Doctors Not Immune to Political Biases When it Comes to COVID

Even physicians can be influenced by their political biases when making decisions about COVID-19 treatments, researchers found.

In a survey with responses from some 400 critical care doctors, those who identified as politically conservative were five times as likely as their liberal and moderate peers to say they would treat a hypothetical COVID-19 patient with hydroxychloroquine, reported Joel Levin, a PhD candidate at the University of Pittsburgh, and colleagues in the Proceedings of the National Academy of Sciences.

“What we’ve shown in this paper is that physicians are human beings like everyone else,” co-author Jeremy Kahn, MD, a critical care physician at the University of Pittsburgh Medical Center, told MedPage Today. “Physicians are not immune to all the biases that affect human decision making. Those biases are innate. They’re part-and-parcel to being human.”

Khan said his experience in the intensive care unit (ICU) during COVID helped inform the study: “It was remarkable to me how many families were coming in, requesting specific treatments that weren’t necessarily supported by the evidence, specifically hydroxychloroquine and ivermectin,” he said.

“The only time I’ve ever been asked about a specific medicine by a patient’s family was during COVID,” he added. “Families never came to me in the ICU and asked if we can prescribe one antibiotic over another. This was a very new experience.”

To get a better sense of whether physicians’ decisions about COVID treatments were affected by political biases, Levin, Khan, and colleagues assessed responses from 410 critical care physicians who were surveyed in three phases from April 2020 to April 2022. They also received responses from 882 laypeople who were surveyed in April 2022.

All participants reported their political ideology on a 7-point scale ranging from “very liberal” to “very conservative.” In each survey, doctors also evaluated a clinical vignette about a severely ill COVID patient and decided on treatment strategy.

Overall, the researchers found that political ideology predicts beliefs about COVID-19 treatments among both laypeople and physicians.

On average, they reported, physicians’ beliefs were less polarized than those of laypeople, and that difference was driven by agreement between liberal and moderate physicians, while conservative physicians displayed polarization that was comparable to that of conservative laypeople, they wrote.

The researchers also conducted an experiment in which they randomized physicians to reading the abstract from the TOGETHER trial — a high-quality randomized controlled trial that showed ivermectin wasn’t effective in COVID-19 — that either identified ivermectin outright or anonymized it as another compound, GL-22. They did the same for laypeople, but using a research summary instead of an abstract.

They found that, overall, responses were more polarized when ivermectin was named compared with when it was anonymized. Those who were more conservative reported that the evidence was less informative, the study was less methodologically rigorous, and the authors were more likely to be biased.

When focusing solely on physicians, the trend remained but didn’t reach statistical significance, Levin said.

These findings highlight the “limits of expertise and exposure to scientific evidence in mitigating polarization,” the researchers wrote.

“It’s easy to have this idea that if only people were smarter, or have more training, had more exposure to relevant information, that their beliefs wouldn’t be polarized,” Levin said. “This paper provides a small but perhaps compelling piece of evidence that that’s not really how it works, and that we’re all susceptible to the same kinds of influences.”

Khan said next steps would be to start thinking about broader interventions that may be able to mitigate the effects of political bias on interpretation of scientific evidence.

“It’s very difficult to think about what kind of intervention is going to make a person a better decision maker when we know that the biases that affect our decision making are really ingrained,” Khan said. “We’re all susceptible to being influenced by factors that we might not think are relevant. If we teach people to be a little more humble in their thinking, then maybe we can overcome some of these biases and come to more of a consensus around best practices.”