Actual and perceived gender differences in virtual tumor board participation

Introduction: Participant characteristics are known to affect group discourse and discussion outcomes. In medicine, many decisions are made by group consensus, therefore an understanding of these factors is highly relevant. We aimed to measure the effects of participant characteristics on tumor board discussions. Methods: We performed a prospective, multi-institution, quantitative study of multi-disciplinary virtual tumor board meetings. Participant characteristics included age, gender, and clinical discipline. Outcomes of interest were speech events, duration, and discourse style. Participant impressions was assessed by a post-hoc survey. Results: A total of 361 cases were discussed across 32 virtual meetings. Of the 283 attendees, 66.4 % were women, and all moderators were men. Women comprised 43 % of the 54 speakers, thus speaking less than male attendees (p < 0.001). No significant differences were detected in the duration or style of speech between men and women. Women participants commented more frequently on cases where the clinical attending was a woman (4.09 comments by women vs. 2.99 comments by men, p < 0.001), and less frequently when the attending was a man (2.48 comments by women vs. 3.20 comments by men, p < 0.001). On post hoc survey, men responded that they introduced ideas, guided discussions, and succeeded in influencing decisions significantly more than women reported that they did. Conclusion: Women physicians were underrepresented in tumor boards as moderators, speakers, and attendings of record. Women physicians commented less on men physicians' patients. Women felt less impactful than their men counterparts, despite having the same duration and style of speech. Prompted participation, moderator feedback, talking points, and limiting the number of cases can be used to balance representation in discussions.