Social media, which consists of Web-based technologies that facilitate idea sharing through collaboration, interaction, and discussion, have increasingly been incorporated into health care and medical education.1 Little is known about the use of social media in graduate medical education (GME).

The millennial generation of residents is unique in both the learning environment in which they train and the ways in which they learn.2 Thanks, at least in part, to the Internet, they are tasked with digesting vast amounts of ever-increasing information while still caring for individuals with complex medical conditions. Duty hours limit the time they are permitted to be in the hospital, leaving fewer opportunities for traditional classroom- and ward-based learning. Social media platforms, which offer ways to address these challenges, are progressively being introduced into GME.

Social media platforms have the potential to influence several domains of GME. Platforms, including wikis (i.e., Web sites offering collaborative modification of content), social networking sites (e.g., Facebook), microblogs (i.e., Twitter), and blogs—to name just a few—offer a venue through which trainees communicate, exchange ideas, learn evidence-based medicine, and promote their scholarship.3,4 Beyond providing educational content, social media are being used by residency programs to establish an online presence and recruit potential applicants.5 Residents are less frequently turning to mailings and the backs of journals to search for jobs; rather, many are using social media sites to obtain information on possible postgraduate opportunities. In addition to its effect on scholarship and recruitment, the use of social media—because of the public nature of platforms6,7—has also brought forth issues related to online professionalism in GME, and the potential for dissemination of protected health information.

Despite the incorporation and use of social media in GME, no study has sought to understand if the use of these Web-based technologies influences residents during their training—and if so, how. In recent years, a few studies have attempted to characterize the effect of social media platforms on medical education at large. A systematic review by Cheston and colleagues1 (2013) examined the effect of social media on medical education—specifically, knowledge and skill attainment. The authors of the study, however, defined medical education as all levels of physician training.1 Two other studies examined the impact of social media on online professionalism. Again, however, they focused on the medical community at large.8,9 While these studies do capture the opinions and attitudes of some resident physicians, they did not concentrate on GME trainees. Although informative, these reviews focused on only one area in which social media have had an effect, leaving other domains unexamined.

Here, we aim to fill this gap by conducting a systematic review of the peer-reviewed literature in order to examine the effect of social media platforms on residency education, recruitment, and professionalism. In doing so, we hope to shed light on the use of social media in GME.

Method

We conducted a systematic review of the peer-reviewed literature to explore the use of social media platforms in GME. We sought to understand how social media platforms affect (1) resident education and learning, (2) resident recruitment, and (3) resident professionalism.

Search strategy

Figure 1 shows the selection and review process for this systematic review. In consultation with a health sciences librarian (D.W.), we performed comprehensive searches of Medline, Embase, and Cochrane between October and November 2015. Major search terms for all databases included social media, graduate medical education, and residency (List 1). We conducted reference and related article searches in Scopus, PubMed, and ERIC. To identify additional manuscripts, we hand searched the bibliographies of included manuscripts.

List 1 Search Terms Used in 2015 Search of the Evidence-Based Literature Examining the Influence of Social Media on Resident Education, Resident Recruitment, and Resident Professionalism

Social Media

Social Media/ social adj2 media social adj2 network* Twitter Facebook blog* Tumblr Instagram wiki* podc*

Graduate Medical Education

Education, Medical, Graduate/ Schools, Medical/ grad* adj2 med* adj2 education med* adj2 school

Residency

“Internship and Residency”/ residen* Figure 1:

The selection and review process for a 2015 systematic review of the Evidence-Based Literature Examining the Influence of Social Media on Resident Education, Resident Recruitment, and Resident Professionalism.

Inclusion and exclusion criteria

We included any English-language articles published through November 2015 that pertained to social media platforms including blogs, microblogs (e.g., Twitter), social networking sites (e.g., Facebook, Yammer), podcasts, video-sharing sites (YouTube), and wiki platforms. We included studies that pertained to residents in any year of training and from any specialty. We did not limit studies to those conducted to the United States. To focus this review on social media platforms, we excluded online and electronic resources that were not interactive (e.g., e-learning modules). We also excluded conference abstracts and letters to the editor. We included only full-text articles in the review, and we identified and excluded duplicates. Two reviewers (M.S. and P.L.) independently reviewed the titles and abstracts of the retrieved articles. With a third reviewer (T.F.B.), they selected items for full-text review (see also Results / Study selection).

