In any other year, Londyn Robinson would have interviewed for her medical residency in person. Instead, an Internet search for tips on how to put her best foot forward online resulted in posting a photo of herself in a bikini top and shorts and setting off an uproar over sexism in science.
With the high-pressure interviews online because of the coronavirus, Robinson, a 26-year-old medical student at the University of Minnesota, went looking for tips on social media etiquette for medical students as part of her preparation.
High in the search results was an article she thought might help. Published in the Journal of Vascular Surgery, it detailed “unprofessional social media content” among young vascular surgeons. But as she read, Robinson said, she became increasingly disgusted.
The authors had studied the social media accounts of surgeons without their knowledge or permission. Among their criteria for “unprofessional” was not only appearing intoxicated or making offensive comments about colleagues or patients but “provocative posing in bikinis/swimwear.”
Robinson took to Twitter, where she expressed her frustration with the mostly male-written paper’s take on professionalism. “I’ll say it: I wear bikinis. I am going to be a doctor,” she tweeted alongside an image of herself wearing a green bikini top and shorts.
She encouraged other medical professionals to share their “professional swimwear” using the hashtag -#MedBikini.
The paper hadn’t generated much comment when it was published online in December 2019.
But #MedBikini garnered thousands of tweets in multiple countries and languages — many including images of doctors and researchers in their own swimwear — and the paper’s retraction a day later on July 24.
It’s one of a recent spate of studies with authors who have been challenged for alleged sexism and other biases — some of which have been retracted or withdrawn. And it’s helping reinvigorate a years-long struggle to make academic publishing more diverse.
Lapses in research aren’t necessarily sexist in nature. But, women researchers say, gender imbalances on research teams and during the publication process can contribute to the publication of biased, flawed research — and even hinder scientific progress.
Women are underrepresented in academic publishing; a 2018 analysis of the authorship of over 10 million academic papers in STEM fields found that it would take 16 years for women to achieve overall publishing parity with men, and up to 258 years in some disciplines like physics.
This gender gap is fueled by everything from inequities in pay, child care, mentorship and advancement to stereotypes about women’s competence in science. Women’s work is judged more harshly during the peer review process. The coronavirus pandemic is expected to make matters worse: One preliminary study found that since January 2020, women have published less than men, a change attributed to an inequitable caregiving load during the stay-at-home period.
Far fewer women are editors of major journals than men, and women are less likely to be invited to be peer reviewers or write the commentary that often accompanies academic articles. As a result, research that applies to women often isn’t read by women who could point out mistakes, biases and design flaws before publication.
“It’s baked into the system,” says Melissa Simon, vice chair for clinical research at the Department of Obstetrics and Gynecology at Northwestern University’s Feinberg School of Medicine and founding director of the Center for Health Equity Transformation at Northwestern. Once flawed research is published, she says, it can have a chilling effect on other scientists and create distrust among patients who might turn down the opportunity to participate in research studies and clinical trials. “Clearly there are papers out there that potentially do harm to advancing science.”
Simon is talking about “Attractiveness of women with rectovaginal endometriosis,” a paper first published in the January 2013 issue of Fertility and Sterility, the highest-ranked journal devoted to reproductive health.
Widely considered the most severe subtype of endometriosis, rectovaginal endometriosis occurs when endometrial tissue grows in both the vagina, the rectum and the area in between. During the Italian-led study, 100 participants with rectovaginal endometriosis who were scheduled for gynecological surgeries were asked about their personal habits, health and sexual histories.
Then their bodies were measured by two physicians. Afterward, a panel of two male and two female physicians rated them on a scale of attractiveness. The results were compared against two control groups — one with other types of endometriosis, others without.
Compared to the controls, more than twice as many women with rectovaginal endometriosis were judged “rather attractive” or “very attractive.” The authors concluded that the women had “a leaner silhouette, larger breasts, and an earlier [age of first sexual intercourse]” than other women studied.
The paper was widely condemned upon publication. Critics asserted that its premise was biased and that its study design was flawed and that it was an example of women being objectified.
Since attractiveness is subjective, there is no way to scientifically measure it, says Simon. “The scientific rigor and the scientific premise are completely faulty,” she adds.
Despite years of criticism, the paper’s authors defended their work. They said that attractiveness could be accompanied by genetic or hormonal markers and that women with “the most feminine phenotype” might be more at risk for endometriosis.
Then, in August 2020, they withdrew their paper and apologized. In a statement provided to The Washington Post, PaolaViganò, Laura Buggio and Giussy Barbara, the three women out of the six-person team of co-authors, expressed surprise at what they characterized as the “harsh” reaction to their work.
