CTA: In your team’s multivariate analysis, you found that speaker and introducer gender per se is driving this trend, correct?
Dr Duma: Yes. The ratio of being introduced by first name for women [vs men] was close to 6 — that is, 6 times more likely. When we control for speaker specialty, the year, their degree, [whether they] are they from the United States or not, the only thing that really mattered — the highest hazard ratio for first name introductions — was just being a woman.
CTA: Your team statistically controlled for nationality in the multivariate analysis. Did you also look at that to see if that is an independent correlate of gender bias in introductions?
Dr Duma: More than a speaker’s nationality, we accounted [for] where the physician is practicing, because it’s very hard to find out speakers’ nationalities by name. I have my husband’s last name. My first name is very unusual. So yes, we did the multivariate analysis and we noticed that people from European institutions are actually less likely to be introduced by first name. They are most likely to be introduced as professors.
CTA: What about race and ethnicity effects in the context of gender bias and speaker introductions? Was it possible to look at ethnicity effects?
Dr Duma: The initial data presented at ASCO were the results of our coders assigning the speakers a race and ethnicity based on their name, features, place of training, and practice. This process was subject to bias itself, so we decided to omit it from the manuscript and we are working on improving this part of the project with the hope of publishing results about the racial and ethnic background of the speakers in the near future.
CTA: Are there other additional variables you’re also going to include in the 2020 follow-up study?
Dr Duma: Yes, age. We would like to account for the ages of the speakers and introducers. But this information was not widely available at the time of data collection.