Data extraction

Two of us (M.S. and P.L.) performed data extraction for each study independently, and a third author (T.F.B.) resolved differences. We extracted the following variables from each study: study authors, year of publication, study design, setting, population studied, control population, social medial platform used, research approach, intervention, key outcomes, and study quality.

Quality assessment

For quantitative studies, we used the Medical Education Research Study Quality Instrument (MERSQI),10 a validated instrument that assesses the quality of medical education research. This 10-item scale assesses the methodological quality of studies in 6 domains: study design, sampling, type of data, validation of evaluation instrument, data analysis, and outcomes measured. Two of us (M.S. and P.L.) separately assigned points to each study, such that 6 represented the lowest quality and 18, the highest (see Supplemental Digital Table 1 at https://links.lww.com/ACADMED/A431). Using this tool, we created a standardized form to extract the data from included studies.

For qualitative studies, we used the Consolidated Criteria for Reporting Qualitative Studies (COREQ), a checklist that consists of 32 criteria, developed to promote explicit and comprehensive reporting of interviews and focus groups. The checklist is divided in three domains; research team and reflexivity, study design, and analyses and findings. Two of us (P.L. and M.S.) separately assessed the presence or absence of each of the 32 items on the COREQ checklist (see Supplemental Digital Table 2 at https://links.lww.com/ACADMED/A431).

We reconciled scoring differences for both the MERSQI and COREQ instruments through discussion.

Results Study selection

Our initial search yielded 534 English-language titles, of which we excluded 107 as duplicates, leaving 427 remaining in our initial search. Next, we hand searched forward citations, which yielded 153 new titles, of which we excluded 61 duplicates. After combining and deleting duplicates, we were thus left with a total of 519 articles. Initially, two of us (M.S. and P.L) disagreed about the inclusion of 5 additional studies (kappa = 0.86); we resolved these conflicts through discussion with a third reviewer (T.F.B.). We excluded two additional studies upon full-text review based on initial exclusion criteria. Ultimately, we included 29 studies11–39 in our analysis (Figure 1).

Appendices 1–3 present the characteristics of the studies including the following: study design, setting, participants, social media platform, intervention, outcomes, and the MERSQI and COREQ scores (as applicable). The mean MERSQI score was 9.57 and ranged from 7.5 to 14.5 (Appendices 1–3). The number of items reported on the COREQ checklist ranged from 21 to 29 (out of 32 items). Supplemental Digital Table 1 provides the MERSQI score components for each of the 22 quantitative studies,11–15,17–20,23–33,36,38 and Supplemental Digital Table 2 provides the COREQ checklist for components of the 7 qualitative studies (https://links.lww.com/ACADMED/A431).16,21,22,34,35,37,39

Appendix 1:

Summary of Evidence Regarding the Influence of Social Media on Resident Education, Based on a 2015 Literature Reviewa

All of the studies were published in 2010 or later. Most studies (n = 22; 75.8%) were descriptive14–16,18,20,22,24–39 (cross-sectional, survey designs, or case studies). The remaining 7 (24.1%) articles evaluated an intervention with pre and post measures.11–13,17,19,21,23 We uncovered no randomized controlled trials. Among the platforms studied, Facebook25,28,30–32,34–37,39 (n = 10; 34.5%), blogs13,21,22 (n = 3; 10.3%), Twitter11,14,16 (n = 3; 10.3%), and podcasts12,20,23 (n = 3; 10.3%) were the 4 most common. While the focus of each study pertained to residents, 9 studies13,17,23,25,28–30,33,39 (31.0%) included medical students and 9 studies14,24,26,27,30,31,33,34,38 (31.0%) included faculty members and program directors (PDs). Among the GME residencies, studies on social media most frequently focused on residents of general/subspecialty surgery27,28,34,35,37 (n = 5; 17.2%), internal medicine11,13,14 (n = 3; 10.3%), and anesthesia12,20,23 (n = 3; 10.3%); radiology, emergency medicine, pediatrics, family medicine, and dermatology residents were studied, albeit less frequently with respect to social media.

