If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Dermatology, Northwestern University Feinberg School of Medicine, Chicago, IllMedical Education Clinical Scholars Program, Northwestern University Feinberg School of Medicine, Chicago, Ill
Department of Dermatology, Northwestern University Feinberg School of Medicine, Chicago, IllCenter for Bio-Integrated Electronics, Northwestern University, Evanston, Ill
Residency applicants often use social media to discuss the positive and negative features of prospective training programs. An examination of the content discussed by applicants could provide guidance for how a medical education faculty can better engage with prospective trainees and adapt to meet the educational expectations of a new generation of digital-native physicians.
OBJECTIVE
The objective was to identify unstructured social media data submitted by residency applicants and categorize positive and negative statements to determine key themes.
DESIGN
The study design was qualitative analysis of a retrospective cohort.
SETTING
Publicly available datasets were used.
PARTICIPANTS
The participants were anonymized medical trainees applying to residency training positions in 9 specialties—dermatology, general surgery, internal medicine, obstetrics/gynecology, plastic surgery, otolaryngology, physical medicine and rehabilitation, pediatrics, and radiology—from 2007 to 2017.
MAIN OUTCOMES AND MEASURES
After we developed a standardized coding scheme that broke comments down into major features, themes, and subthemes, all unstructured comments were coded by two independent researchers. Positive and negative comments were coded separately. Frequency counts and percentages were recorded for each identified feature, theme, and subtheme. The percent positive and negative comments by specialty were also calculated.
RESULTS
: Of the 6314 comments identified, 4541 were positive and 1773 were negative. Institution was the most commonly cited major feature in both the positive (n = 767 [17%]) and negative (n = 827 [47%]) comments. Geography was the most cited theme, and City, Cost of Living, and Commute were commonly cited subthemes. Training was the next most cited major feature in both positive (n = 1005 [22%]) and negative (n = 291 [16%]) comments, with Clinical Training being more commonly cited compared to Research Opportunities. Overall, 72% of comments from all were positive; however, the percent of comments that were positive comments varied significantly across the 9 specialties. Pediatrics (65%), dermatology (66%), and internal medicine (68%) applicants were more likely to express negative comments compared with the global average, but physical medicine and rehabilitation (85%), radiology (82%), otolaryngology (81%), and plastic surgery (80%) applicants were more likely to express positive comments.
CONCLUSIONS AND RELEVANCE
This qualitative analysis of positive and negative themes as posted by applicants in recent matching years is the first and provides new detailed insights into the motivations and desires of trainees.
The social media environment is expansive and rapidly changing, with weekly social media use among 18- to 24-year-olds increasing from 89% to 98% from 2014 to 2016.
Villanti AC, Johnson AL. Social Media media Use use and Access access to Digital digital Technology technology in US Young young Adults adults in 2016. J Med Internet Res. Jun 7 2017;19(6):e196.
Villanti AC, Johnson AL. Social Media media Use use and Access access to Digital digital Technology technology in US Young young Adults adults in 2016. J Med Internet Res. Jun 7 2017;19(6):e196.
To this end, medical students and residents are using social media to study, network, discuss patient encounters, job hunt, share research, and more. However, research surrounding medical trainees and social media has largely been limited to professionalism.
Attitudes and practices of surgery residency program directors toward the use of social networking profiles to select residency candidates: a nationwide survey analysis.
Listening to the voice of patients on platforms such as Yelp has been suggested as a way to understand their needs and improve the experience of health care.
Similarly, listening to the digital voice of students and residents may allow for improvement in the delivery of medical education.
For most medical students, the transition from medical student to resident is formative, and the selection of a residency program represents a critical decision, but there is a limited understanding of how modern medical school seniors use social media to assess residency programs. We hypothesized that unstructured social media data created by medical students regarding residency selection will provide new, valuable, and renewable insights on their expectations and preferences for residency training. The National Resident Match Program (NRMP) Applicant Survey provides some insights into candidate preferences, but it is only periodically published and has limited breadth, given its structured design and limited dissemination.
National Resident Matching Program, Data Release and Research Committee, Results of the 2017 NRMP Applicant Survey by Preferred Specialty and Applicant Type. Washington, DC: National Resident Matching Program; 2017.
