. As the trial progressed, the better-performing doctors made small adjustments to their treatment rules following both successes and failures, while low performers made large rule adjustments, and made significantly larger adjustments following successes than failures.
A very striking and unanticipated result was that the more experienced a doctor was, the poorer they performed during the trial.
So what's the take home message from the study? Probably that people can learn as much from their mistakes as their successes but tend to think that their successes are much more meaningful.
What about the fact that less-experienced doctors were better at solving the puzzle? The meaning of that was beyond the scope of the study and could have many explanations. Perhaps younger minds are more facile at solving puzzles. Or maybe older doctors simply are less fond of computer tasks. Nobody knows for sure. It's one more point to consider in the seemingly unanswerable question of whether people are better off with a younger doctor or a more experienced one.