Resident education and knowledge

Of the 29 studies, 1311–23 (44.8%) attempted to use social media to enhance the educational value that residents receive (Appendix 1). Of these, Twitter11,14,16 (n = 3; 23.1%), podcasts12,20,23 (n = 3; 23.1%), and blogs13,21,22 (n = 3; 23.1%) were the most frequently used platforms. Wikis17,18 (n = 2; 15.4%), Skype19 (n = 1; 7.7%), and YouTube15 (n = 1; 7.7%) were studied less frequently. The average MERSQI score for studies pertaining to residency education was 10.65; the range was 7.5 to 14.5.

Within the domain of education, social media platforms, particularly Twitter and blogs, have been used to promote clinical concepts, disseminate evidence-based medicine, and circulate conference material to residents. We found that blogs13 were used both to complement case-based teaching during morning report and as a vehicle to support online journal clubs22 through which residents, authors, and other members of the health community could discuss research content. Bergl and colleagues11 surveyed internal medicine residents regarding their attitudes toward a one-year chief-run Twitter feed. Residents generally found the chief residents’ tweets informative, and 69% of 61 residents agreed that Twitter enhanced their overall education in residency. Residents reported that tweets about “pearls” from morning report, medical news, grand rounds, and evidence-based medicine were most informative to their learning.11

In addition to serving as an adjunct to traditional residency learning, Twitter is being adapted at medical conferences that residents attend.14,16 A high-quality case study (MERSQI = 14) by Desai and colleagues14 sought to determine the reach of tweets from participants at the 2013 Association of Program Directors in Internal Medicine meeting in 2013, whose attendance included a mixture of internal medicine faculty members, PDs, and residents. The authors found that most tweets were from faculty rather than from residents and that faculty members more frequently had their messages retweeted compared with those of residents. However, predicting the influence that tweets had on resident learners at the conference was difficult, and the investigators could not gauge how amplified tweets (i.e., tweets retweeted by conference attendees) affect learners after the conference because they did not measure such outcomes.

Several studies examined the effects of podcasts and wiki platforms on resident knowledge and skills. One of the first studies to examine the effects of podcasts on residency education was a small (n = 10), intervention-based pilot from the University of Kentucky.12 Bensalem-Owen and colleagues12 studied the effect of electroencephalogram (EEG) podcasts on resident knowledge outcomes. Ultimately, they found no significant difference in mean test scores compared with conventional lectures on EEG interpretations. A few years later, a high-quality study (MERSQI = 14.5) by Vasilopoulos and colleagues23 tested the effect of an EEG podcast on residents’ knowledge acquisition, which the investigators measured with test scores. In this small study population (n = 21), EEG interpretation scores improved after viewing the podcasts, and 100% of the residents found the experience either positive or neutral. In addition to demonstrating the effect of podcast technology on knowledge, the authors also found that prior experience with podcast technology was associated with greater gains in EEG evaluation scores. A cross-sectional study by Matava and colleagues,20 in which the authors surveyed 169 Canadian anesthesia residents, also aimed to assess the impact of podcasts on education. The authors found that 60% had used medical podcasts in the past, and 72.3% of these users found podcasts valuable because they afforded residents the “ability to review material whenever” they wanted.

Although wikis are familiar to most residents, their use and effect on resident education appear to be minimal. Kohli and Bradshaw18 conducted a case study at the University of Indiana to evaluate radiology residents’ comfort with, access to, and use of an internal wiki site. They found that while 78% of 51 residents knew how to edit pages, only 12% of residents used the internal wiki site for educational content. A University of Colorado study by Karimkhani and colleagues17 of medical students and residents found that a wiki about dermatology was highly rated among medical students, but less so among residents who favored patient-based exposure to cases and skin findings, rather than online content.

