Abstract
Purpose:
Many US medical schools have adopted learning communities to provide a framework for advising and teaching functions. Faculty who participate in learning communities often have additional educator roles. Defining potential conflicts of interest (COIs) among these roles is an important consideration for schools with existing learning communities and those looking to develop them, both for transparency with students and also to comply with regulatory requirements.
Methods:
A survey was sent to the institutional contact for each of the 42 Learning Communities Institute (LCI) member medical schools to assess faculty opinions about what roles potentially conflict. The survey asked the role of learning community faculty in summative and formative assessment of students and whether schools had existing policies around COIs in medical education.
Results:
In all, 35 (85%) LCI representatives responded; 30 (86%) respondents agreed or strongly agreed that learning community faculty should be permitted to evaluate their students for formative purposes, while 19 (54%) strongly agreed or agreed that learning community faculty should be permitted to evaluate their students in a way that contributes to a grade; 31 (89%) reported awareness of the accreditation standard ensuring “that medical students can obtain academic counseling from individuals who have no role in making assessment or promotion decisions about them,” but only 10 (29%) had a school policy about COIs in education. There was a wide range of responses about what roles potentially conflict with being a learning community faculty.
Conclusion:
The potential for COIs between learning community faculty and other educator roles concerns faculty at schools with learning communities, but most schools have not formally addressed these concerns.
Faculty who advice and counsel medical students in official or unofficial capacities may encounter situations when educational and administrative activities potentially conflict with their roles as advisors. For example, faculty charged with advising students may serve as course or clerkship directors or may directly assess medical student’s progress. Roles often inform one another and that may have benefits. 1 A teacher who has made firsthand observations in a clinical setting can offer specific and impactful insights to a student, such as using knowledge of the student’s strengths and weakness to recommend electives to pursue or specialties to consider for future training. However, the dual role of advisor and assessor may discourage students from seeking out assistance from this person, based on students’ concern that sharing sensitive information could potentially negatively impact assessment of their academic performance or adversely affect recommendations for future training. In one study, students with advisors in clinical specialties who also served on a residency selection committee expressed discomfort with the advising relationship and admitted to being more likely to make misleading statements about training program preferences than if the advisor had no role in resident selection. 2 Likewise, some have argued that student affairs officers should not oversee the preparation of the Medical Student Performance Evaluation (MSPE) document that compiles assessment data for graduating students. The concern raised is that such an arrangement may discourage students from seeking out “sensitive health services” if the student affairs office is the conduit by which students access those resources. 3
The Liaison Committee on Medical Education (LCME), the accreditation organization for US and Canadian medical schools, has published standards proposed to help schools serve both the interests of the general public and of enrolled medical students. Many of the standards are also meant to prevent conflicts of interest (COIs) for medical educators. One particular LCME Standard 11.1 relating to COI stipulates the following: A medical school has an effective system of academic advising in place for medical students that integrates the efforts of faculty members, course and clerkship directors, and student affairs staff with its counseling and tutorial services and ensures that medical students can obtain academic counseling from individuals who have no role in making assessment or promotion decisions about them. (Italics added)
4
Meeting this standard can be challenging, particularly if there are a limited number of faculty members qualified to conduct the teaching mission of an academic medical center. Schools outside of the United States and Canada that do not have a related accreditation standard may also have an interest in navigating COIs for medical educators as academic advising is an important element for all higher education institutions.
Many US medical schools have adopted learning communities to provide a framework for advising and teaching functions. Learning communities are defined as “intentionally created group of students and/or faculty who are actively engaged in learning from each other.” 5 A 2012 survey of 151 US medical schools found that 66 had learning communities and 29 were considering instituting them, 5 and a more recent survey of medical students reported that 102 of 140 schools (72.9%) had learning communities. 6 Learning community structure and function vary considerably across schools. Most have both advising and curricular elements that emphasize wellness, connectedness, professional identity formation, career planning, and clinical skills development to varying degrees. Learning communities have been associated with student perceptions of a more positive learning environment 6 and increased connectedness among students. 7 The growing number of schools with learning communities, and interest in developing them, has led to the development of the Learning Communities Institute (LCI), 8 a multidisciplinary organization of medical school faculty, administrators, and students promoting and studying the concept of longitudinal relationship-centered education across the continuum of health sciences learning.
As learning communities blend advising and curricular elements, many learning community faculty have both advisory and assessment responsibilities. Indeed, a recent survey found that 38% of schools have learning community faculty who are involved in formal assessment or grading. 5 These faculty often possess additional advising and assessment responsibilities within the medical school but outside the learning community sphere, creating potential COIs across a spectrum of educator roles. Considering potential COIs among these roles is important for schools with existing learning communities and those planning to develop them, both to maximize transparency with students and to comply with accreditation requirements.
The goals of this study are to (1) to assess the opinion of learning community faculty as to what constitutes COIs in medical school educator roles and (2) to assess whether schools with learning communities have existing processes and procedures that currently address such COIs. This study is important because it focuses on the under-studied area of potential COIs between educator roles and not on the better described area of teacher-student relationships and personal boundaries.9,10
Methods
Participants
Targeted participants were 41 designated institutional contact members at each Learning Communities Institute (LCI) medical school. The institutional contact members were usually involved with their learning community in a leadership capacity as a dean, learning community director, or learning community leader.
