Assessment of Perceptions of Professionalism Among Faculty, Trainees, Staff, and Students in a Large University-Based Health System

This qualitative study used a mixed-methods approach to analyze survey and narrative data on how professionalism standards are operationalized and perceived in diverse health care work and learning environments.

RESULTS For the survey, there were 3506 respondents from a pool of 18 550 potential respondents (18.9% response rate). Of 3506 survey respondents, 2082 of 3231 (64.4%) were women, 331 of 3164 (10.5%) identified as gender or sexual minority groups, and 360 of 3178 (11.3%) were non-Hispanic Black individuals. In adjusted analyses, women compared with men (adjusted odds ratio [aOR], 1.8; 95% CI, 1.4-2.3) and Asian individuals (aOR, 2.0; 95% CI, 1.7-2.3) and Hispanic individuals (aOR, 2.0; 95% CI, 1.4-2.7) compared with non-Hispanic White individuals were more likely to value institutional professionalism. In addition, gender identity and sexual minority groups compared with heterosexual respondents (aOR, 1.5; 95% CI, 1.2-1.8) and non-Hispanic Black individuals compared with non-Hispanic White individuals (aOR, 1.3; 95% CI, 1.2-1.4) were statistically significantly more likely to consider changing jobs because of unprofessional behavior at work. The qualitative analysis of narratives revealed that marginalized populations (including but not limited to women, gender and sexual minority groups, racial/ethnic minority groups, those who identify as having a disability, and religious minority groups) reported (1) greater infringements on their professional boundaries, as well as increased scrutiny over their professional actions, and (2) a tension between inclusion vs assimilation.

+ Invited Commentary + Supplemental content Introduction
Professionalism is an important unifying principle in medicine and has been historically described as "the basis of medicine's contract with society." 1(p243) This contract for professionalism gained prominence in the 1990s, [2][3][4][5] when members of the medical field agreed on and promised to uphold a set of "ethical values and competency standards" that "the public and individual patients can and should expect from medical professionals." Thereafter, medical professionalism was implemented as a core competency for undergraduate medical education and graduate medical education to govern how members of the profession conduct themselves in public, be it with patients or each other. [6][7][8] Despite these efforts, the medical field lacks a concise, unifying, and operational definition of professionalism. Professionalism remains a fluid and contextual notion that is thereby often misused or overused, [9][10][11][12] with greater emphasis in practice placed on the delivery of patient care rather than interprofessionalism within the workplace. 13 Professionalism is a historical construct. The term professionalism is often used to describe a broad set of behavior, dress, language, affect, and hierarchies that are deemed appropriate in medicine. Historically, medical institutions did not have women and members of minority groups in positions of leadership and authority, which resulted in professional norms largely being centered around White male identity and the broader cultural norms associated with whiteness. [14][15][16] The current conceptualization of what is deemed professional often does not include diverse groups and thereby can be noninclusive or discriminatory. 14 For example, these cultural norms may result in perceiving the way that certain groups wear their hair or speak as unprofessional. 15 Thus, there is a need to evaluate how professionalism standards are applied in different populations and cultures, as well as to reexamine how professionalism is defined given the increasing diversity of the health care workforce. 16 Consensus is lacking on how to effectively integrate inclusion into developing and operationalizing professionalism within health care work and learning environments. 13 Therefore, the objective of this study was to examine the perceptions and experiences of professionalism among faculty, trainees, staff, and students in diverse health care work and learning settings.

Methods
We conducted a qualitative study using a mixed-methods approach to characterize perceptions of professionalism with data from 2 distinct sources. First, a quantitative analysis of existing survey data 17 was performed, followed by a qualitative analysis of a subset of narrative data collected in June 2017. The University of Pennsylvania Institutional Review Board approved our study protocol and the recruitment language that detailed public use of only deidentified narratives. Participants consented for public use of their anonymized data by submitting a response, as indicated in the instructions given in the email. We followed the Standards for Reporting Qualitative Research (SRQR) reporting guideline in this study.

Independent Variables
The following self-reported respondent characteristics were evaluated: gender identity, sexual orientation, race/ethnicity, position, generational age group, length of employment at institution, disability status, belief system or religion, and primary site of work or study (where one spends most time working).

Dependent Variables
The key outcomes of interest were respondents' perceptions of professionalism in the workplace captured by responses to the following 3 statements: (1) "I have considered changing jobs due to inappropriate, disruptive, or unprofessional behavior by a coworker or supervisor." (2) "I value institutional initiatives, policies, and/or educational resources related to professional behavior in the workplace." (3) "My institution supports a culture of professionalism." The 5-point Likert-type scale response options to the statements were 5 (strongly agree), 4 (agree), 3 (neither agree nor disagree), 2 (disagree), or 1 (strongly disagree).

