COUNTERPOINT: Should Fellowship Interviews Remain Exclusively Virtual? No.

There are few aspects of our daily lives that the COVID-19 pandemic has not upended. This was certainly true in the medical education world as the 2020 recruitment season approached and the world was still in the grips of a deadly virus. The traditional interview process was suddenly unthinkable. To protect applicants and abide by the travel restrictions, the Coalition for Physician Accountability’s Work Group recommended a shift to virtual interviews (VI), a process novel to both candidates and programs.

The adoption of VI in 2020 demonstrated that the process was feasible and could be implemented at scale. Virtual interviews also come with advantages, which include a decreased carbon footprint resulting from less travel and financial savings to both programs and applicants.

However, VI are not the panacea that they appear to be, and it would be ill advised to exclusively adopt them in the near future without carefully considering the downsides and possible repercussions for both candidates and programs.

Fit

A major purpose of the interview for fellowship is finding a good fit. Fit is defined as the compatibility between an individual and a work environment that occurs when their characteristics are well-matched. Candidates are well aware of the importance of fit and cite it as one of the top factors in ranking a program. There are four types of fit that have been shown to have moderate and strong correlation with the important long-term outcomes of job satisfaction, overall performance, intent to quit, and indicators of strain: (1) person-job fit, (2) person-organization fit, (3) person-group fit, and (4) person-supervisor fit (Table 1 ). Person-job fit assesses the competencies and cognitive skills that make a candidate well-prepared for the job. It is easily assessable by reviewing the application, letters of recommendation, and prior experiences and is readily amenable to the virtual interview process. The other three types of fit consist of intangibles that are harder to evaluate through VI. Person-organization fit assesses the congruence between a candidates’ personality, attitudes, and goals and those of the program. Person-group fit assesses the interpersonal compatibility of the candidate with their work group (fellows, faculty, and other members of the health care team). Person-supervisor fit assesses congruence of values, personality, and goals between a person and the supervisor, in this case the program director, division leaders, and other key clinical faculty members. All three types of fit require a wider, less structured contact with the organization and more casual interactions with its people than can be afforded by the virtual space to assess adequately. In addition, the virtual platform does not really allow one to meaningfully observe the interactions among fellows and faculty (the casual banter, the malignant whisper during a conference, or the supportive smile during a presentation), and prevents one from getting a good “feel” for a place. In a medical specialty such as Pulmonary and Critical Care Medicine, where burnout is climbing, it is our responsibility to ensure that applicants have the ability to optimally assess their own fit within a program before committing. This is especially important for under-represented in medicine candidates who depend on the interview day to assess how comfortable they feel in a place and to note any microaggressions.

Table 1

Correlation of Four Types of Fit With Long-term Outcomesa

Outcome	Person-Job Fit	Person-Organization Fit	Person-Group Fit	Person-Supervisor Fit
Job satisfaction	+++	+++	++	+++
Overall performance	+	++	++	++
Intent to quitb	++	+++	++	No data
Indicators of strainb	++	++	No data	No data

+++ = strong correlation; ++ = moderate correlation; + = poor correlation.

Data summarized from Kristof-Brown et al.

Inverse correlation.

Perceptions About Virtual Interviews

Several studies have assessed program directors’ and applicants’ perceptions of VI. Although most agree that VI were easy and convenient and should continue to be used in the future, the majority of candidates and program directors did not want to see VI used as the only modality and would prefer to see it used as a screening tool. , Most applicants (61%) reported that VI did not allow for an accurate representation of the program and was not sufficient to select a program. Compared with in-person interviews, applicants felt that VI were less conducive to understanding the culture of a program (100% vs 64%), and only 54% of applicants felt that VI were sufficient to make ranking decisions (vs 92% for in-person interviews). In a study from a surgical program that used VI for 3 years (2015 to 2017) while others were still using the traditional format, 34% of applicants felt that VI had a negative impact on their ranking of that program. This concerning result raises the issue of program ranking inequities, favoring in-person interviews, if some programs chose to exclusively use VI while others do not.

Geographic Location

The most obvious downside of VI is the inability to visit the hospital and training facilities and to explore the town where one could be living for the next 3 to 4 years. Geographic location was the top-ranked factor considered by applicants when ranking programs. Virtual hospital and city tours, while an acceptable alternative during a pandemic, are hardly a replacement for the experience of exploring a city and surrounding areas, which are needed activities to form an impression of whether one could be happy living there.

Soft Skills

When it comes to ranking interviewed candidates, four of the top 10 factors cited by Pulmonary and Critical Care Medicine program directors are related to soft skills, such as interactions with faculty and house staff, interpersonal skills, and feedback from current fellows. This is not surprising, given that interpersonal and communication skills, as well as professionalism, represent two of the six core competencies expected from medical residents by the Accreditation Council for Graduate Medical Education and are the most prevalent problematic issues that result in termination. They are also the hardest to remediate. The interview day remains the main avenue by which programs evaluate these skills. Switching to a virtual process, which significantly limits interactions, can take away an important tool that programs depend on for candidate selection.

Interview Allocation Disparities

Because of the minimal financial and time commitments needed to go on virtual interviews, some applicants chose to apply to more programs in 2020. For the same reason, there was a drop in the number of interview cancellations (unpublished data). These trends could lead to interview allocation disparities when programs interview their “first pass” candidates, who are more likely to be the traditional high achievers, at the cost of diversity, equity, and inclusion. The solution to this would be for programs to increase the number of interviews that they conduct, which would negate the time-savings of VI and increase interview stress for all.

Conclusion

Most of us would feel uncomfortable taking up a mortgage to buy a house that we have only toured virtually. We would much prefer to use a virtual platform to tour several houses, narrow down the choices, then visit a few before picking the perfect one.

Similarly, the choice of fellowship training is a high-stake decision for trainees; one that will have repercussions on their entire career. As program leaders, we owe them a fair but flexible interview model that takes into consideration their preferences and allows them to gather all the information they need to make one of the most consequential decisions of their careers. The exclusive use of VI does not fit that model. Instead, programs should take the time to design an equitable interview format that not only uses the benefits of virtual interviews but also incorporates in-person visits for those who need it.