Attractiveness likely influences selection for residency interviews, according to researchers who used mock radiology residency applications to test for potential bias. "The facially attractive and nonobese applicants had twice the chance to get an interview," lead author Charles M. Maxfield, MD, vice chair of education in the Department of Radiology at Duke University in Durham, North Carolina, told Medscape Medical News. Less attractive, nonobese applicants in the study by Maxfield and colleagues had a 12% chance for an interview, whereas those who were both facially attractive and nonobese had a 24% chance. "If you are obese and unattractive, you had only a 10% chance of an interview," he said. That is particularly significant since choosing those who get interviews is already highly selective, with only 15% or so of all applicants moving on to the interview stage, Maxfield said. Maxfield and colleagues published their findings online May 28 in Academic Medicine. Maxfield noted that radiology was studied only because it was the field he is in, but said, "I am certain this is not a radiology-specific finding. I don't even think it's a medicine thing. I think this is about higher education in general." How Attractiveness Was Determined Researchers pulled from the Internet 170 open access/stock photos (head and shoulders view of someone dressed professionally) of men and women that represented various levels of facial attractiveness and obesity and a mix of races. They asked eight radiologists to each rate the photos for attractiveness (1-5 scale) and obesity levels (1-3 scale). Panel members were chosen to reflect the demographic distribution of the core radiology faculty at the five centers (four men, four women; five white, two Asian, one African American; age range 26-56 years). The academic centers included University of Indiana in Indianapolis; Mayo Clinic in Rochester, Minnesota; University of New Mexico in Albuquerque; Duke University in Durham, North Carolina; and Stanford University in California. The researchers discarded photos if there was large disagreement among panel member ratings. They then attached 76 photos to fixed identities that were inserted into abbreviated versions of applications that would typically be seen on the Electronic Residency Application Service (ERAS). The researchers randomly combined these fixed identities with academic variables to determine the effect of attractiveness and obesity on the likelihood of selection. "We randomized academic variables important in the selection of radiology applicants for interview (preclinical class rank, clinical clerkship grades, Alpha Omega Alpha (AOA) Honor Society membership, and quantity of research publications) for each application and reviewer, such that each reviewer saw a different combination of academic variables associated with any given photograph," the authors explain. The applications did not contain information such as medical school performance evaluations, deans' letters, personal statements, United States Medical Licensing Examination (USMLE) Step 2 scores, or additional advanced degrees. The reviewers were told that the applications were abbreviated for efficient review, but still had enough information for screening. The reviewers accessed the mock profiles through a website designed to look like the ERAS website. Applications Reviewed by 74 Core Faculty The 5447 mock applications were reviewed by 74 core faculty in September and October 2017. Volunteers, who were experienced in evaluating residency applications, were recruited to review the applications under the guise that they would be screening actual applicants and would be helping the department. Each of the volunteer reviewers evaluated an average 74 applications and gave them a 1 to 5 score on interview desirability. The researchers then compared reviewer scores and application variables using linear mixed fixed and random effects models. The strongest predictor of whether an applicant was chosen was how the applicant scored on the USMLE Step 1 (B = 0.35 [standard error (SE) = 0.029]). Notably, the second most influential factor was race (B = 0.25 [SE = 0.059]). Reviewers gave preference to black and Hispanic candidates, which demonstrates their thoughtful approach to the process, Maxfield said. "We gave no instructions to be race conscious," Maxfield said. The third most influential factor was facial attractiveness (attractive vs unattractive, B = 0.30 [SE = 0.056]; neutral vs unattractive, B = 0.13 [SE = 0.028]). "Facial attractiveness was more important than any of the other individual academic metrics [aside from USMLE Step 1 score], which includes your class rank and your third-year clerkship scores," Maxfield said. Attractiveness More Important Than AOA Reid A. Waldman, MD, with the Department of Dermatology at the University of Connecticut Health Center in Farmington, told Medscape Medical News that when he was going through the match he was told that "AOA [Alpha Omega Alpha honor society] was make or break on whether you matched." After reviewing this article, which he called "phenomenal," he said, "It sounds like people should have been saying to me 'you need to get more attractive and leaner.' "If this is more significant than AOA — generally considered to be the largest academic honor that a medical student can receive — that is enough that should give anyone pause, and it's something we need to be talking about," he said. As previously reported, Waldman warned of this kind of bias in a letter to the editor published last November in Academic Medicine, in which he advocated for getting rid of the photos. Waldman said the study, done in the specialty of radiology, which is less patient-facing and has no cosmetic component, may understate the problem. The bias, he said, may be even more pronounced in fields such as dermatology. He praised the study methods for so closely mirroring the match process —with two exceptions. One was that the reviewers in this study were instructed not to discuss the candidates with colleagues, which they would do in real situations and that might have an effect on bias. The second is that, as the authors state, letters of recommendation and personal statements were not included, and those do carry weight in the process, he said. Why Use Photos? Waldman said the only reason he has ever heard for using the photos is for identification purposes and for remembering the candidate at the end of a long day of interviewing. "I call B.S. on that," he said, adding that he has yet to hear a valid reason for using them for screening. He noted that the US Equal Employment Opportunity Commission says you can't legally require photos before the interview. "I do think once people show up, it's reasonable to get a photo of the applicant because we do see 35 people in the same day all wearing suits that look pretty identical." But as for the early process, he said, "I think it's time for photographs to be eliminated." "I do think we have a moral obligation as educators and as healthcare providers to take the steps in our power to do what's right," he said. Maxfield takes a different view. He said there has been some talk of eliminating photos from ERAS, but he says he doesn't advocate for that because the bias is just as likely to surface in the interview stage. "I just think we need to be aware the bias is there," he said. The authors and Waldman have disclosed no relevant financial relationships. Academic Medicine. Published online May 28, 2019. Abstract Source
