Of 75 institutions with neuroradiology fellowship programs included in the 2016–2017 NRMP, eight neuroradiology groups had at least one social media account. Among the social media platforms, 4 of 75 neuroradiology programs were engaged on Facebook, and four used Twitter. By contrast, at these same institutions, 31 neurosurgery and 21 neurology institutions had at least one social media account. Neurology programs were engaged in Facebook (n = 13) and Twitter (n = 12) less than neurosurgery programs were (21 of 75 had Facebook and 19 of 75 had Twitter accounts), but both had more participation than neuroradiology (Fig. 1). The programs in all three fields had limited activity on Instagram and LinkedIn. In neuroradiology three institutions had Instagram accounts, and two had LinkedIn accounts. Neurosurgery and neurology participation on Instagram (eight and two programs) and LinkedIn (three and two programs) was also limited (Fig. 1). The difference between social media presence across fields was statistically significant: neuroradiology had the lowest participation (p < 0.001). Neuroradiology training programs with the largest number of followers on different platforms are shown in Table 1.

Logistic regression analysis showed that both neurosurgery (odds ratio, 5.9; 95% CI, 2.5–14.0) and neurology (odds ratio, 3.2, 95% CI, 1.3–7.9) had significantly greater likelihood of social media presence than did neuroradiology at the same institution. Neuroradiology training programs at larger institutions (≥ 5 fellowship positions) had significantly greater likelihood (odds ratio, 7.6,; 95% CI, 1.6–36.4) of having social media accounts than those with fewer than five positions. There was a small but statistically significant association (coefficient, 0.005; 95% CI, 0.001–0.009) between the number of followers and size of the neuroradiology training program. Larger programs had larger numbers of followers.

Among 75 neuroradiology, neurosurgery, and neurology division chiefs, 41, 46, and 43 had at least one social media account. Division chiefs had more accounts on LinkedIn than on other social media. However, few neuroradiology chiefs were actively engaged professionally on Facebook and Twitter (Fig. 2). The difference in division chief social media presence across fields was not statistically significant (p = 0.71). There was no significant association between the size of a neuroradiology training institution and the social media presence of the chief (odds ratio, 1.7; 95% CI, 0.6–5.2). There was a trend showing that if the chief had a social media account, there was greater likelihood that the same program also had a social media account, but it was not statistically significant (odds ratio, 1.84; 95% CI, 0.98–3.43).

We found 13 accounts for neuroradiology programs in social media (four Face-book, four Twitter, three Instagram, and two LinkedIn). In a subanalysis of these 13 neuroradiology accounts, we reviewed the content of these accounts. Four did not have any posts during the May 1, 2017, through April 30, 2018, time period. The University of Southern California Twitter account had the largest number of posts (417 posts) in the interval specified (median per month, 36; IQR, 19–50) followed by the Stanford University Instagram account (157 posts; median per month, 13.5; IQR, 11–15), John Hopkins Facebook account (135 posts; median per month, 9.5; IQR, 7–15.5), Stanford LinkedIn account (115 posts; median per month, 10.5; IQR, 4.5–13.5), University of Southern California Instagram account (19 posts), University of Utah Facebook account (seven posts), Mount Sinai Twitter account (five posts), and University of California San Francisco and Stanford Facebook accounts (each with one post in the 12-month period) (Fig. 3).

Of the 857 total posts from all neuroradiology accounts in the defined time period, general news and publications made up 51% (95% CI, 47–54%) of posts followed by other (17%; 95% CI, 14–20%), networking (14%; 95% CI, 12–16%), department news (12%; 95% CI, 10–15%), and educational posts (6%; 95% CI, 4–7%). Figure 3 shows the types of neuroradiology training program social media activity. The median number of responses per post (number of likes, comments, and shares combined per post) to different types of posts were as follows: eight (IQR, 2–16) for general news and publications posts, 11 (IQR, 5–20) for motivating and other posts, 11 (IQR, 7–18) for networking posts, 11 for department news posts (IQR, 7–23), and 10 (IQR, 6–29) for educational posts.

Almost all of the posts on LinkedIn (96%; 95% CI, 92–99%) and most of the Instagram posts (77%; 95% CI, 70–83%) were general news and publication posts. General news and publication posts also had the highest frequency on Twitter (35%; 95% CI, 30–39%) and Facebook (28%; 95% CI, 21–36%), but other activities, including networking, department news, and educational posts, made up the largest proportion of posts (Fig. 3). The difference between types of activities among social media platforms was statistically significant (p < 0.001). In addition, the audience engagement patterns differed across platforms. LinkedIn had the least audience engagement (Fig. 4).

Regarding neuroradiology society activity, all six neuroradiology subspecialty societies had at least one social media account. Facebook and Twitter were the two most active and followed platforms. The American Society of Neuroradiology had the largest number of followers (Table 2).

We investigated neuroradiology engagement in social media by the profile presence of all 75 current neuroradiology fellowship programs in the United States and their division chiefs. We also assessed U.S. neuro-radiology societies' social media engagement. Most of the academic neuroradiology programs are not actively using social media platforms to share information and connect with their colleagues or the public. Assessing social media engagement predictors, we found that the size of a neuroradiology program based on the number of neuroradiology fellowship positions is an important driver for having a social media account: programs with five or more training positions had significantly greater likelihood of having social media engagement. There was also a small but statistically significant association between the size of a program and having more followers.

