Journal clubs in the digital age: Twitter for continuing professional development
ABSTRACT
The departmental journal club (JC) is a well-established form of continuing professional development (CPD). Social media offers a range of interactive online platforms, allowing the traditional JC to move from a formal educational meeting with local health professionals to a digital platform with users across the world. The authors created the General Internal Medicine JC (@GIMJClub) on Twitter and following a year of activity retrospectively analysed the participation and impact of this medium of JC delivery. There were 61 different participants across different continents, specialties and levels who participated in the 12 JC sessions and sent 1,543 tweets in total. Factors that appeared to influence the success of an individual JC session included choosing diverse, topical papers to discuss and a wide range of hosts. This work demonstrates the success of a Twitter-based general internal medicine JC for CPD. @GIMJClub facilitated unique and diverse interactions not otherwise available.
Introduction
Journal clubs have long been the mainstay of medical and academic teaching. That is, a group gather and discuss the merits and pitfalls of a recent research publication of interest. Within a department's journal club (JC) there would be a weekly or monthly meeting and a cross section of experience, roles and seniority. These forums offer the opportunity to both teach literature critical appraisal and keep physicians up to date with cutting edge evidence-based medicine.1
Social media comprises a range of online platforms allowing real-time interactions between users for an immediate exchange of information. Social media is a widely-utilised technology and as such may provide a novel supplemental learning modality. Social media as a learning tool has been a consequence of further developments in e-learning. E-learning modules traditionally provided a didactic approach to tutorials with or without a self-assessment component. Web 2.0 software enables discussion and interaction between individuals.
It has been more than a decade since the advent of social media. Twitter is a micro blogging site and is one of the most commonly used forms for this medium with over 230 million users.2 Twitter allows users to have a small biography (bio) about themselves, seek out key word descriptors, ‘follow’ updates from others and converse in real time. The conversation takes the form of short public posts called tweets with a maximum of 140 characters, which can include pictures, videos, hashtags or hyperlinks. Naturally, this has evolved beyond simply a social forum and now involves professional interactions and discussion with the rapid dissemination of knowledge and opinions. Table 1 defines commonly used terms relating to Twitter.3 Specialty-specific Twitter journal clubs have become well established:4,5 urology (@iurojc, #urojc), cardiology (@Heart_BMJ, #HeartJC) and plastic surgery (@prsjournal, #PRSjournalclub) to name a few.
The aim of this work was to assess the success and engagement of a general internal medicine twitter journal club and assess what factors contribute to success of a regular social media-based learning event.
@GIMJClub conception
The authors of this piece established the general internal medicine twitter journal club (@GIMJClub) on 1 December 2015. The authors’ intention was to create a journal club across the breadth of general internal medicine to allow healthcare professionals from different backgrounds to discuss critical questions beyond their own specialty. The format is that on a pre-determined monthly date, a clinician in a particular field chairs a specific JC, focusing on one paper. The paper is chosen by the host and is agreed with the journal club moderators. The host is rotated each month. Topics were chosen to represent a range of challenges and concepts encountered in a variety of general medical settings. The JC ran on a designated Sunday every month either at 20:00 or 21:00 (GMT/BST) for approximately 1 hour. The chair would advertise the session through the @GIMJClub account, and frequently via their own account, prior to the session. The session would use the hashtag #genmedjc and four or five questions would be devised by the chair to be released during the hour to initiate discussions and debates around the chosen topic.
Methods
A retrospective observational study of the @GIMJClub activity and engagement over 12 months (December 2015–December 2016) was undertaken. Metrics for engagement (Table 1) and activity were assessed using the Twitter search function and analytic data. Only tweets using the #genmedjc on the day of the session were analysed. Tweets were identified by using the Twitter search function with the hashtag and the date range in question, eg ‘#genmedjc since: 2015-12-01 until:2015-12-31’. The search was undertaken during March and April 2017. Participants’ demographics were collected manually by reviewing their biographies and the followers of the account during March 2017 were located using Tweepsmap online software (Canada).6 The tweets analysed had to include the hashtag #genmedjc to be included in the analysis because of the limitations of the Twitter search engine. Frequently, a minority of participants would start and continue an interaction without the hashtag. Furthermore, participants, on occasion, misspelt the hashtag title (eg #gimjc and #genmedjp). This meant that not all participants and tweets could be included in the analysis and unfortunately is a limitation of this work. Consent was obtained from the participants whose tweets we have published.