YouTube, known for its online video-sharing capability, appears to be another social media platform being used to promote skill attainment and self efficacy with respect to mastering procedural techniques. A study conducted by Fischer and colleagues15 evaluated the educational value and accuracy of arthrocentesis videos published by health institutions from 2008 to 2012. Of the 13 videos reviewed, the majority (n = 8; 61.5%) were considered to be of moderate quality by two reviewers (one internal medicine resident and one rheumatologist), and 8 (61.5%) were considered useful with respect to resident education. Although nearly half (n = 6; 46.1%) demonstrated sterile techniques, only 1 video (7.7%) was rated to be of excellent quality. Overall, 5 of the videos (38.5%) were classified as educationally unhelpful.

Resident recruitment

Of the studies, 6 (20.7%) explored how social media platforms are being used to address residency recruitment24–29 (Appendix 2). The majority of the studies24–27,29 (n = 5; 83.3%) used surveys to ascertain the attitudes of trainees or PDs toward social media as either (1) a mechanism for residency programs to enhance their online visibility or (2) a means of screening residency applicants. The average MERSQI score for resident recruitment studies was 8.66, with a range of 7.5 to 10.

Appendix 2:

Summary of Evidence Regarding the Influence of Social Media on Resident Recruitment, Based on a 2015 Literature Reviewa

Across GME, programs are acknowledging the presence of social media platforms and appear to be integrating them into aspects of their training programs. In one study of radiology PDs,24 38% of 132 associate PDs report social media use, and roughly a quarter felt that program Facebook pages would be of value. Similarly, a cross-sectional survey study by Schweitzer and colleagues29 of osteopathic applicants, interns, and residents found that a majority of applicants and residents are using social media sites for application information and postgraduate job searches. Commonly used platforms include Facebook, Twitter, LinkedIn, and Student Doctor Network blogs.

In addition to offering information about programs or jobs, several studies demonstrated the extent to which social media are being used by GME PDs to screen applicants during the selection process.26–28 In a study in the Journal of Surgical Education, Go and colleagues surveyed 250 PDs of general surgery and surgery subspecialty residency programs on their use of social media sites (Facebook, Twitter, MySpace, etc.) to screen applicants.27 They reported that 17% visit social media sites to gain info about applicants, and that upon doing so, 33.3% of those PDs ranked applicants lower after a review of their social media profile(s). Similarly, another study by Go and colleagues26 in Medical Education surveyed Accreditation Council for Graduate Medical Education (ACGME) PDs on their use of social media platforms during the intern selection process. They found that 16.3% of 196 PDs had reviewed Internet resources to learn more about a candidate’s application. In this study, a high proportion (38.1%) of PDs ranked applicants lower as a result of their social media profile(s). Of the platforms used, the majority of PDs used Facebook to screen applicants. A case series by Golden and colleagues28 of ear, nose, and throat (ENT) residency applicants at University of Alabama at Birmingham found that of 112 profiles searched, 11% of ENT applicants had questionable content (e.g., alcohol intoxication and wearing Halloween costumes portraying specific negative ethnic stereotypes); however, the content did not affect the applicants’ match outcomes.

Resident professionalism

Ten studies30–39 (34.5%) explored the effects of social media on professionalism in residency (Appendix 3). Most of these studies were exploratory in nature and highlighted issues of patient and resident privacy, particularly with respect to Facebook. The average MERSQI score for studies on resident professionalism was 8.67, and the range was 8.0 to 10.0.

Appendix 3:

Summary of Evidence Regarding the Influence of Social Media on Resident Professionalism, Based on a 2015 Literature Reviewa

Four of these studies (40%)30–32,36 surveyed residents about their social network behavior with respect to their patients (searching for patients or accepting friend requests). A study by Ginory and colleagues32 surveyed 182 psychiatry residents through the American Psychiatric Association about their social media (Facebook) use in the context of clinical care. Of those surveyed, 18.7% reported looking up patient profiles on Facebook, and 9.7% reported having received a friend request from a current patient; none of the residents accepted these requests. In addition, the majority of residents reported not having guidance regarding social media use during clinical training and said that more guidance would be welcome. A case series by Jent and colleagues33 in the Journal of Adolescent Health explored the attitudes of pediatric residents (n = 80) and pediatric faculty members (n = 29) toward social media use in general and toward seven specific fictional social media profiles. They found that more trainees used social media compared with faculty, but that both groups generally believed it was not an invasion of privacy to look at social media profiles of colleagues and patients. Only trainees, however, reported conducting social media site searches of patients.

