Across the Twitter-verse: Is Twitter an equitable tool in academic medicine? A scoping review

The goal of this study was to investigate what is known about the demographic characteristics of Twitter influencers in academic medicine. We conducted a literature search and scoping review exploring the demographic characteristics of Twitter influencers in academic medicine. Included studies evaluated Twitter influence by any metric and reported associated demographic characteristics. There were no date or language restrictions. Data points included metrics of influence, demographics, and study characteristics. From 1656 records, thirteen met inclusion criteria. Influence was variably defined based on followers, measures of centrality, measures of engagement, or through third-party software. Studies were conducted at single time points and relied heavily on facial recognition for classification of sex or gender, with no studies exploring race as a variable of interest. Twitter influencers in academic medicine were found to be disproportionately male: significantly fewer influencers were female at all levels of professional status. Two studies identified more female influencers among trainee populations. Female users were less likely to develop influence based on followers, retweets, likes, or other measures of engagement at academic meetings, despite equal or greater representation at these meetings. Results on associations between Twitter influence and number of publications or h-index varied considerably. No included studies explored race or ethnicity in relation to Twitter influence, which presents a significant gap in the literature. Our findings highlight the need for larger, user-engaged studies of inequities related to social media influence in academic medicine, especially as social media is increasingly incorporated into academic portfolios.