Assessing the Use of Twitter to Share Canadian Residency Match Information During the COVID-19 Pandemic

Purpose In their final year, medical students explore prospective residency programs by completing visiting electives and attending interviews during the Canadian Resident Matching Service (CaRMS) process. Due to COVID-19, visiting electives and in-person interviews were suspended, leaving residency programs searching for alternate ways to share CaRMS information with applicants. This study evaluates the utility of Twitter to share CaRMS-related information prior to and during the pandemic. Methods Primary tweets published from three CaRMS cycles between 2018 and 2021 were identified using the analytics tool Vicinitas. The type, content, and language of tweets and the date and location of publication were extracted. Demographic data about tweet creators were determined using provincial regulatory college databases and institutional websites. Descriptive statistics were employed for categorical variables. All tweets were deductively analyzed. Results Of the 1,843 tweets, 603, 472, and 768 were published during the 2018-2019, 2019-2020, and 2020-2021 cycles, respectively. Most tweets were written in English (97.4%) and by medical students (29.5%) affiliated with Ontario universities. The most common types of tweets were supportive messages (29.1%), reflections about CaRMS (24.7%), and positive match results (20.8%). Rurally located institutions experienced the greatest increase in the total number of tweets between the pre- and full-COVID cycles. Conclusion Since COVID-19, Twitter has been increasingly used by medical professionals to share CaRMS-related information, primarily to promote programs and advertise CaRMS events. Given the environmental and financial benefits, CaRMS interviews will likely remain virtual, which highlights the ongoing need for residency programs to use social media platforms to share information with prospective applicants.


Introduction
During the final year of medical school, senior medical students (SMS) must select their preferred specialties, identify schools of interest, and apply to the Canadian Resident Matching Service (CaRMS) portal [1]. Traditionally, exploration of residency programs occurs during visiting electives where SMS travel across Canada to complete a series of two-week-long rotations at hospitals outside of their home institution [2,3]. SMS can also interview in-person with prospective residency programs and attend program informational sessions and socials [2,4]. Visiting electives and in-person interviews are therefore a key component of the CaRMS process, as they provide applicants with the opportunity to interact with staff and residents, understand the institutional culture, explore the city, and ultimately assess their "fit" with different programs [5]. However, in March 2020, the Association of Faculties of Medicine of Canada (AFMC) suspended visiting electives and in-person interviews indefinitely to limit the spread of COVID-19 [3,6]. This forced residency programs and applicants to find alternate avenues to share CaRMS-related information with one another.
Since the COVID-19 pandemic, the number of Canadian social media users increased from 27.66 million in 2019 to 35.81 million in 2023 [7]. Specifically, there has been increased use of platforms such as Instagram and TikTok by the healthcare community [8,9]. Twitter, a microblogging application that allows users to share short messages known as "tweets" [10], is cited as the most popular form of social media used by the medical community [8,9]. Throughout the pandemic, the number of global Twitter users increased 12-fold [11], as breaking COVID-related news was often shared first on Twitter [12]. Tweets can be tagged with a hashtag to allow users to easily filter for posts with related content [13]. Many healthcare-related hashtags (i.e., #COVID, #Surgery) are used by healthcare professionals on Twitter to share emerging research, circulate information, and connect with colleagues [8,9,12,14].
Since the suspension of visiting electives and in-person interviews, residency programs have increasingly relied on social media to advertise themselves [8,15]. Due to its accessibility and wide reach, Twitter has become a useful tool for residency programs to share real-time updates with CaRMS applicants regarding virtual events and deadlines [15][16][17]. To date, no study has analyzed whether the content of CaRMS-related tweets and user demographics has changed since the COVID-19 pandemic.
As such, the purpose of this study is to assess the content of CaRMS-related tweets from three cycles prior to and during the COVID-19 pandemic. Specifically, we aim to provide a better understanding of how educators and applicants are utilizing this social media platform to share CaRMS-related information and provide universities with information regarding their uptake of Twitter relative to other Canadian institutions.

Ethical statement
This study did not require ethics approval as data was collected from publicly available documents and online sources.