A better understanding of the contemporary expectations of medical students would allow residency programs to better recruit and engage prospective learners while adapting to meet the educational expectations of a new generation of physicians.
METHODS
We performed an Internet search of key social media sources: Twitter, Reddit, Facebook, Student Doctor Network (SDN), Aunt Minnie, and Otomatch. We selected postings using the following terms: “residency match list,” “interview experiences,” and “residency rank list.” We identified postings related to 9 specialties: dermatology, general surgery, internal medicine, obstetrics/gynecology, plastic surgery, otolaryngology, physical medicine and rehabilitation, pediatrics, and radiology. All searches were conducted within 1 week of the rank list deadline for the 2017 NRMP match (February 22, 2017). This study is exempt from Northwestern Institutional Review Board approval, because it used only publicly available datasets.
A standardized coding scheme was developed (see Supplement 1, available online). The scheme included 3 nested levels: major features, themes, and subthemes. Comments were coded to the highest level of specificity possible; for example, “The city is expensive” is coded as Institution (major feature) > Geography (theme) > Cost of Living (subtheme). Two authors (LA, BD) coded all unstructured comments using NVIVO v11 (QSR International, Burlington, Mass). Discrepancies were discussed (SX, LA, BD) until a consensus was reached. We determined the proportion of positive comments by specialty using Stata v14 (StataCorp, College Station, Texas) with a logit transform to calculate 95% confidence intervals. We used a logistic regression model to compare the proportion of comments that were positive within each specialty. Specialties were identified as having a greater proportion of positive comments than overall (odds ratio [OR] > 1.0, P < .05), a lower proportion of positive comments than overall (OR < 1.0, P < 0.05), or not significantly different than average.
RESULTS
Three online resources—SDN, Aunt Minnie, and Otomatch—yielded results that met our search criteria. The most recent relevant posting groups were identified, and up to 3 per specialty were mined for comments. Focusing on the 3 most recent forums and threads by year allowed for a more contemporary analysis. For dermatology, plastic surgery, otolaryngology, internal medicine, and radiology, the forums used corresponded with the 2014–2015, 2015–2016, or 2016–2017 NRMP match cycles. For the remaining specialties, earlier groups were used representing match cycles 2007–2008 (obstetrics/gynecology), 2011–2012 (physical medicine and rehabilitation), and 2012–2013 (general surgery and pediatrics). These postings yielded 6314 total comments (4541 positive and 1773 negative entries) specifically related to individual residency programs. On SDN alone, these comments accounted for more than 1.2 million aggregate views.
Of the positive comments (Figure 1A), Institution was the most commonly cited major feature (n = 1323 [29%]), and Geography was the most cited theme (n = 767 [17%]). City (n = 312 [7%]), Cost of Living (n = 121 [3%]), and Cultural and Leisure Activities (n = 104 [2%]) were the most commonly cited subthemes. Training was the next most cited major feature (n = 1005 [22%]), and Clinical Training (n = 655 [14%]) was more commonly cited than Research Opportunities (n = 238 [5%]). Resident Experience was another commonly cited feature with important themes and subthemes, including Likeable Residents (n = 248 [5%]), Successful Fellowship Placements (n = 197 [4%]), and Perceived Resident Happiness (n = 188 [4%]). Table 1 highlights representative comments from each of these themes and subthemes. Institution was also the most commonly cited negative feature (n = 827 [47%]; see Figure 1B), with Geography being the most cited theme (n = 637 [36%]), and City (n = 191 [11%]), Commute (n = 112 [6%]), and Cost of Living (n = 88 [5%]) being the most cited subthemes (Figure 1B). Training was the next most cited major feature (n = 291 [16%]); again, Clinical Training (n = 226 [13%]) was cited much more commonly than Research (n = 49 [3%]). A summary of major features by specialty is included in the supplementary materials available online.
Figure 1Positive (A) and negative (B) factors considered across all specialties shown in sunburst hierarchical plots. The inner rings represent broader themes, and the outer rings represent subthemes. Of the 6314 comments coded, 4541 were positive and 1773 were negative as specifically related to applicant descriptions of individual residency programs. These comments accounted for more than 1.2 million aggregate views.
