Our search retrieved 9634 potential citations, which based on our criteria were reduced to 628, the full-texts of which were analysed (See Fig. 1 for a flow diagram of the results). We were unable to locate the full-text of 11 publications despite searching 5 separate institutional libraries and emailing corresponding authors three times. These publications were excluded.
Many of the publications (n = 228) were not geographically anchored (e.g., they did not refer to a specific region of the world). For the remainder, the majority were in North American (n = 187) or European (n = 88) contexts, although there were publications from many other jurisdictions [Asia (n = 31), Australia/New Zealand (n = 20), Middle East (n = 11), Africa (n = 5), South America (n = 4)].
Prevalence over time
Over the past decade, publications featuring social media for knowledge translation and education of physicians and medical trainees have increased (See Fig. 2).
Specialties of origin
We identified a range of specialties, with surgical specialties most often represented. (See Table 1 for specialties or groups of specialties that were included in more than 10 publications.)
The majority of articles targeted attending (fully qualified and practising) physicians (n = 346) and postgraduate trainees (e.g. residents and fellows, n = 242). Of the articles, 195 described interventions aimed to educate or disseminate new knowledge to medical students in addition to other learners, with 20 of these publications focusing on preclinical basic science, and 11 focusing exclusively on undergraduate medical education. Eighty-three articles addressed our target audiences plus at least one other health professional group (e.g., nurses, dentists). The majority of papers within all of the groups tended to address similar issues (i.e., using social media for learning, uptake and usage), although the articles targeting trainees/students tended to have more content regarding professionalism (many of which were excluded in our final analysis).
Social media strategies used for knowledge translation and education
Prevalence of types of social platforms
A wide range of social media platforms were featured, including Twitter (n = 157), blogs (n = 104), Facebook (n = 103), podcasts (n = 72), video archival platforms (e.g. YouTube, n = 68), and Wikipedia (n = 57) (See Table 2 for a comprehensive platform listing). The majority of articles focused on the technology itself (77.5%, n = 487) (e.g., an article that introduces a medical specialty to Twitter [30, 31]), whereas the remainder focused on the individuals using the technology (22.5%, n = 141), such as the rate at which residents used podcasts or blogs [10, 32].
Strategies described using social media platforms to engage in knowledge translation and education
Reviewed publications described many social media strategies. These strategies focused on ways to push out information or foster engagement. For example, several studies described organized efforts by journals, professional societies, and individuals to push journal articles to their audiences, such as creating a podcast to showcase a recent article. Other common strategies focused on fostering interaction between individuals. For example, multiple residencies reported using social media to facilitate engagement around journal articles using a virtual journal club format. In several cases, strategies contained both efforts to push information and foster engagement. Table 3 features examples of identified strategies. This table specifically arose because of our expert consultation process.
Types of social media scholarship in knowledge translation and education
The bulk of included publications (n = 242) were descriptive studies of physicians’ social media usage for the purposes of knowledge translation and education. Another prominent scholarship type (n = 192) was conceptual or narrative reviews describing the application of social media for connecting physicians and their trainees. Innovation reports were the next most prevalent type of scholarship (n = 122). A minority of studies included elements of clarifying between types/formats of social media for knowledge translation and education (n = 5), critical appraisal of social media-based content (n = 5), and integrative reviews of social media (n = 3). Table 4 lists types of scholarship in knowledge translation and education.
The majority of empirical studies (n = 353) applied quantitative methods (79.9%, n = 282). Forty-four studies utilized mixed methods (12.5%, n = 44) and a minority featured qualitative methods (7.5%, n = 27). Five studies critically appraised online content and two reported consensus building using modified Delphi methods.
Methods used in studies about knowledge translation and education for physician audiences were varied. The predominant study design was the quantitative survey (n = 180 studies), which was often mixed with other methods such as free-text survey questions. Usage analytics were also popular (n = 79). Other common study types included applying objective observations or text analysis of substantive written texts (e.g., blog posts or archived narratives), interviews, and micro-text (e.g. Tweet) analysis. Table 5 reports all methods used.
Within empirical studies, multiple themes were discerned via our analysis of these types of articles. We identified the feasibility of social media for practice improvement as the most prevalent theme (n = 330). Multiple publications also described specific elements of a particular technology (n = 327), such as podcasts or e‑learning. Examples of these studies were: a) Lien and colleagues comparing blog posts to podcasts; both groups showed increased knowledge retention, with no preference for one media over the other ; b) multiple studies examined the use of mixed media (podcasts, video) for medical students and resident translation of learning, ultimately concluding that learners rated the media tool more highly than traditional learning [34, 35]; c) other studies have examined use of Twitter as a study aid or engagement tool in student courses, but have not shown improved performance [36, 37].
Table 6 features major themes identified and provides a reference to exemplar papers. Some of the rarer themes included: challenges and pitfalls aside from professionalism (n = 5), humanism (including medical humanities and reflective practice, n = 3), costs of the innovation (n = 2), correlation of social media with bibliometrics (n = 2), health policy change and advocacy (n = 2), informatics (n = 1), and mentorship (n = 1).
Outcomes measured in social media for knowledge translation and education studies
Most (134; 70.1%) articles measuring outcomes were at the acceptability level. For example, in one study researchers developed audiovisual podcasts, which were reviewed by medical students to assess their usability . Forty-eight studies (24.7%) measured knowledge acquisition. For example, one research team conducted a prospective, nonblinded, three-arm randomized trial to compare use of Wikipedia, UpToDate, and a digital textbook for knowledge acquisition among pre-clerkship students to assess knowledge acquisition . Six studies (3.1%) measured behavioural change. Finally, four studies (2.0%) measured organizational or patient outcomes. For example, one study compared the use of emergency department communications and peer-to-peer learning via WhatsApp and standard telephone for educating peers during consultations .
Diffusion of an innovation through scholarship
There appeared to be a diffusion of innovation that has moved into more scholarly avenues—a diffusion of innovation through scholarship. Akin to Rogers’ framework for diffusions of innovation [48, 49], we found a pattern in our review of the literature. We have assembled our findings into a new conceptual framework, which highlights specifically how educational innovations seem to diffuse through the scholarly system. For instance, we observed the trend of narrative papers commenting about the potential and usage of social media, moving into more descriptive papers about the phenomenon via surveys and demographics-based studies, with patterns of usage and preferences for usage appearing in tandem with the more narrative literature. Soon after, the one-off, single-centre innovation publications emerged—variably describing interventions, reporting new ideas for application, with a mind for others to replicate and reproduce. After the appearance of descriptive publications come the justification articles (which seek to substantiate why a certain innovation is worthy) . Finally, there are the clarification papers (which seek to clarify ‘what works’)  and the critical appraisal publications, which seek to set standards and examine the innovation via a more critical lens. Fig. 3 depicts the scholarly sequence that social media for knowledge translation and education seemed to have appeared in our review; the stages were not discrete and often overlapped as the field tended to mature over time.