Study design
This study analyzed the content of CaRMS-related tweets and the demographics of the tweet creators prior to and during the COVID-19 pandemic. This study was exploratory in nature, given that it is the first study known to analyze the use of Twitter for sharing information related to the CaRMS process.

Procedure
Three CaRMS cycles of interest and their relevant hashtag were defined: (1) pre-COVID, (2) partial-COVID, and (3) full-COVID cycles ( Figure 1). The partial-COVID cycle corresponds to the 2019-2020 CaRMS year, where only the latter half of the cycle was impacted by the COVID-19 pandemic. The analytics tool Vicinitas was used to extract all tweets published within the predefined date range with the hashtags of interest. Tweets from all cycles were retrospectively collected from Vicinitas on June 18, 2021. Only primary tweets, defined as original posts made by a Twitter user, were eligible for inclusion. Retweets (i.e., reposting of an original tweet) or replies to original tweets were excluded. Basic tweet information including type (i.e., primary tweet, retweet, reply) and content of the tweet, date and time of publication, language, and username of the tweet creator was extracted. Vicinitas automatically extracts virality information, which details the number of times each tweet was favorited, retweeted, and replied to; however, it was not utilized in the final analysis. Using the date and time of publication extracted by Vicinitas, all included tweets were manually categorized into the appropriate CaRMS cycle by one study author (LV). The demographic information of the tweet creator including their profession (i.e., medical student, resident) and academic affiliation or the type of institution (i.e., residency program, national organization) and location were also manually extracted. If the user's profession or academic affiliation could not be clearly identified in the tweet or user profile, external sources such as provincial regulatory college databases (i.e., College of Physicians and Surgeons of Ontario) and institutional websites were consulted.

Data analysis
Descriptive statistics were used to describe the demographic information of tweet creators. Categorical variables were summarized using frequencies and percentages. One study author (LV) deductively analyzed eligible tweets into one of the following nine codes: (1) CaRMS-related experience, (2) event advertisement, (3) information related to CaRMS, (4) negative match result, (5) positive match result, (6) program promotion, (7) reflection about CaRMS, (8) supportive message, or (9) other (i.e., physician sharing their match story). Tweets that could be categorized into multiple codes were coded in consultation with a second author (KA). The coding process was iterative and comparative, as sample tweets were presented to all study authors to ensure consistent interpretation of the codes and to identify overarching themes.

Results
This study examined a total of 1,843 tweets published over three CaRMS cycles. Tweet characteristics are displayed in Table 1. Overall, 603 tweets (32.7%) were published during the pre-COVID cycle, 472 tweets (25.6%) in the partial-COVID cycle, and 768 tweets (41.7%) in the full-COVID cycle. Across all three cycles, the majority of tweets were written in English (97.4%) and posted by medical students (29.5%) and residents (16.4%) who were most often affiliated with the University of Toronto (11.6%) or McMaster University (11%). Tweet creators were located in 16 different countries; however, 94.3% of tweets originated from Canada.

TABLE 1: Tweet characteristics from three Canadian residency match cycles
The total number of tweets from each cycle categorized by academic affiliation is presented in Figure 2. Figure 2A demonstrates tweets created by creators affiliated with institutions outside of Ontario. Figure 2B depicts the total tweets from creators affiliated with institutions outside of Ontario. There was an overall increase in the total number of tweets from the pre-COVID cycle to the full-COVID cycle from all institutions outside of Ontario. Except for the Northern Ontario School of Medicine (NOSM), there was no notable increase or decrease in total tweets from Ontario institutions. The institutions with the greatest increase in total tweets between the first and third cycles were Memorial University Newfoundland (MUN), NOSM, and the University of Saskatchewan (U Sask), which saw a 25-, 10-, and eightfold increase, respectively.

FIGURE 2: Trend in total published tweets based on academic affiliation over three CaRMS cycles: (A) total tweets from creators affiliated with institutions outside of Ontario before and during COVID-19 and (B) total tweets from creators affiliated with Ontario institutions before and during COVID-19
CaRMS  Table 2 and summarized graphically in Figure 3. The number of tweets coded in each of the nine categories across all three cycles is displayed in Figure 4. Four of the nine most coded categories are described narratively below. The number of tweets from these categories was either consistently high across all cycles (i.e., supportive message) or changed considerably between cycles (i.e., event advertisement, program promotion, reflection about CaRMS).