Survey design and content
A 25-item survey was developed to ask whether learning community faculty should assess students, either in a formative fashion or in a way that contributes to a grade, and whether being a learning community faculty conflicts with other educator roles common to medical schools (course director, clerkship director, promotions committee member, etc). The survey was informed by a workshop on COIs in medical education conducted at the 2015 Annual Learning Communities Institute meeting and had input from LCI leadership. For the purposes of this study, a COI was defined as any real or perceived discrepancy that exists between roles and/or responsibilities that a faculty member may have regarding medical students, either singly or collectively. In addition, the survey asked participants whether they were aware of the LCME Standard 11.1 and whether their schools had existing policies concerning COIs in medical education.
The Institutional Review Board at the Icahn School of Medicine at Mount Sinai reviewed the study protocol and deemed it exempt from further review or monitoring.
Survey administration
The survey link was sent by email in April 2016 to the designated institutional contact of each LCI member school. Annonymous survey responses were collected using the REDCap electronic data capture tool. 11 Three reminders were sent through REDCap, and then, study authors directly emailed members who had not completed the survey. No compensation or incentive was offered for completing the survey.
Data analysis
Frequencies and percentages for demographic and COI items were computed. A “0” was assigned to no conflict, “0.5” to possible conflict, and “1” to definite conflict for each of the 17 role-comparison items. These numerical ratings were averaged across the 17 items and multiplied by 100 to create a percent COI score, and any item with a “not sure/no opinion” rating was omitted for each respondent. Conflict of interest percent scores closer to 100% represent more perceived COIs between the role of learning community faculty and the other role types. Independent sample t-tests were used to compare COI scores between respondents (1) who strongly agreed/agreed to those who disagreed/strongly disagreed that learning community faculty should be permitted to assess their students in a way that contributes to a grade, (2) who were aware of the LCME standard to those who were not aware of the LCME standard, and (3) who had a COI policy at their school to those who did not. All data were analyzed in SPSS, version 23 (Armonk, NY: IBM Corp.) Survey comments were coded by 3 investigators (PG, KM, and JC-G), and grounded theory was used to identify themes. Any disagreement in coding was discussed until consensus was reached.
Results
Complete survey data were available for 85% (35 out of 41) of all LCI institutional contacts. Most of the respondents held the role of learning community faculty or learning community leader (40%, 14); 23% (8) of respondents held the dual roles of student affairs dean and learning community faculty or learning community leader, while an additional 26% (9) were student affairs deans only. The remaining respondents were 3 clinical faculty members (9%) and 1 dean (3%).
Table 1 provides frequencies of respondents’ COI ratings for 17 role types. Most of the respondents (86%, 30) strongly agreed or agreed that learning community faculty should be permitted to assess their students for formative purposes. However, only 54% (19) strongly agreed or agreed that learning community faculty should be permitted to assess their students in a way that contributed to a course grade. These 19 respondents had an average COI score of 37% (SD = 18%), which was not significantly less than the average COI score (46%, SD = 14%) for the 13 respondents who did not agree with the statement, t(30) = 1.37, P = .180. Three respondents had no opinion and were omitted from the comparison.
Medical education roles that may conflict with LC faculty role.
Abbreviations: COI, conflict of interest; LC, learning communities.
Most of the respondents (89%, 31) were aware of LCME Standard 11.1, while 11% (4) of the respondents were not aware. These 4 respondents who were not aware of LCME Standard 11.1 had a significantly lower COI score (17%, SD = 16%), suggesting less perceived COI for role types than the 31 respondents who were aware of the LCME standard (43%, SD = 17%; t(33) = 2.94, P = .006, 5% confidence interval [CI] of 26% difference = 8%-45%).
Most of the respondents (86%, 30) were currently engaged in discussions about COI at their school, with 6% (2) not engaged and 9% (3) unsure if discussions were occurring at their school. In all, 10 respondents (29%) already had a COI policy at their school, with most respondents (37%, 13) unsure whether they had a policy at their school and 34% (12) not having a policy. The average COI score for those with a COI policy (48%, SD = 19%) was not significantly different than the average COI score for those who did not have a policy (39%, SD = 14%, t(20) = 1.26, P = .222).
In all, 11 (31%) responded to the open-ended item, and there were a total of 13 comments. Three main themes emerged from the coding. Most of the comments (6 out of 11) were about ways to mitigate potential conflicts. Some respondents wrote that potential conflicts can be handled by having alternate faculty for specific tasks or for specific students. (“If as a course director or clerkship director, there is a committee or group such that you would not be in a position to assess your own mentees, that is, a possible way to get around this conflict.”) Other roles or obligations may be more difficult to manage. (“Advisors should not be treating physicians or mental health providers to students. Advisors shouldn’t be disability officers or ombuds-officers for students.” “I think that role where an individual (research mentor) interacts with a mentee, and there are no others who can act in this assessment role presents a conflict with the [learning community] mentor role.”) Three (out of 11) comments concerned the difficulty or undesirability of avoiding situations where learning community may have dual roles. (“Because we have a small faculty, it is difficult to exclude all assessing faculty members from student support roles.” “[T]he notion that some of the people at a medical school who are most knowledgeable about how to advise students cannot do it because of some theoretical conflict will ultimately lead to suboptimal advising for students.”) Finally, 3 (out of 11) comments described the difficulty of generalizing potential conflicts in roles when schools structure learning communities in such different ways. (“In some schools, [learning community] faculty have only an advising role, while in other schools, [learning community] faculty teach students in the curriculum. Therefore, it seems potential that COIs would depend on how the individual school structures its LC.”)