Approximately one-half of neuroradiology division chiefs have at least one social media profile; however, most profiles are on LinkedIn, which is isolated and is not used for intercommunication as Facebook and Twitter are. We found higher odds for having an account on social media for the neuroradiology training programs in which the chief also was engaged in social media, but the difference was not statistically significant. We looked at the division chiefs' profiles as an indicator of their acquaintance with social media. The small number of neuroradiology training program chiefs engaged in highly active platforms like Twitter and Facebook suggests either a lack of awareness or a belief that social media are not a valuable tool for content dissemination. On the other hand, in a survey on using social media among radiologists [2], nearly 60% of the respondents used social networking for medical and professional purposes, and, interestingly, they were disproportionately men; only one in four users were women.

Comparing neuroradiology with two closely related fields, we found that both neurosurgery and neurology programs are more involved on social media than neuro-radiology is. Nevertheless, the social media engagement of these two fields is also not widespread. Among the social media platforms, Facebook was the most popular within the studied groups. Facebook engagement was the highest in neurosurgery, followed by neurology and neuroradiology programs. This might have been because Facebook is more generally popular [3]. Twitter was the second-most prevalent platform used by the studied programs.

A previous study [4] that included 41 academic neurosurgery institutions showed that the neurosurgery private practice sector is more active in social media than is academic neurosurgery [4]. This pattern also was observed in the radiology field [5], possibly reflecting barriers to social media adoption at academic institutions.

We found that neuroradiology engagement on social media is more prominent through professional organizations and societies (American Society of Neuroradiology, American Society of Head and Neck Radiology, American Society of Pediatric Neuro-radiology, American Society of Functional Neuroradiology, Society of Neurointerventional Surgery, American Society of Spine Radiology) than through institutional programs. Given that the numbers of followers of four of the six societies were fewer than 1000 and that the updated content was inconsistent, we believe that there is potential for greater use of social media platforms by organized neuroradiology compared with other radiology groups [4].

Browsing the contents of the 13 neuroradiology divisions that had an active social media account, we realized that the postings are mostly about general and publication news in the field (51%), but these types of posts attract less attention. Fewer posts are dedicated to educational content (6%), such as MR images, but our analysis showed that the number of responses to educational posts is greater than that to general news posts. If more of the followers of these accounts are in the training phase (i.e., medical students, residents and fellows), more attention should be paid to providing educational content in the posts.

Twitter, Facebook, LinkedIn, and Instagram have played a substantial role in sharing medical and educational content in medicine for both providers and consumers. Each platform has unique value and limitations. Aligned with our hypotheses, we found that Facebook and Twitter are the two social media platforms most commonly used to serve the neuroradiologic communities.

Although Facebook has the most active users of any social media platform, a previous study [6] showed that it is not as commonly used as Twitter for medical communication. However, our study results showed that the numbers of Facebook and Twitter profiles among the neuroradiology and neurology institutions were equal, but Facebook had a larger number in neurosurgery. This could be attributed to the fact that Facebook can provide medical information for patients and families, whereas Twitter users are more likely to be professionals communicating among themselves.

We found that LinkedIn is the most popular platform among division chiefs. LinkedIn is static compared with the real-time interaction of Twitter and Facebook. It is useful for making initial contact with people with common interests, but its platform for exchange is limited. Instagram has the same limitation as LinkedIn. Additionally, people currently tend to use Instagram for entertainment and personal matters rather than for education. It also seems that there is no place for patient engagement on the closed professional LinkedIn site.

To many health care professionals and scholars, social media activities have been presumed to be risky medicolegally, a waste of time, or both [7]. Some disagree. Encouraging neuroradiology training programs to be more involved in social media can help to recruit more medical students and residents to the field. It also helps fellows and practicing neuroradiologists to extend their professional networks in addition to keeping them up to date about neuroradiology news, publications, and opportunities at different institutions.

Several limitations of this study are worth noting. First, some division chiefs' Facebook or Instagram accounts were in private mode, and we could not assess their numbers of followers or the content of their posts. With LinkedIn, some chiefs did not accept our friend requests, or we had to pay to be able to send a friend request to those who were not on the network. We were only able to identify whether the chiefs had an account. Moreover, we explored neurology and neurosurgery social media activity only for the institutions that also have a neuroradiology fellowship program, and the findings may not necessarily be generalizable to all available neurology and neurosurgery residency programs. Finally, especially for fellowship programs with smaller numbers of positions, neuroradiology-related social media may be driven through the institution's residency program or other departmental and institutional accounts.

Social media are ubiquitous in the experience of young people in the current century. They are a potentially valuable educational and communication tool that modern medical scholars should use to disseminate knowledge. Most academic neuroradiology programs are not actively using social media and have not benefited from the opportunity for using social media as a channel for training, interaction, recruitment, and research promulgation. Neurosurgery and neurology have significantly more activity but also have not wholeheartedly embraced social media aggressively. With the diversity of platforms and the extensive user base, there is an opportunity for neuroradiology programs to reach and teach a larger number of individuals on social media platforms than with more traditional methods of communication.