The pipeline starts in medical school: characterizing clinician-educator training programs for U.S. medical students

ABSTRACT In the past forty years, clinician-educators have become indispensable to academic medicine. Numerous clinician-educator-training programs exist within graduate medical education (GME) as clinician-educator tracks (CETs). However, there is a call for the clinician-educator pipeline to begin earlier. This work aims to identify and characterize clinician-educator track-like programs (CETLs) available in undergraduate medical education (UME). We developed an algorithm of 20 individual keyword queries to search the website of each U.S. allopathic medical school for CETLs. We performed the web search between March to April 2021 and repeated the search between July and September 2021. The search identified CETLs for 79 (51%) of the 155 U.S. allopathic medical schools. The identified CETLs commonly address the clinician-educator competency of educational theory (86%, 68/79), are formally organized as concentrations or analogous structures (52%, 41/79), and span all four years of medical school (37%, 29/79). The prevalence of CETLs varies with geography and medical school ranking. We provide an overview of the current state of CETLs as assessed from institutional websites. To create a future with a sustainable output of skilled clinician-educators, UME must continue to increase the number and quality of CETLs.


Introduction
In the past forty years, clinician-educators have become indispensable to the medical education enterprise [1,2]. Clinician-educators comprise the majority of academic physicians [3], fulfill both educational and clinical missions of academic medical centers [4], and receive recognition through distinct promotional tracks [1,5]. A recent consensus definition for the 21st-century clinician-educator [6] highlights the multifaceted nature of the role: an individual who participates in clinical practice, applies theory to education practice, engages in scholarship, and consults on education issues. Indeed, with major national organizations such as the Carnegie Foundation calling for significant reforms to medical training [7], a sustainable pipeline of clinicianeducators who are education scholars -in addition to being exceptional teachers -is necessary to advance medical education [8,9].
However, the current paradigm of undergraduate medical education (UME) rarely teaches students the educator skills essential for the practice of future clinician-educators [6,10,11]. Clinician-educators often must pursue master's or doctoral programs in health professions education [12,13], faculty development programs [14], and/or post-residency medical education fellowships [15]. Over the past two decades, clinicianeducator-training programs have begun appearing in residency programs as clinician-educator tracks (CETs) [16,17]. This emergence may be due to the recognition that residents skilled in clinician-educator competencies are needed to soon join the clinician-educator workforce. Moreover, enhanced recruitment of future clinician-educators in residency may counteract low faculty clinician-educator retention rates [18], thereby bolstering the number of clinician-educators at academic medical centers [16]. This is in addition to the ongoing need to develop resident teaching skills -up to one third of medical student teaching is performed by residents [19] -coupled with accreditation organization requirements for resident teaching [20,21].
Recent literature has called for the clinician-educator pipeline to begin earlier than residency: in undergraduate medical education [22,23]. Cultivating educator identities in medical students may be a solution for generating sustained interest in the clinician-educator pathway during training. Exposure to and identification with educator role models fosters educator identity and produces future clinician-educators, according to the community of practice framework [24,25]. Furthermore, clear career pathways and early research exposure are known to positively influence the choice to become an academic physician [26,27].
The 2020 supplement to Academic Medicine [28] highlights the recent expansion of student enrichment tracks in UME. Other works document the rise in scholarly concentration programs, including those in medical education [29,30]. These emerging formal structures show promise for extending the clinicianeducator pipeline into earlier stages of medical training, encouraging educator identity formation in medical students as CETs do for residents. Our study will refer to these structures for medical students as clinician-educator track-like programs (CETLs).
Existing literature mainly centers on programs that train students to be better teachers [10,31,32]. However, with increasing attention in UME to programs developing the educator skills (e.g., scholarship) and knowledge (e.g., pedagogical theory) vital to the 21st-century clinician-educator [6,[33][34][35], a comprehensive study of the current state of CETLs nationwide is needed. This work aims to identify and characterize CETLs in UME at U.S. allopathic medical schools in order to understand how medical schools prepare students for careers as clinician-educators and to inform future directions for CETL development. repeated the search in July-September 2021. Prior to the search, we obtained the list of U.S. allopathic medical schools from the Liaison Committee on Medical Education (LCME) website [36]. We independently executed the search algorithm for distinct school lists and exchanged schools for the second iteration of the search. In developing the search algorithm, we adhered to established recommendations for conducting systematic web searches [37]. Prior works studying UME scholarly concentration programs [29,38] and reviewing medical school websites [39] informed our search methodology.