Statistical Analysis
We examined the association between the self-identified respondent characteristics (independent variables) and each of the 3 professionalism statements. After examining the distributions of the dependent variables and performing sensitivity analyses evaluating the outcome as continuous vs binary, the outcomes were dichotomized for ease of interpretation. Responses were categorized as either affirmative (5 [strongly agree] or 4 [agree]) or not affirmative or neutral (3 [neither agree nor disagree], 2 [disagree], or 1 [strongly disagree]). We fitted multivariable logistic regression models to estimate the associations of respondent characteristics with our binary professionalism outcomes, adjusting for the independent variables specified previously, using generalized estimating equations with robust SEs to account for clustering by site. Two-tailed P values and 95% CIs are reported for all covariates. P < .05 was considered statistically significant.

Qualitative Analysis of Narratives Study Design
To better understand the factors that may be contributing to differences in responses to professionalism questions, we analyzed anonymous narratives collected in June 2017 through a campaign entitled "Please Tell Us Your Story of Inclusion." Across 2 health systems and 4 health professional and graduate schools, faculty, trainees, staff, and students submitted solicited personal narratives (n = 315) in response to 2 stimulus questions designed to assess experiences with professionalism and inclusion at workplace or school. Content analysis was conducted on completed narratives stripped of any identifiers, producing a thematically coded data set. 20,21 Further methods are reported elsewhere. 22

Sampling and Analysis
There was no overt code for professionalism in the parent study 22 ; therefore, we began by identifying narratives that mentioned the term professional or professionalism (n = 13). Then, narratives that mentioned these terms were examined to identify salient clusters of co-occurring themes, yielding the following 3 previously coded themes: hierarchy, interpersonal dynamics, and social norms. Finally, the remaining 302 narratives from the parent study were reviewed to identify other narratives coded for these 3 themes (n = 39), producing a final sample of 52 narratives associated with professional behavior in the workplace or lack thereof.
Throughout a series of in-person meetings, a codebook was iteratively developed until reaching thematic saturation. 21 Next, the research team (D.A.A., M.D.K., and J.A.) reviewed the final group of narratives to identify salient patterns associated with professionalism, its causes, and its consequences. We examined these patterns by respondent characteristics to assess differences by subpopulations. We specifically focused on comparisons between those subgroups that would be classified as marginalized, defined as subgroups systematically excluded from mainstream social, economic, cultural, or political life, with subsequent unequal power in group relationships within society. 23 Marginalized populations would include but are not necessarily limited to women, gender and sexual minority groups, racial/ethnic minority groups, those that identify as having a disability, and religious minority groups.
There were statistically significant unadjusted differences in responses to all 3 professionalism statements by respondent gender identity, sexual orientation, race/ethnicity, and position. Table 3 lists all unadjusted comparisons by respondent characteristics. In response to the statement "I have considered changing jobs due to inappropriate, disruptive, or unprofessional behavior by a co-worker or supervisor," 14.3% of respondents that identified as female, 14.8% of non-Hispanic Black individuals, 15.5% of LGBTQ individuals, and 16.3% of staff agreed in contrast to 7.9% of males (P < .001), 11.2% of non-Hispanic Whites (P = .002), 11.4% of heterosexuals (P = .007), and 10.4% of faculty (P < .001), respectively. In response to the statement "I value institutional initiatives, policies, and/or educational resources related to professional behavior," 52.3% of females, 54.1% of non-Hispanic Blacks, and 54.3% of staff agreed more compared with 45.4% of males (P < .001), 49.0% of non-Hispanic Whites (P = .002), and 47.3% of faculty (P < .001), respectively. Last, in response to "My institution supports a culture of professionalism," 54.2% of females, 58.2% of non-Hispanic Blacks, 54.8% of LGBTQ, and 55.6% of staff agreed in contrast to 50.0% of males (P < .001), 54.2% of non-Hispanic Whites (P < .001), 52.6% of heterosexuals (P < .001), and 49.2% of faculty (P < .001), respectively. Table 3 lists the number and percentage of the remaining respondents for each professionalism question from our unadjusted analyses.
In adjusted models, respondents who self-identified as female, LGBTQ, non-Hispanic Black individuals, and staff compared with male, heterosexual, non-Hispanic White individuals, and faculty were statistically significantly more likely to report considering changing jobs because of unprofessional behavior. For example, gender and sexual minority groups compared with heterosexual respondents (aOR, 1.5; 95% CI, 1. and Hispanic individuals (aOR, 2.0; 95% CI, 1.4-2.7) compared with non-Hispanic White individuals were more likely to value institutional professionalism. No statistically significant adjusted differences were found among respondents who agreed with the statement "My institution supports a culture of professionalism." Table 4 lists the odds ratios and 95% CIs for all of the adjusted comparisons.

Qualitative Analysis
Among the 52 narratives, 33 respondents (63.5%) self-identified as female, 7 (13.5%) as LGBTQ, 28 (53.9%) as non-Hispanic White, and 7 (13.5%) as non-Hispanic Black (Table 2). Two themes emerged from the narratives. Marginalized populations reported (1) greater infringements on their professional boundaries, as well as increased scrutiny over their professional actions, and (2) a      tension between inclusion vs assimilation. In addition, analyses revealed that religious minority groups, specifically those not identifying as Judeo-Christian, struggled in similar ways. We found the same patterns as described above.

