Editor's Note: The online version of this article contains a table of examples of the 5 themes in the reflection domain.

Introduction

Reflective writing is used to promote learning and professional growth in medical education.1 The rationale for reflection, evidence supporting reflection, and general tips for teaching reflection have been described in recent reviews.1–3 Sharing reflective writing with peers and supervisors facilitates feedback that enhances understanding. Current mechanisms for sharing reflections include small group meetings; printed or electronic documents; e-mail discussion threads; or web-based discussion boards and blogs.2

A mechanism for sharing reflective writing that has not been studied much is social networking. Platforms such as Facebook or Twitter facilitate the communication of thoughts and feelings with many people. Social media sites are readily accessible on desktop or laptop computers, smartphones, or tablet devices and are familiar to millennial learners who commonly use social media for communication.4

Although social media has great potential to connect learners and teachers, popular platforms, such as Facebook or Twitter, are open to the public and relatively unsecure. Holding honest and thoughtful dialogues through those platforms is difficult, particularly for discussions about patient care. Many educators have voiced concerns regarding social media and professionalism.5

We explored the feasibility of using a secure social media platform for sharing reflections and promoting reflective discussions in an emergency medicine (EM) residency training program.

Methods

Results

Almost all (36 of 37, 97%) residents consented to participate in the investigation, created an Edmodo account, and joined the EM Reflection Group. One male resident did not participate. Twenty-one of 34 (62%) faculty (7 of 13 women, 54%; 14 of 21 men, 67%) consented to participate, created an Edmodo account, and joined the EM Reflection Group. There were 54 posts and 140 replies during the 2-month study period (table 1).

TABLE 1 Basic Descriptive Analysis of the Number of Posts and Replies During the 2-Month Pilot

View Fullscreen

Eighty-nine percent of resident participants (32 of 36) and 90% of faculty (19 of 21) completed the postpilot survey. Desktop or laptop computers were the most commonly used devices to access the software, and approximately half of both groups reported using smartphones to some extent. Only a few participants used tablet devices.

Most faculty (79%, 15 of 19) and residents (72%, 23 of 32) reported using at least 1 social media site (in addition to Edmodo) on a regular basis, and 32% of faculty (6 of 19) and 9% of residents (3 of 32) used more than 1 social media site regularly. Facebook was the most commonly used social media site for both groups.

We asked participants to rate their experience with Edmodo and reflective writing using 7 items with Likert response sets. Mean responses to each of the 7 items did not differ significantly across the 2 groups (table 2).

TABLE 2 Descriptive Statistics By Role and Results of Independent t Test Comparisons for Opinion Items About Edmodo Pilot Experience

View Fullscreen

Qualitative analysis yielded 2 global domains—reflection and nonreflection—with 43 posts (80%) categorized to the reflection and 11 (20%) to the nonreflection domain. The 3 themes of the nonreflection domain were sharing information for teaching purposes (literature, electrocardiograms, images), announcements/notifications, and fun commentary. The reflection domain contained 5 themes: clinical case management, challenges of clinical work, expression of emotions, teaching and learning, and teamwork (provided as online supplemental material).

Discussion

Efforts to promote reflection in medical education have increased significantly in recent years. Our piloted study used secure social media as an alternative platform for sharing reflective writing and promoting discussions. Faculty and residents reported that the platform was engaging and that the reflections were valuable to their professional development. This is consistent with experiences from continuing medical education suggesting that social and collaborative learning is central to promoting change in professional practice.9

An unexpected and interesting finding of our study was that faculty used the Edmodo platform more frequently than residents did. We suggest a number of potential reasons for this. Participation rates (21 of 34 faculty, 62%; 36 of 37 residents, 97%) may indicate a selection bias for faculty with an interest in reflection, social media, and/or technology. Although the study was not designed to explore differences between individuals, it was clear that some faculty and residents were more active in the discussions than others. Most prior studies have involved medical students with required reflective assignments.10 The voluntary nature of our investigation and variable use of the platform by individuals may suggest reflection to be more important for some. In this context, the question of mandatory versus voluntary participation in the use of social media for reflection is worth exploring further. Perhaps a minimal requirement of posts and replies is required to both keep the conversation going and to ensure a level of exposure to reflection for all learners. Those who find benefit can then participate beyond that required level.

Another potential explanation for faculty predominance may be resident perceptions of the security of the Edmodo platform. Perhaps residents felt inhibited by faculty participation or were concerned their reflections would be used against them in some manner. Despite the lower frequency of posts and replies written by residents, survey data were positive and were not significantly different than faculty responses. This suggests some value in simply reading discussions. Future research is needed to look at the composition of reflection groups. Perhaps only select faculty—those viewed as nonthreatening by residents and skilled in facilitating discussions—should be included.

Given the positive survey data, we expect the frequency of postings to improve over time, and suggest a variety of measures for increasing learner participation. First, because only half of study subjects accessed the platform from their smartphone, we might increase participation by getting participants to download the program application (app) for their phone. An incentive, such as a gift card, could be offered to those who do so. A second suggestion is to have faculty provide discussion prompts. Perhaps learners would find it easier to reply to discussion topics than to initiate them.

Our experience with residents suggests that even the appearance of a mandated change will be met with resistance. Accordingly, we intentionally designed the pilot to be open-ended, relatively unstructured, and voluntary.

The pilot data revealed that women were less likely to comment than men, even after accounting for participation rates. That trend requires further research. Should men's comments continue to be overrepresented, efforts to facilitate a balance in participation across both groups would be needed.

Another interesting finding of our investigation was that 25% of posts were not reflective in nature, despite the intent of the pilot study. Faculty and residents found uses for the technology that we did not anticipate, such as sharing educational resources and contributing announcements and nonreflective commentary. Future areas for study include exploring the effect of unanticipated uses of this technology, including uses to promote collaboration or community building.

This pilot study is limited in that it involves 1 residency program at a single institution for a short time, and the outcomes were primarily what Kirkpatrick refers to as level 1 evidence (reaction/participation/satisfaction).11 The survey instrument lacks evidence to support the validity of responses and our interpretation of the results. Research suggests that many physicians inherently desire to reflect and share their experience with peers. Providing a safe and convenient venue for that to occur is important and may prevent perceived lapses of professionalism through the use of traditional social media sites, such as Facebook.12

Conclusion

Secure social media offers a feasible option for sharing reflections and facilitating reflective discussions in medical education.