In addition, six studies explored the use of social network sites by residency programs as a vehicle for identifying and censoring unprofessional behavior of trainees.33–35,37–39 A study by Langenfeld and colleagues35 in the Journal of Surgical Education searched Facebook profiles of 319 general surgery residents for unprofessional behavior. The study, which was of higher quality (MERSQI = 10), found that 73.7% of residents had profiles with no unprofessional content, 14.1% had profiles with potentially unprofessional content (drinking alcoholic beverages), and 12.2% of residents had profiles with clearly unprofessional content (e.g., Health Insurance Portability and Accountability Act violations, binge drinking, and sexually suggestive material). Another study by Ponce and colleagues37 assessed the effect of unprofessional online content on residency match outcomes. They reviewed the Facebook profiles of 153 orthopedic surgery applicants to the University of Alabama at Birmingham and rated profiles on a subjective professionalism scale of 1 to 3, where 3 = “no professionalism issues,” 2 = “questionable content,” and 1 = “definite violations of professionalism.” They were guided by ACGME’s “Components of Professionalism.”40 They found that applicants had a mean professionalism score of 2.82 and that 16% of applicants had at least some unprofessional content on their Facebook profile. The authors reported no significant difference in professional scores among those who matched compared with those who did not. In their research, Landman and colleagues34 took surveying for unprofessional social media content one step further. In addition to analyzing the use of Facebook among general surgery residents and faculty members, they also discussed the formulation of and proposed individual and department-wide guidelines for social media use at Vanderbilt University.34

Discussion

We conducted this systematic review of the peer-reviewed literature to examine the use of social media platforms in GME. Of the 29 studies we reviewed, most (n = 13; 44.8%) studied the effect of a particular social media platform on residency education. Within the context of resident education, we found Twitter, podcasts, and blogs to be the most frequently used platforms. Across studies, these platforms were used to share clinical teaching points, disseminate evidence-based medicine, and circulate conference materials to residents. At medical conferences, Twitter was the most frequently used platform to promote conference themes (via hashtags) and research content to attendees of whom some were residents. The majority of studies, however, were exploratory and used hashtags to analyze the frequency with which conference attendees accessed the platform, not its effect on learning. Studies that examined the effects of wikis and podcasts on resident education found that residents are most often using the platforms as a mechanism to review material on their own, at any point in time. For residents, one advantage of podcasts and wikis seems to be their comfort with these platforms. In several studies, trainees reported prior exposure to these platforms, and we wonder if perhaps this familiarity aided in reducing program-wide start-up costs when programs adopted them to promote learning or improve skill sets. However, comfort with a particular platform does not appear to always translate into frequent, sustained use or increased knowledge.

In general, most studies which pertained to social media and education were of modest quality and offered mixed results in terms of resident satisfaction and knowledge attainment. Additionally, 6 of the 13 studies that pertained to residency education had a mixed sample population that contained input from residents as well as faculty members and medical students. Thus, to assess the true effect of social media platforms on residency education is difficult—and generalizing these findings is even more difficult.

The majority of the studies assessing residency recruitment sought to examine the attitudes of trainees or PDs toward social media platforms, particularly through institutional-specific surveys. In addition, the authors of several studies interviewed PDs regarding the degree to which they screen applicants on the content of their social media platforms, such as Facebook. Results seemed to vary minimally across GME programs (e.g., medicine vs. surgery). Among the 10 studies that focused on resident professionalism, the majority used Facebook and explored the extent to which residents posted or disseminated information on this platform. A few of these studies raised concerns about privacy, but only one offered an attempt at providing guidelines. Overall, the studies focusing on residency recruitment and professionalism were of poor quality, and half were generated from single-institution-based surveys focusing on just one type of social media platform.