CETL definition and exclusion criteria
Using the clinician-educator definition of Sherbino and colleagues [6] and criteria for CETs established by Friedman and colleagues [16], we defined CETLs as programs that target UME learners and address three core clinician-educator competencies: direct teaching, educational scholarship, and educational theory. We categorized CETLs as full or partial depending on the number of criteria met. Full CETLs meet all 4 criteria (longitudinal, direct teaching, educational scholarship, educational theory) whereas partial meet 3 or fewer (Table 1). We established criteria to exclude specialty-and knowledgespecific programs, non-dual degree programs in education, and programs meeting specific CETL criteria that are not substantive enough to constitute a CETL ( Table 2).

Web search algorithm
The web search used an algorithmic approach ( Figure 1). We first located the website for each medical school through the Medical School Admission Requirements (MSAR) online database [40] and confirmed it with a Google search. Entering the website name into the Google search bar ('site:[website name]') limited the search to the institutional website. We conducted a series of queries using 20 keywords, in order, until we identified a full CETL for each school. We chose keywords in consultation with an experienced research librarian (J.M.S.). For each keyword query, we clicked on all links on the first page of Google results and scanned the directly linked web pages for content that fit our CETL criteria.
If an individual keyword query was negative (i.e., did not uncover content consistent with a CETL), we continued through the keyword list until a query was positive or until all keywords were exhausted. When a keyword query was positive, we extracted the following information in a Google form: (1) keyword resulting in CETL hit; (2) website link for CETL; (3) number of CETL criteria filled; (4) type of CETL (full/partial); (5) brief program description; (6) length and timing of program; and (7) explanation of how the program fulfills 'direct teaching,' 'educational scholarship,' and/or 'educational theory' criteria. We Table 1. Definition of a clinician-educator track-like (CETL) program in undergraduate medical education (UME).
CETL Program Definition: a longitudinal (≥4-week) program targeting UME learners that addresses 3 core clinician-educator competencies Direct Teaching Effectively uses scholarly teaching techniques in the clinical and extraclinical environments Educational Scholarship Contributes to the development, dissemination, and translation of health professions knowledge and practice Educational Theory Maintains knowledge of education theory, psychology, and principles, and applies this knowledge to education practice *Full CETL programs fulfill all the above criteria; partial CETL programs lack one or more criteria. Dual degree education programs are included in analyses for schools without a full CETL program also recorded when all keyword queries were exhausted for a school without hitting a CETL program ( Figure 1A).

Data analysis
We reviewed extracted data as a group and calculated Cohen's kappa statistics to measure interrater reliability. Group discussion resolved disagreements and ensured consensus. An independent reviewer (A.W.O.) verified and compiled CETL information for each institution postextraction. We performed descriptive data analysis in Google Sheets and identified medical school subgroups by institution characteristics (geographic region, private/public, ranking). Twosample z-tests for independent proportions assessed for subgroup differences in CETL availability. (1) Specialty-specific programs (2) Programs requiring skill/knowledge prerequisites not shared by all medical students at the targeted level of training (3) Degree programs in education not listed as dual degree offerings (4) Programs meeting the 'longitudinal (≥4-week) program' criterion and/or 'direct teaching' criterion without meeting any other criteria

Web search
We analyzed inter-rater reliability during the two iterations of the search to evaluate reproducibility. We observed a median percent agreement of 87% for 2 raters performing the web search algorithm for the same institutions (range across 5 rater pairs: 80-100%) and a Cohen's kappa of 0.75 (range: 0.59-1.00). This reflects substantial agreement between raters in finding CETLs through the web search [41]. More detailed reporting of inter-rater reliability can be found in Table A1 in the Appendix.
We also examined when CETLs were identified during the execution of the search algorithm. Fortynine percent (39/79) of CETLs were identified after 1 keyword search: 'scholarly concentration.' Ninetyone percent (72/79) of CETLs were identified within 4 keyword searches ( Figure 1B).

CETL identification
The web search identified CETLs at 79 (51%) of the 155 U.S. allopathic medical schools. Thirtynine (25%) have full CETLs; 40 (26%) have partial CETLs, including 3 that also offer a dual degree in education; and 2 (1%) do not have CETLs but offer a dual degree in education ( Figure 2). Individual program summaries and links for all 79 CETLs identified in this study are in Table 3.