Figure 2 shows the percent of positive comments across the 9 specialties. Overall, 72% of comments were positive; however, there were significant differences across specialties. Pediatrics (65%), dermatology (66%), and internal medicine (68%) applicants were more likely to express negative comments compared to the global average, whereas physical medicine and rehabilitation (85%), radiology (82%), otolaryngology (81%), and plastic surgery (80%) applicants were more likely to express positive comments.
Figure 2Percentages (and 95% confidence intervals) of positive comments across specialty type out of a total of 6314 comments; numbers in parentheses indicate the comments in that specialty type. The dashed vertical line indicates the overall proportion of positive comments.
Understanding the preferences of residency candidates can provide guidance to residency programs. Medical students are highly active in social media within the context of rating specific residency programs across numerous specialties, and our results provide a review of positive and negative themes posted by applicants in recent matching cycles. Despite differences in the length of training and nature of training in primary, subspecialty, and surgical specialties, the valued features and themes were largely concordant, including Geography, Clinical Training, and Resident Experience.
Nested within Resident Experience, Likable Residents and Perceived Resident Happiness were repeated themes across several specialties. The Accreditation Council for Graduate Medical Education has made resident wellness initiatives a recent key focus,
and our data suggest that improving the well-being of current residents would positively impact resident recruitment. Geography was the most commonly cited positive and negative theme and also the most important factor in prior studies, including the 2017 NRMP Applicant Survey.
National Resident Matching Program, Data Release and Research Committee, Results of the 2017 NRMP Applicant Survey by Preferred Specialty and Applicant Type. Washington, DC: National Resident Matching Program; 2017.
Although location is intrinsic and immutable, subtheme analysis illuminates how a program could best present itself, a nuance lacking in the NRMP dataset. As an example, programs could address the issues of cost of living, commuting, and cultural activities by highlighting reasonable living options, public transportation or ridesharing, concerts, sporting events, and outdoor activities during interviews and on their websites.
The variation in positive comments by specialty has several implications. Programs with more critical applicants could focus on addressing perceived negative features during interviews. Programs with more positive applicants could focus their attention on accentuating their strengths. Ultimately, the skew toward more positive or negative comments may also reflect applicants of a certain specialty being more critical or more willing to share negative opinions.
There are several limitations to this analysis. The majority of comments originated on a single site, SDN, with the notable exceptions of radiology and otolaryngology. Because comments are posted anonymously, we were unable to control for characteristics of the commenter. Although we cannot say how many unique posters contributed to the comments, the total number of views of each page would suggest that the information shared was disseminated widely. Studies of “lurkers” in social media settings have found that many find just reading the content is enough, suggesting that individuals who view without posting may still be influenced by what they read.
The significant range in the number of comments among the different specialties reflects specialty-specific variability in the use of social media. Finally, our data cannot provide the relative importance of each theme to respondents. Although NRMP provides a weighted average of the importance of a feature, the majority of items fall within a narrow range.
The use of social media comes naturally to millennial learners, who tend to be collaborative, inclusive, and desiring of feedback.
It is not surprising that residency applicants commonly use social media to discuss and obtain peer feedback with regard to potential residency programs. Our results demonstrate that social media can be a useful, unstructured data source that offers some advantages compared to traditional, structured survey approaches. Such data allow for a more stratified understanding of the motivations and desires of current applicants. In medical schools, faculty have already begun to adapt to the needs of millennial learners by soliciting their feedback, leading to the de-emphasizing of traditional lectures in favor of small groups and team-based activities.
Residency programs might consider the online comments of applicants in order to recruit the strongest applicants and adapt to their changing preferences and expectations for post-graduate training. Medical educators themselves should consider participating directly in social media to, for example, dispel rumors, in addition to establishing a safe, anonymous platform for engaging with prospective residents.
Villanti AC, Johnson AL. Social Media media Use use and Access access to Digital digital Technology technology in US Young young Adults adults in 2016. J Med Internet Res. Jun 7 2017;19(6):e196.
Attitudes and practices of surgery residency program directors toward the use of social networking profiles to select residency candidates: a nationwide survey analysis.
National Resident Matching Program, Data Release and Research Committee, Results of the 2017 NRMP Applicant Survey by Preferred Specialty and Applicant Type. Washington, DC: National Resident Matching Program; 2017.
30508-4&id=gr1.jpg){kind=link}
30508-4&id=gr2.jpg){kind=link}