Whereas previous reviews have explored the use of social media in the health care environment and on undergraduate medical education, few have done so with respect to GME. Residents are a unique population of physicians, with a different set of needs and goals than their undergraduate counterparts and members of the medical community at large. As such, examining the literature that pertains to the use of social media platforms in GME is an important first step in understanding the effect of this relatively new phenomenon.

In doing so, we highlight a few notable themes. First, residents and residency programs across specialties are increasingly using social media. Today’s residents train in a complex learning environment characterized by a high volume of information and fast-paced delivery. Further, because of duty hours, residents have less time for formal, classroom-based learning. For this generation of millennial trainees, who are both comfortable and versatile with technology, the incorporation of social media into GME appears logical. Second, we found that despite many conference abstracts and editorials calling for research about social media and residency, few studies have attempted to formally study the adoption and/or use of social media in GME. Of those that have, few have used rigorous methods. Third, of the studies we reviewed, most offer mixed results and provide medical educators and residents with little guidance on how best to incorporate social media platforms into the residency experience. Finally, within the three domains we examined (education, recruitment, and professionalism), study design and outcomes varied tremendously.

Our study is not without limitations. Given the relatively recent emergence of social media and the rapid rate in which platforms develop, we have possibly missed studies published since our search that pertain to GME. Moreover, despite our efforts to include all relevant search terms, we may have unintentionally excluded keywords and thus relevant studies. An additional limitation is that much of the relevant works that emerged from our initial search were ultimately excluded because they were not peer reviewed but, rather, constituted conference abstracts. Finally, our intent was to capture the effect of social media platforms on GME, with a focus on residents. Some of the studies included—even though they pertained directly to GME—captured the attitudes of PDs, not residents. Additionally, a small percentage of the studies included some medical students in their study population.

In spite of these limitations, our systematic review adds to the current understanding of social media use in GME. Although interest in social media across GME seems to be wide and growing, its effect on education, recruitment, and professionalism remains inconclusive and understudied. Of the peer-reviewed studies we analyzed, most are descriptive in nature, highlighting resident attitudes toward social media in these three domains. Of the few studies that did include an intervention, the sample sizes were small and often lacked controls. Moreover, the results realized few tangible benefits for trainees—either in knowledge gains or in satisfaction scores. Among the studies that pertained to recruitment and professionalism, most were cross-sectional in design and used online surveys to capture resident attitudes toward platform-based content. Apart from the identification of trends, these studies do not allow for any associations or causal inferences.

Despite increasing use of these technologies by residents and medical educators, the adoption of social media into GME remains in its early stages. Further high-quality research is necessary such that the effectiveness of the platforms can be measured with validated instruments. In addition to moving toward intervention-based studies, researchers ought to be consistent in the outcomes they use so that results across studies can be compared. Given that this is a relatively new area of research, a qualitative approach offers value, particularly with hypothesis generation. However, among the qualitative studies included here,16,21,22,34,35,37,39 only two21,22 of the seven used focus groups or one-on-one interviews. Additionally, only one study conveyed their findings to the participants whom they studied. Future studies might follow the lead of Sherbino and colleagues22 and reengage the GME community such that findings can shape attitudes and practice.

Beyond improvements in methodology, future studies might also focus on pragmatism. What might be most useful to residents is if study findings offered practical instruction on how they should incorporate social media platforms into their resident experience in real time. Overall, further research is needed such that a best practice approach can be developed for trainees and program leaders.

References 1. Cheston CC, Flickinger TE, Chisolm MS. Social media use in medical education: A systematic review. Acad Med. 2013;88:893901. 2. Wutoh R, Boren SA, Balas EA. eLearning: A review of Internet-based continuing medical education. J Contin Educ Health Prof. 2004;24:2030. 3. McGowan BS, Wasko M, Vartabedian BS, Miller RS, Freiherr DD, Abdolrasulnia M. Understanding the factors that influence the adoption and meaningful use of socia