Cognitive biases are unconscious. They change our understanding of events. One example is anchoring, where people naturally place weight on the first information they receive, even if order is not important. Therefore, if you see a car priced at $20,000 and then later see one for $10,000, you may have the natural tendency to assume that $20,000 is a reasonable price and $10,000 is cheap. If you saw the cars in reverse order, you might tend to assume that $10,000 is a reasonable price and $20,000 is expensive.
Doctors are human, and they are subject to cognitive biases like other people. Unfortunately, when their biases translate into medical decisions, those decisions may not be appropriate.
Consider what psychologists call “availability bias.” In one study, doctors changed their prescribing behavior. When one patient experienced a rare side effect of a drug, doctors were less likely to prescribe that drug in other patients even though the side effect was uncommon enough that it would not be expected in many patients.
In another study, obstetricians changed their delivery recommendations. When one of their patients experienced an adverse event while giving birth, most obstetricians tended to change delivery methods for the next patient regardless of its objective appropriateness. In this study, the change of delivery method was associated with worse outcomes.
Recently, a new study was published in the New England Journal of Medicine on a cognitive bias known as “left digit bias.” This is the same bias that explains why people buy more when the price is $4.99 than when it is $5.00. People naturally round down to the left-most digit.
It turns out that doctors round to the left-most digit, too, when considering treatments in light of a person’s age. For example, doctors tended to recommend coronary artery bypass surgery significantly more often for people who were a few weeks shy of their 80th birthday than for those who had just passed theirs, even though they are objectively the same age. Presumably, the doctors perceived the pre-birthday patients as “in their 70s” and the post-birthday patients as “in their 80s.”
What that meant for patients in the study was that people who were objectively only a few months younger were given the more aggressive treatment — and they were less likely to die within 30 days. Those just past their 80th birthdays were left out of an effective treatment and had worse outcomes.
What can be done to prevent cognitive biases?
The crucial way to prevent cognitive biases from affecting decision-making is to become aware of potential biases and fight them. That requires education and dedication, but it would be well worth it if we could improve outcomes in a clinical setting.
Another way to deal with unconscious cognitive biases is to create systems and structures that work against them. For example, one study found that changing the default options on an electronic medical record resulted in doctors prescribing fewer opioids.
Ultimately, doctors are expected to make medical decisions in their patients’ best interest and treat them to at least a minimum standard of expected care. If a cognitive bias interferes with doing that, it could lead to medical malpractice.