Event advertisement
Within the pre-COVID cycle, 32 (5.3%) tweets were coded as event advertisements. Most tweets from this cycle advertised in-person socials held by the residency program. Within the partial-COVID cycle, nine (1.9%) tweets were coded as event advertisements and highlighted both in-person and virtual events. Within the full-COVID cycle, 112 (14.6%) tweets were coded as event advertisements, and all advertised online events held by the residency program such as informational sessions.

Program promotion
Within the pre-COVID cycle, 36 (6%) tweets were coded as program promotion and most commonly written to welcome students to in-person interviews and advertise redeeming qualities about the program. Within the partial-COVID cycle, 55 (11.7%) tweets were classified as program promotion. Similarly, tweets from this cycle advertised in-person interviews and promoted particularities of the city where the program was located. Within the full-COVID cycle, 118 (15.4%) tweets were coded as program promotion. Tweets from this cycle included links to the institution's website and YouTube videos that highlighted supplemental program information.

Supportive message
Within the pre-COVID cycle, 115 (19.1%) tweets were coded as supportive messages and often displayed words of encouragement to support the SMS as they traveled across the country for CaRMS interviews. Within the partial-COVID cycle, 134 (28.4%) tweets were coded as supportive messages. Several tweets from this cycle congratulated students on matching to a program and wished luck to those awaiting results. Within the full-COVID cycle, 123 (16%) tweets were coded as supportive messages and encouraged applicants to persevere despite COVID-related obstacles that impacted the CaRMS process.

Reflection about CaRMS
Within the pre-COVID cycle, 62 (10.3%) tweets were coded as a CaRMS reflection. These tweets were from applicants who shared their experiences traveling across the country and interviewing with programs. A minority of tweets described individuals' perspectives about the imperfections of the general CaRMS application process (i.e., faults of the matching algorithm, amount of time spent traveling). Within the partial-COVID cycle, 123 (19.3%) tweets were coded as a CaRMS reflection. Tweets from this cycle were similar to those pre-COVID. Within the full-COVID cycle, 162 (21.1%) tweets were coded as a CaRMS reflection. A large proportion of these tweets from both applicants and medical professionals (i.e., residents, physicians) mentioned the impact that COVID had on the application process. Tweet creators highlighted the benefits of having virtual events (i.e., better for the environment, convenience) and the disadvantages of this new format (i.e., inability to visit the school, lack of face-to-face interactions).