Results
In March 2017, the JC account had a total of 880 followers. Fig 1 shows a map representation of the location of all the followers: 62.9% were situated in the UK and 14.7% in the USA. Twelve sessions were run by different chairs over the year of observation (one per month). The article selected for discussion was at the discretion of the chair and often reflected their area of interest or expertise. Table 2 summarises the articles discussed7–18 and the informatics about the chairs’ Twitter accounts. Sessions are labelled 1–12 based on sequential order of sessions.
After the inaugural session, the most popular month for attracting new participants was the 10th month (Fig 2) when a renal specialty trainee hosted the session, looking at early warning alert systems for acute kidney injury (AKI). This attracted predominantly renal specialists as new participants. Worth noting is that month 10 also attracted the biggest diversity in nationalities.
Online it may be challenging to assess how engaged in a discussion participants are. Twitter's 140-character limit leads to users sending multiple tweets and as such can be used to assess how intense a conversation is with the number of exchanges recorded. A total of 1,543 tweets were sent across the sessions using the unique hashtag, averaging 128 tweets per session. Fig 3 shows overall number of hashtag uses during each session and a breakdown of this by host or participant. As may be noted in a conventional teaching session in a room of people, the teacher/host may need to prompt or encourage participants to encourage discussion. During month 1 and month 9, the hashtag was used the most. More pertinent is the difference seen in month 2 and month 5 between tweets sent by participants and hosts (Fig 3). In month 2, there were ardent debates between participants on appropriate blood pressure targets when the host provided very little prompting to stimulate discussion. Whereas in month 5, the host comparatively had a lot to say about the appropriate use of statins in older patients, as did participants. Using month 2 as an example, there were 148 exchanges with only 29 coming from the host. This would correspond to a lively and interactive conventional teaching session.
Each tweet sent from a Twitter account has a metric analysis available to the account holder (definitions in Table 1). This allows for assessment of the impact a particular tweet has had – such as how often it has been retweeted and replied to. The average impressions and engagements of the tweets in each session was obtained and plotted (with standard deviation) for each month (Fig 4). Month 11 had the most impressions, ie the #genmedjc appeared on the most twitter accounts’ timelines, and month 3 had the least. Interestingly, this is not a direct correlation with the number of participants or the total number of hashtag uses.
Every host mentioned chairing JC at least once in their own twitter account; therefore, analysis of their overall Twitter activity was undertaken. This included their own personal total number of followers, the number of tweets they have posted and how long they have been on twitter (Table 2). These latter factors were taken as representation of how established a reputation they had and how experienced they were at using the Twitter interface. As would be expected, the number of followers a host had, largely related to how many tweets were sent from their personal account. As such, a greater number of followers suggests a bigger Twitter presence and larger sphere of influence, potentially increasing the number of participants in an organised event such as a JC.
Comments from JC hosts
The authors sought comment from the JC hosts regarding their experience. With consent, their opinions are below.It was a challenge to run a JC for people I didn’t know… but it was fulfilling to explore a topic with others, and get some different insights.
As an ICU physician, it was wonderful sharing a paper and discussing it with general physician colleagues. I really enjoyed the discussion generated and it's certainly changed my practice.
Wholly educational interactions with doctors from outside your specialty are surprisingly rare so this was a great opportunity to teach and learn from a diverse group of doctors.
It required lots of attention to keep up with everyone's comments but was also more rewarding than expected. Comments from others helped me to see different views from the same data.
It was a fantastic experience… as it not only gave opportunity to discuss diabetes care with many other professionals but also an opportunity for me to discuss others’ views too. In a short concise format, it appeared to work well.