CETL program components
We aimed to understand how CETLs differ according to their program components (longitudinal, direct teaching, educational scholarship, educational theory). Of the 79 CETLs, the 39 full CETLs contain all 4 components. The 40 partial CETLs include 26 that have 3 CETL components. Seventeen longitudinal programs incorporate direct teaching and educational theory, 7 longitudinal programs incorporate educational scholarship and educational theory, and 2 longitudinal programs incorporate direct teaching and educational scholarship. Twelve CETLs have 2 components. Eight longitudinal programs incorporate educational scholarship, 3 non-longitudinal programs incorporate direct teaching and educational theory, and 1 longitudinal program incorporates educational theory. Two partial CETLs contain only 1 component (1 educational scholarship program and 1 educational theory program). Overall, 94% (74/79) of CETLs are longitudinal, and educational theory (86%, 68/79) is more prevalent than direct teaching (77%, 61/79) and educational scholarship (72%, 57/79) ( Figure 3A).
Thirty of the 79 (38%) CETLs are designated as electives, the majority of which occur exclusively during the fourth year of medical school (70%, 21/ 30). Three of the 79 (4%) CETLs are research programs. Two of these occur during the summer between the first and second year of medical school. One occurs during the third year. Two of the 79 (3%) CETLs are school requirements (a 4-year Student-As-Teacher program and a month-long Teaching Practicum). The remaining 2 (3%) CETLs represent uniquely branded experiences: a Scholars program and a 4-year course not designated as an elective.
CETLs vary tremendously in length. A plurality of programs (37%, 29/79) span all 4 years of medical school. Eighteen of the 79 CETLs (23%) last 4 weeks and 15 (19%) are between 4 weeks and 1 year. A smaller number of programs are longer than 1 year but shorter than 4 years (10%, 8/79) and even fewer are shorter than 4 weeks (6%, 5/79). Timing of CETLs during medical school is also quite heterogeneous. Although a majority of CETLs start in the first year and continue through the fourth year (37%, 29/79) or occur exclusively during the fourth year (30%, 24/79), programs occur at all points during medical school ( Figure 4).

CETLs by institution characteristics
We additionally analyzed whether CETL program availability is associated with institutional geographic location, public or private status, and ranking among medical schools according to U.S. News & World Report [42].

Dual degree programs
Among the 116 schools without a full CETL, 4% (n = 5) offer a dual degree program for medical students. The dual degree programs encompass 2 general education programs and 3 health professions education programs.

