Physicians’ political beliefs may impact patient care

Issue: January 25, 2017

When it comes to politicized health issues, whether a physician is a Democrat or Republican may impact how they care for patients.

According to a study published in the Proceedings of the National Academy of Sciences, political bias affects how issues such as drug use, firearm safety and abortion are handled by primary care physicians.

Based on past research showing that such an impact was possible, Eitan D. Hersh, PhD, assistant professor of political science at Yale University, and Matthew N. Goldenberg, MD, MSc, assistant professor of psychiatry at Yale School of Medicine, hypothesized that physicians would react differently to a series of issues presented as clinical vignettes based on their political leanings.

They found that physicians not only assessed the seriousness of politicized health issues in different ways — Democrats worried more about gun ownership; Republicans about drug use and abortion — they also were willing to offer different treatments based on their affiliation.

“Our hypotheses were confirmed. In that sense, we weren’t surprised,” Hersh told Infectious Disease News. “But we were a bit surprised how strong the results were.”

Obtaining party affiliations

Hersh and Goldenberg focused on primary care physicians (PCPs) in the specialties of internal medicine, family medicine, general practice and adult medicine. They oversampled Democratic and Republican PCPs who were in practices together to control for the idea that “physicians with different ideological worldviews might sort into different practices and thus encounter different patient populations.”

To do this, Hersh and Goldenberg linked the records of more than 20,000 PCPs in 29 states to a voter registration database to obtain the physicians’ party affiliations. This also allowed them to avoid having to ask doctors how they were registered.

According to researchers, party affiliation affects how some physicians deal with politicized health issues such as drug use, firearm safety and abortion while caring for patients.

Source: Shutterstock.com

In late 2015, they mailed surveys to 1,529 of these PCPs and invited them to respond by paper or online. (According to Hersh, the study was timed to meet a journal publication date and was not designed so that it would be published close to Election Day.)

They received responses from 20% of the physicians, with Democrats more likely to respond than Republicans, although Hersh and Goldenberg said nothing in the survey indicated it was politically oriented. They did not solicit responses from independent voters, believing that the most apparent partisan differences would occur between Democrats and Republicans.

Despite a small sample size of 231 to 233 physicians, Hersh and Goldenberg said the results of the survey provided “strong and consistent support” for their hypotheses.

“Although we approach small samples with caution, such caution here is balanced against the consistency and strength of the evidence,” they wrote.

The results

In the surveys, PCPs were presented with nine clinical vignettes involving hypothetical patients — a man aged 38 years and a woman aged 28 years — without any known prior chronic medical issues who arrive for their first visit and provide a patient history.

Of the nine vignettes, three involved hot-button issues. The male patient smokes marijuana but denies any related physical concerns. He says he keeps several firearms in a house with two small children. The female patient says she had two elective abortions in the last 5 years but denies any physical complaints or complications and is not currently pregnant.

The physicians were asked to use a 10-point scale to rate both the seriousness of each issue and also how likely they would be to engage in several management options. According to Hersh and Goldenberg, large differences were seen only in the three vignettes involving marijuana, firearms and abortion, where Democrats and Republicans offered “substantially different” ratings, even when the results were stratified by sex, church attendance and patient socioeconomic status. Results for other vignettes, such as those involving alcohol, obesity and helmets, were not skewed based on political leaning.

Republicans were more likely to be concerned than Democrats about the man’s marijuana use and also were more likely to consider the woman’s abortions to be a serious problem, Hersh and Goldenberg said. Democrats were likelier to be concerned about the man keeping firearms in a house with young children.

In general, Hersh and Goldenberg noted, Republicans were more likely than Democrats to say they would engage in active treatment options. The two parties differed in terms of how likely they said they would be to discuss certain options when it came to the politicized health issues.

For instance, Republicans were more likely than Democrats to discuss the health risks and legal implications of marijuana use — much more than for alcohol and cigarettes— and to encourage the women not to have any more abortions. They also were more likely to warn about the legal risks of having intercourse with sex workers. Democrats were more likely to say they would urge patients not to keep guns in the house, but Republicans were likelier to ask about how safely the weapons were stored.

Hersh and Goldenberg said the results show that patients can expect to get different medical care based on the political party affiliation of their PCP. Hersh told Infectious Disease News that there also is evidence from other studies that care varies base on the gender of the provider.

“To be clear,” Hersh and Goldenberg wrote, “we cannot say with any certainty what the root cause of partisan differences in treatment is. However, regardless of the underlying mechanism that affects physician judgement, the evidence suggests a clear effect from the patient’s perspective.” – by Gerard Gallagher

Reference:

Hersh ED and Goldenberg MN. Proc Natl Acad Sci USA. 2016;doi:10.1073/pnas.1606609113.

Disclosures: The researchers report no relevant financial disclosures.