Each host mentions a positive experience from the JC and an overall feeling of fulfilment. Further to this, the benefit of obtaining diverse opinions was a common theme (Box 1).
|
Discussion
Twitter is a micro-blogging website which has a 140-character limit to each post, meaning interactions are normally concise. Twitter accounts can be private or public, personal or professional. Twitter can be used for medical education in various ways: for publishing factual soundbites, links to recently published work, as a study resource, a supportive platform between students and teachers or student peers, or simply for developing one's professional network.19 Social media is a tool that facilitates the formation of associations and relationships, and enables discussion and collaboration to happen and a viable method of delivering interactive medical education. Real-time interactions enable social media platforms to host interactive conversations and as such facilitate professional discussion and collaboration. With the ability to download transcripts, Twitter JCs could provide a method of accredited CPD, especially for those who can’t take part during the time of the JC.
Many of our participants were based in the UK and there are various reasons why that may be. It could be the Sunday evening GMT/BST time slot was desirable. It could be because the discussions were in English or occasionally studies discussed related particularly well to working in the NHS. This created some open discussions on resource availability and costs in differing regions or countries. This was particularly pertinent in the session discussing the use of computerised tomography coronary angiography to evaluate chest pain presentations (month 3). This session generated discussion of service implications and costs within the emergency and radiology departments if such a study were to be implemented nationwide. This suggests the Twitter JC format allows for different domains of medical education to be covered, such as clinical governance and resource allocation, in addition to explicit knowledge exchange.
In addition to this, month 9 was hosted by our only non-UK based host. It is clear from Fig 1 that diverse nationalities follow the account. Of the five new participants, this session (Fig 2), four were of non-UK nationality and this trend continued into month 10. It is difficult to know why more followers did not formally participate in sessions. One can surmise that others participated passively in the JC by following discussions without contributing themselves (colloquially known online as ‘lurking’), perhaps because of language barriers, the British time zone, a lack of confidence, inability to commit time or a simple preference to observe anonymously. However, participants did comment that they felt less intimidated by asking questions than they would in a traditional JC setting, demonstrating the power of social media to democratise information and remove barriers, perceived or real. Similarly, participants and hosts alike commented on the usefulness of the Twitter-based JC to enable the consideration of other opinions and obtain information from specialists (see examples of tweets in Fig 5).
The ‘Net Generation’ theory suggests that those born after the mid-1980s are more likely to use social media.20,21 As such, it would be a logical hypothesis to think that most participants would be a junior doctor if using level of seniority as a surrogate for age. 41% of our session participants could be identified as being of training grade (data not shown), suggesting a large proportion of the participants were more senior. This number suggests that senior doctors are embracing social media for medical education. With this comes benefits such as easily accessed CPD relevant to participating clinicians and also improved confidence and communication skills for the juniors who are taking part. Furthermore, it is likely that the numbers using social media will increase as the ‘Net Generation’ progress through seniority levels. According to the most recent data, over 90% of medical students, 80% of doctors in training, over 60% of GPs and 40% of specialists use some form of social media.22–24
The predominance of nephrologists could be due to there being a well-established and successful renal specialty JC (@nephJC).25 The papers discussed by @nephJC include modern chronic kidney disease classification26 and AKI in pregnancy.27 The hosts have a website with further reading material relating to the area discussed and compared with @GIMJClub, there are a larger group of coordinators who run it. Furthermore, it was established in April 2014 and currently has around 5,000 followers (compared with 880 of @GIMJClub in March 2017). The success of @nephjc is summarised by Topf et al25 and there are some overlapping topics between both JCs, with both groups discussing one paper – the SPRINT trial.8 The @GIMJClub SPRINT discussion was one of the busiest, or most successful, weeks as suggested by the number of participants (Fig 2), by the number of times the #genmedjc hashtag was used by participants and by how little host prompting was required (Fig 3⇑). It was also one of the busiest discussions at #nephjc. SPRINT is a study that overlaps between multiple general internal medicine subspecialties and has been an extensively discussed trial in the last 18 months. These factors will have undoubtedly contributed to both JCs having such successful sessions when SPRINT was the chosen topic. Interestingly, in addition to the presence of nephrologists, the only two surgical colleagues who participated were urologists and, again, there is established urology social media use.28–30
The first Twitter medical JC was established in 2011, called @twitjournalcub (#twitjc) but has since stopped.31 There are also several other medical twitter JCs available, such as @gpjournalclub (#gpjc) discussing primary care articles once a month since 2015 with over 900 followers, Sheffield Teaching Hospitals Anaesthetics and Critical Care Department has its own twitter account, @STHJournalClub, discussing anaesthetic and critical care issues (although it does not appear this account has ever actually hosted a scheduled JC), the journals Circulation: Cardiovascular Quality and Outcomes (#OutcomesJC) and BMJ Heart (#HeartJC) both have twitter journal clubs. Another includes @hpmjc (#hpm) for hospice and palliative medicine with over 1,300 followers. Although there is no published evidence on which to base this, the authors observe that there is greater representation of some specialties compared with others. It would be intriguing to investigate this from a social, educational and industry point of view.