Discussion
Medical educators have raised concerns about the potential for COIs among faculty who are tasked with student advising and who also share a role in student assessment. 3 In addition, the LCME has made clear that US and Canadian medical schools must have advisors available who have no role in assessment. 4 The limited evidence from learners’ perspective shows that learners are less comfortable with advisors who have a role in assessment than those who do not and that such an assessment role can influence a learner’s willingness to be open and honest in an advising relationship. 2 This is the first study we are aware of that examines opinions of faculty leaders on the perceived COIs among medical educator roles and the acceptability of assessing students within the context of learning communities.
Our results demonstrate that faculty involved with learning communities are concerned about COIs. The vast majority reported awareness of LCME Standard 11.1 and discussions by their schools about how to handle potential COIs. Despite this concern and discussion, less than a third of respondents reported that their schools definitely have a policy on COIs. Consensus on what kind of COIs are important to monitor for, and mitigate if possible, may help guide both schools with learning communities and other institutions with various advising practices.
There was a wide range of responses about what roles potentially conflict with being a learning community faculty. Although student mental health director/faculty had the highest percentage of respondents perceiving a possible or definite COI (83%), it is surprising that all respondents did not perceive a COI in this setting, given there are additional LCME standards about this specific role. 4 Likewise, respondents were not unanimous in their opinions about student health director/faculty, and/or disability officer, all roles that have access to highly personal and sensitive student information. A quarter of respondents did not feel a role on a promotions committee posed a conflict. Conversely, all roles listed in the survey have some perceived COI. Does this mean that there is implicitly some level of COI in all potential relationships? If so, where should the line be drawn? The variabilty in responses likely reflects the heterogenity of learning communities in their focus and how they are structured.
Most respondents felt that learning community faculty can assess medical students and are more comfortable with giving formative assessments than assessments that contribute to a grade. There were no statistically significant differences between respondents’ perceptions of COIs averaged among the 17 role combinations and (1) their opinions about learning community faculty assessing students in a way that contributed to a grade or (2) whether their schools were engaged in discussions about COIs. Learning community faculty who were not aware of Standard 11.1 had a significantly lower COI score compared with members who were aware of the standard. Again, the structure and function of the learning community could have influenced how respondents answered these questions. For example, a learning community model used for delivering a clinical skills course versus for advising may have affected the faculty’s perceptions of COIs and opinions about learning community faculty assessing students. Those learning community faculty who teach clinical skills might consider assessment as more inherently part of their role and thus have a higher threshold for perceiving COIs, whereas learning community primarily tasked with advising students may have a lower threshold for perceiving COIs.
There are limitations of this study. The survey was sent to representatives of the Learning Communities Institute member schools. This may not be representative of the larger number of schools that have learning communities but are not members of this organization. However, the schools in the LCI represent a broad spectrum of schools in size, geographic location, and public and private status. 6 The survey only solicited a response from the institutional contact and not from all learning community faculty at a given school. Surveying all learning community faculty may have allowed for a larger set of responses and advanced analyses, but it would have been difficult to ensure an appropriate response rate for generalization. In addition, there are limitations with the survey as we did not ask about features of a school’s learning community, such as structure, focus, or curricular elements. These features could have influenced answers to the questions about assessing students and the opinions on COIs among roles as noted above. The survey also did not ask about whether conflicts could be mitigated, and strategies for mitigation, which could have been useful to derive potential best practices.
In summary, the potential for COIs between learning community faculty and other educator roles clearly concerns those schools with learning communities, but little has been done to formally address these concerns at most schools. Knowledge of accrediation standards may influence the way faculty perceive COIs, but little consensus exists about what roles have the potential to conflict. Given the level of student distress, mental illness, 12 burnout, 13 and mistreatment, 14 it is important that we are mindful of institutional stuctures and practices that allow students options for support and to alleviate their potential concerns about whether seeking help could impact their assessment. As this was a survey of faculty perceptions, further discourse on this topic will benefit from incorporating perceptions of learners. Studying students’ perceptions of COIs and the impact of such perceptions on their practices for seeking out advice and assistance will be instrumental in further developing best practices in this area.
Footnotes
Declaration of Conflicting Interests:
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding:
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author Contributions
PG: Study design, data analysis, manuscript draft.
JCG: Study design, data analysis and statistics, edit manuscript.
KM: Study design, data analysis, edit manuscript.
MB, WC, SE, AF, MK, LO, MP: Study design, edit manuscript.