Discussion
Our web search provides an exploratory look at the current state of CETLs in UME. The search algorithm was reproducible, with a median percent agreement between raters of 87% (range across rater pairs: 80%-100%), despite 3-6 months separating the 2 independent searches and program availability changing throughout the time period. Approximately half (51%, 79/155) of U.S. allopathic medical schools have identifiable structures to prepare their students for careers as clinician-educators. Prior works demonstrate that formal training in educational theory, curriculum design, scholarship, and teaching produces better educators [44][45][46] as well as better learners [47]. Medical students soon will be residents and faculty whose educator competencies will benefit trainees and patients alike [47]. These constitute compelling reasons for all U.S. allopathic medical schools to implement -and continually improve -CETLs and dual degree programs in education for their medical students.
Most commonly, CETLs address the clinicianeducator competency of educational theory, are formally organized as concentrations (or analogous structures), and span all four years of medical school. Nevertheless, programs exhibit significant heterogeneity in their naming, timing, and content. Friedman and colleagues [16] report on heterogeneity in the timing of CETs in graduate medical education (GME), proposing that future research examine the type and amount of time required in CETs for sufficient clinician-educator skill-building and professional identity formation. We encourage similar research for CETLs in UME. Applying the community of practice framework [24,25], a greater number of hours devoted to a CETL or dual degree program in education may result in deeper interactions within the medical education community of practice, stronger professional identity formation, and a higher likelihood for medical students to become clinician-educators. Clinician-educators commonly describe their entry into the position as serendipitous and characterize the career pathway as ill-defined [48]. This underscores the critical role that time-intensive CETLs could have in spurring early entry into the clinician-educator career pathway. Future research should explore this role by assessing the long-term impact of CETLs and dual degree programs in education on career trajectory.
With CETLs and CETs increasing in UME and GME, respectively, coordination is necessary to promote continuous development along the educator skill spectrum and reduce curricular redundancy across these two stages of training. Soriano and colleagues [32] found that UME student-as-teacher programs have remarkably similar content and teaching formats to resident-as-teacher programs [49]. A clear framework of educator competencies for both UME and GME may facilitate the alignment of CETLs with CETs to achieve optimal outcomes. The key roles, domains of competence, and core competencies for clinician-educators defined by Sherbino and colleagues [6] may serve as a foundation for institutions seeking to design their own competency frameworks. Standardizing such frameworks would create a shared language for clinician-educator competency building and transform how future clinician-educators are developed. It may therefore behoove educational  leadership at institutions with CETs and CETLs to collaborate on a consensus competency framework for use across UME and GME nationwide. Models of successful CETLs are useful references for institutions aiming to implement their own programs. In recognition of this, our study aims to effectively summarize 79 distinct CETLs representing the diversity of offerings at U.S. allopathic medical schools. Numerous works discuss the challenges of such programs. Faculty doubts about student teaching competence may drive institutional resistance to implementation [32], though students are known to be as good as faculty in certain teaching contexts [50]. It is reasonable to suppose that this capability may extend to other clinicianeducator domains. In their report on a scholarly concentrations program at the Vanderbilt University School of Medicine, Gotterer and colleagues [51] describe additional challenges: deciding on program timing, tracking student progress, and funding program expenses, among others.
Nevertheless, suggestions for actualizing such programs despite their challenges have also been published. Freret and colleagues [52] describe 12 tips for implementing student-as-teacher programs, including defining program scope, promoting longitudinal trainee and faculty relationships, and collecting feedback from all stakeholders. These tips may be adapted to enhance nascent or existing CETLs. We recommend educational leaders draw from the pool of existing CETLs, anticipate challenges to implementing and maintaining their own program, and adhere to published guidance to maximize the chances of CETL success.
While CETLs are most immediately pertinent to medical students, their relevance may begin before medical school. Celebi and colleagues [53] found that implementation of a CET in their residency program was associated with increased applicant interest, larger application numbers, and enhanced applicant quality. We speculate that CETLs may have a similar impact on medical school admissions,  particularly for programs that are prominently featured on medical school institutional websites. The content of residency program websites is known to be important for residency applicants [54], and as most medical school applicants use the Internet as an information source when applying [39], the importance of website advertising likely applies to medical school recruitment as well. In our study, the prevalence of CETLs interestingly varied with institutional geography (more common in the Northern region of the U.S. than the Southeast) and ranking (greater presence at top 50 research institutions than lowerranked/unranked institutions), but not with institutional status as public or private. Medical school applicants aspiring to become clinician-educators may benefit from awareness of these trends when curating their list of prospective schools.

Limitations
Several limitations exist for this study. First, our data do not account for CETLs and CETL components that may be in place but are not evident on institutional websites. Despite this, our CETL sample is large: 79 programs representing 51% of all U.S. allopathic medical schools. This exceeds the number of formal medical education training programs reported in similar works [32]. Our data are also descriptive of various CETL characteristics and are as current as institutional websites, although some websites may be outdated. Second, our web search was not sensitive to program information (e.g., outcomes) typically reported by surveys and case reports. For this work, we desired to have a comprehensive and contemporary picture of institutional CETLs at allopathic medical schools across the U.S. Web surveys are known for low response rates [55,56], and CETLs are sparse in the current literature. Nevertheless, further surveying of educational leadership at medical schools and systematic literature review would add invaluable program information and, importantly, clarify the scope of individual CETLs to enable further analysis. Third, website analysis is inherently subjective, and the personalized algorithms of search engines risk the introduction of selection bias [57]. We employed a comprehensive, algorithmic web search strategy to counteract this, with high inter-rater reliability scores indicating search reproducibility and low rater subjectivity. Finally, our definition for CETL was limited and did not include the many ways medical students may prepare for future clinician-educator careers, such as serving on their institution's curriculum committee and developing courses [11,22]. Future research should explore these opportunities as well as programs that address clinician-educator competencies beyond the scope of our work (e.g., leadership).

Conclusion
In conclusion, we provide an overview of the current state of clinician-educator track-like programs (CETLs) in undergraduate medical education (UME) as assessed from institutional websites. Only half of U.S. allopathic medical schools appear to have formal structures to prepare their students for careers as clinician-educators. To create a future with a sustainable output of skilled clinician-educators, UME must continue to increase the number and quality of CETLs and dual degree programs in education. Future studies should explore CETL outcomes, including end-ofprogram learner satisfaction, demonstrated learner proficiency in clinician-educator competencies, and long-term impact on career trajectory.