A possible downside of a Twitter JC could be that it is currently not recognised formally for CPD. Further to this, many of those who don’t use Twitter see it as frivolous and time wasting therefore attracting participants outside of work hours is a potential challenge. This is supported by current data that suggest only 24% of clinicians use social media to actively seek new medical information and even fewer (12%) contribute new information to social media. Furthermore, doctors appear to use social media predominantly for interactions with family and friends (83%), for entertainment (60%) and only 15% for CPD; however, data specifically for Twitter are not known.32,33 The authors feel that the number of participants and level of interaction in @GIMJClub disproves this. Possible issues unique to the Twitter JC format include choosing a time that maximises access across the world, language barriers and difficulty in following conversations unless familiar with the software. The name itself is perhaps suboptimal, as many medical specialty curricula in the UK no longer include general internal medicine. Additionally, the term may not be well known outside the UK, for example in America and Europe it would be termed purely ‘internal medicine’. Further to this, given that some specialties are disproportionately represented on Twitter, finding hosts from a range of general internal medicine specialties and keeping the topics interesting and varied may present challenges for future sessions.
In conclusion, we feel our analysis demonstrates the success of a Twitter-based general internal medicine JC as a means of CPD. @GIMJClub attracted many grades and specialties of clinician and facilitated diverse interactions that would be impossible with conventional teaching formats. Social media-based medical education offers an accessible and unique method to increase learning, but currently is not formally accredited.
Conflicts of interest
The authors have no conflicts of interest to disclose.
Author contributions
EM undertook the data analysis and initial preparation of the manuscript. RF assisted in figure preparation and preparation of the manuscript. DRW contributed to manuscript preparation. CJM devised the concept of the JC, hosted a session of the JC and contributed to manuscript preparation. Each author hosted a month of the Twitter JC.
The authors tweet under the following monikers:
Dr Eilidh McGinnigle = @EilidhPinkChic
Dr David Warriner = @DrDavidWarriner
Dr Rohin Francis = @MedCrisis
Dr Christopher McAloon = @DrChrisMcAloon
Acknowledgements
CJM devised the concept of the GIMJClub. CJM and the other authors must thank their co-founders Dr Stephen Makin and Dr Adam Seccombe with a special mention to Dr Karl Norrington, who has sadly died since the inception of the journal club. Many thanks also to each of our hosts Dr Kit Byatt, Dr Thomas Oates, Dr Amy Davidson, Dr John Ryan, Dr Olusegun Olusanya, Dr Partha Kar and Dr Nitin Arora.
- © Royal College of Physicians 2017. All rights reserved.
References
- ↵
- ↵
- Duggan M,
- Ellison NB
- ↵
- ↵
- Leung EY
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- Investigators N-SS
- ↵
- McAloon CJ
- ↵
- ↵
- ↵
- Avci K
- ↵
- ↵
- ↵
- Topf JM
- ↵
- Hall YN
- ↵
- Tangren JS
- ↵
- Loeb S
- ↵
- Rivas JG
- ↵
- Borgmann H
- ↵
- Bower C.
- ↵
- ↵
Article Tools
Citation Manager Formats
Jump to section
Related Articles
- No related articles found.
Cited By...
- No citing articles found.