“After eight years, we have lost our hopes to convince readers that this study was based on a reasonable biological rationale,” they said. “A constructive debate on the findings never started and, most likely, never will.”
Paolo Vercellini, who led the research, is past president of the World Endometriosis Society. The group’s current president, Neil Johnson, says its board of directors “welcomes” the decision to withdraw the paper.
“We’ve seen this stuff in the past, but there does seem to be a cluster” of papers being challenged for their biases, says Ivan Oransky, vice president at Medscape and editor of Retraction Watch. “We’re being sold a bill of goods that all those checks and balances work.”
He’s referring to the multiple layers of review that any paper must pass on the road to publication. Colleagues and academic advisers approve research before it even begins. So do institutional review boards designed to protect human subjects and ensure ethical research. Academics also present their proposed research to obtain needed funding.
After the article is written, journal editors and peer reviewers weigh in, ask questions and make revisions. Sometimes, papers are also uploaded to preprint servers for additional comments from fellow researchers in the weeks and months before publication.
Both papers went through similar processes. But in the case of the professionalism paper, the Boston University School of Medicine’s institutional review board waived the need to gain informed consent. The researchers gathered the social media account information from a list of graduating vascular surgeon trainees compiled by a vascular surgery association; the surgeons were not told their information would be used by the researchers.
An Institutional Review Board approved the research protocol of the rectovaginal endometriosis paper, and it received funding from the University of Milan School of Medicine. But the authors pressed forward with the study even after 62 potential participants dropped out. About half of those were unwilling to fill out the sexual history portion of the questionnaire; the other half refused to participate in the physical assessment portion of the study, which included measuring the circumference of patients’ breasts.
And though participants consented to the research, they were not informed their attractiveness would be measured. In the paper, the researchers wrote that this was to “limit potential unintentional seductive behaviors that might have swayed the raters’ judgment.”
“Informed consent is a hallmark of good research. If we are not able to [get] informed consent, we will have trouble getting people into our research studies,” says Beth Linas, an infectious-disease epidemiologist in Washington, D.C. Women are underrepresented in clinical trials, especially in areas like cardiovascular health.
Both Linas and Simon say studies that fail to obtain fully informed consent or proceed from biased premises might make it harder for other researchers to recruit a diverse group of potential research participants in the future.
“Unprofessional social media content not only reflects poorly on the individual, but on the medical profession as a whole,” the vascular surgery paper warned. But with the torrent of #MedBikini posts, its authors and the Journal of Vascular Surgery faced a professional social media scandal of their own.
As Twitter timelines filled with images of doctors in bikinis and engaging in conduct the paper’s male data collectors had deemed unprofessional, its authors apologized. The journal’s editors followed.
In a statement on Twitter, they wrote that they believe that the review process didn’t flag “errors in the design of the study with regards to conscious and unconscious bias,” leading to a perception that the published paper “[broadened] inequities in surgery.” Alik Farber, the paper’s corresponding author, did not respond to multiple interview requests.
“It was hurtful, isolating, stigmatizing and frankly disappointing,” says Julie Ann Sosa, chair of the department of surgery at the UCSF School of Medicine and editor of the World Journal of Surgery. “I think you have to look at a peer review process that must be broken.”
Sosa and the editors of 107 surgery journals signed a recent pledge committing to diversity and inclusion throughout the editorial and peer review process.
“The fact that male editors sent this flawed and biased manuscript for peer review, accepted it, and then asked a man to write the invited commentary raises major concerns,” wrote surgeons Melina Kibbe and Julie Freischlag in an article in JAMA Surgery accompanying the pledge. Just 6.7 percent of surgery journals have woman editors.
Sosa hopes that the furor over studies like the one that prompted #MedBikini speeds up equity initiatives in science. “This is ignition for change,” she says. “If we do not remain contemporary as a community, then shame on us.”
“Journals need to wake up to the reality that people are going to read this stuff,” says Oransky. He says that social media can fuel the fire — and that journal editors who are blindsided when work on controversial topics is closely scrutinized should be better prepared.
“I would not put the onus on patients or the public to call this out,” says Simon. “The scientific community needs to call this out.”
As for Robinson, she was gratified by the conversations sparked by the hashtag she created — and hopes the social media moment helps foster equity and spur change. She’ll find out the results of her medical residency applications in March 2021.
“We’re in medicine,” she says. “This is not a Miss America contest.”