Discussion
This study examined 1,843 tweets published across three CaRMS cycles prior to and during the COVID-19 pandemic to evaluate Twitter's utility as a platform to share CaRMS-related information. To date, no study has analyzed the content of CaRMS-related tweets and the demographics of tweet creators. Previous research has evaluated the use of social media over the course of the pandemic by specialty-specific programs (i.e., general surgery, anesthesiology) and found that residency programs had increased use of social media in the years following the COVID-19 pandemic [15][16][17]. While the overall increase in total tweets noted in this study is in keeping with the literature, our data differ slightly as a transient decrease in total tweets between the pre-COVID cycle (2018-2019) and the partial-COVID cycle (2019-2020) from 608 to 472 was noted. We hypothesize that this unexpected decrease may be because medical professionals had a sudden increase in workload when COVID-19 was first declared a global pandemic [18,19], which resulted in less time spent using social media during 2020.
When examining the demographics of tweet creators, Twitter saw an increase in the number of tweets contributed by medical schools (fourfold), residency programs (fourfold), medical students (twofold), and program directors (twofold). These findings are in line with the literature [15][16][17], reporting increased use of social media during the pandemic by individuals directly involved with CaRMS applications. Interestingly, tweets created by organizations (i.e., College of Family Physicians of Canada) saw a threefold decrease between the pre-COVID and full-COVID cycles. This may be because some organizations narrowed their social media focus during the pandemic to convey information that is more closely aligned with issues directly related to COVID-19. For example, @CFPC_e shifted their focus to promote COVID-related content, such as mask-wearing and how to avoid burnout. A minority of organizations continued to tweet about CaRMS-related topics but excluded the CaRMS hashtag and therefore were not captured in our data.
There was an incidental geographic trend noted. Institutions located in cities with a population of less than 250,000, namely, Memorial University Newfoundland (MUN), NOSM, and University of Saskatchewan (U Sask), saw the greatest increase in total tweets between the pre-and full-COVID cycles. Although the lack of visiting electives and in-person interviews led to decreased visibility of all programs, our data suggest that institutions with the greatest need for self-promotion through social media were smaller institutions located in relatively less populated cities [20,21]. These geographic trends should be studied following the reinstatement of visiting electives to see if they persist. With the exception of NOSM, there was no overall trend in total tweets from creators affiliated with Ontario institutions. This may be due to the high use of social media pre-pandemic or no increased need for self-promotion during the COVID-19 pandemic.
The qualitative analysis demonstrated an increase in the number of tweets coded as "program promotions" and "event advertisements" from 6% and 5.3% during the pre-COVID cycle to 15.4% and 14.6% during the full-COVID cycle, respectively. The content within the tweets also evolved throughout the pandemic to include links to institutional websites, YouTube videos, and sign-up pages for virtual events. Interestingly, the number of tweets coded as "supportive tweets" overall decreased over the three cycles. While the reason is unclear, it is well documented that healthcare workers are reporting increased stress due to the emotional and physical toll associated with providing healthcare services in Canada [22]. Given that the full-COVID cycle took place during the second year of the pandemic, it is possible that an element of compassion fatigue carried over to social media and contributed to the decrease of supportive tweets toward SMS progressing through CaRMS [23,24]. Alternatively, much of social media's support was directed to frontline workers; thus, it is possible that SMS were simply overlooked as a group in need of support [25].
Given that virtual CaRMS interviews have important financial and environmental benefits [26], they are likely to persist beyond the COVID-19 pandemic. As such, this provides an ongoing utility for tweets to convey important program-specific information to applicants who invariably cannot complete visiting electives at all institutions across the country. While this study provides useful insight into the utility of Twitter as a medium to share CaRMS-related information, our study is not without limitations. First, this study only examined the content of tweets with the following hashtags: #CaRMS2019, #CaRMS2020, and #CaRMS2021. Thus, it is possible that CaRMS-related tweets were not captured in our data because they did not contain one of these hashtags. However, given that a total of 1,843 tweets were examined in this study, our data likely represent overall trends throughout the COVID-19 pandemic. Second, this study does not examine the number of Twitter users who interacted with each tweet. Future studies may consider assessing how the number of individuals interacting with CaRMS-tagged tweets via replies or retweets has changed since the pandemic. Future studies may consider analyzing the use of Twitter following the reinstatement of visiting electives in the 2023-2024 CaRMS cycle to assess if these numerical and geographic trends prevail. Analyzing tweets over several additional cycles may help to establish long-term trends that may differ from those noted in this study, specifically analyzing the number of tweets by academic affiliation to identify changes in the geographical trends and in the common types of tweets (i.e., supportive tweets, event advertisements). Third, a qualitative study may be undertaken to evaluate how often CaRMS applicants interact with the CaRMS-related tweets identified in this study and whether applicants consider the information shared during the application process. Finally, determining if the residency programs with increased Twitter use also had increased interest in their program as measured by the total number of CaRMS applications may be valuable.

Conclusions
This study examines Twitter as a platform to share CaRMS-related information prior to and during the COVID-19 pandemic. Overall, the number of tweets published by medical students, residents, and residency programs increased and were primarily created to promote programs and advertise CaRMS events. Institutions located in relatively rural locations had the greatest increase in tweets. Although visiting electives will be reinstated, CaRMS interviews will likely remain virtual given their financial and environmental benefits. This highlights the ongoing need for Twitter to convey important program-specific information to CaRMS applicants.

Additional Information Disclosures
Human subjects: All authors have confirmed that this study did not involve human participants or tissue. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.