Infringements on Professional Boundaries
Many narrators who self-identified as members of marginalized populations expressed greater infringement on their professional boundaries during interactions at work or learning environments.
The infringements reported ranged from microaggressions to blatant racism, sexism, xenophobia, homophobia, and harassment. Whether subtle or aggressive, the infringements on professional boundaries disproportionately impacted women and gender, sexual, and racial/ethnic minority groups ( Table 5) Narratives consistently described that women and minority groups navigating health care work and learning environments were subject to multiple experiences of unprofessionalism because of their identity. One narrator expressed: "I have felt a gender and age bias on several occasions…. When I am working with older men…I am spoken down, disrespected, or outright ignored. I am generally made to feel that my thoughts are invalid or not helpful…."

Scrutiny Over Professional Actions
Many narratives stated that professional standards were applied differently to certain groups, and those groups perceived that they were subject to greater scrutiny. Examples of differential treatment varied in degree ( Yet another narrative stated: "The candidate we knew is a woman of color from a working-class background who has some idiomatic speech patterns and has mostly learned in the job, while the candidates we were sent with a real push to hire instead were White, from middle-class backgrounds with speech patterns of that cultural group and more formal education (but less relevant on-theground experience). It gave a strong sense that looking/speaking like our current entirely White administration was valued over the known skills of our candidate."

Theme 2: Tension Between Inclusion vs Assimilation
We found that respondents perceived the need to assimilate into their environments as opposed to the workplace adopting an inclusive environment. Many respondents perceived pressure to conform to the dominant norms of their work environment or their positions would be at stake. This pressure often meant adhering to norms associated with White identity. Words frequently used in the narratives were "fit," "pressure," and "individuality." For example, a respondent stated how there was a "strong push toward uniformity at the expense of individuality and the energy that comes from the interaction of people of different backgrounds." Narrators expressed that they did not perceive being welcomed for their differences but rather Another narrator expressed: "Many internationals compromise personal dignity and values so they don't get kicked out, but ultimately it only promotes intolerance and behavior aimed at silencing their voices and opinions rather than making them feel included and welcome" ( Table 5).
Narratives also highlighted the tension between perceiving pressure to conform vs leaving the institution. For example, one narrator expressed: "As a queer person of color…[I am mostly made to] feel 'other' in one way or another! That is one of many reasons why I will be leaving soon…it's too much." Such external pressure to conform often manifested in reported disciplinary action. One narrative stated: "I was called into a standard/probation meeting on a Muslim learner….
[I]t was commented on the case that the individual 'would not even eat our foods' as a reason to question their academic performance…as 'the learner only ate halal foods'." Narratives consistently described use of the word "fit" to suggest that individuality was not appreciated or sought out. For example, a respondent stated that the institution is "inclusive if you fit the mold of what is expected." Another noted: "First I watched the residency director tell their faculty colleagues that a resident of color was just a bad fit in their residency. They didn't say why, just a bad fit."

Discussion
Professionalism in medicine has been championed as a contract between all members of the profession that governs the membership's public conduct, be it with patients or each other. [1][2][3][4][5] The objective of this mixed-methods analysis was to examine the perspectives of a diverse set of faculty, clinical providers, trainees, staff, and students to assess if there is a need to revisit this contract. We found that individuals self-identifying as female, underrepresented racial/ethnic minority groups, 24 gender and sexual minority groups, and staff value professionalism equally or even greater than their counterparts, yet they perceive or experience it differently within their organizations. Our findings highlight the tension between inclusion and assimilation. The need to conform to a set of noninclusive professional standards emerged as a key challenge for women and minority groups (gender, sexual, and racial/ethnic) alike. Seminal work in understanding the nature of belonging or inclusion by  describes inclusivity as the opposite of conforming or assimilation. By this definition, professionalism standards are not inclusive when underrepresented members of a profession have to assimilate or conform in an effort to adhere to them. The concept of covering has also been used to describe this same notion. Covering, a word coined by Erving Goffman, details how individuals with known stigmatized identities perceive greater pressure and put more effort into downplaying those identities to assimilate better into majority groups. 40,41 Understanding how and why women and minority groups cover and assimilate can help break down barriers and lead to a culture of inclusion and authenticity.
In addition, findings of the present study suggest that the medical field must revisit and expand professionalism standards to ensure that the increasingly diverse membership helps to develop and agree on a "belief system in which group members declare shared competency standards and ethical values they promise to uphold in their work." 3 We believe that such competency standards and ethical values must include fair treatment of patients and colleagues regardless of position or personal characteristics, such as gender identity, sexual orientation, and race/ethnicity. However, this goal is not enough. Health care organizations and medical institutions can also engage their diverse membership to ensure professionalism standards are equitably applied.
Professionalism and the development of associated standards were devised with a narrow lens given the historical homogeneity of academic medicine. 26,[42][43][44][45] As a result, there is some evidence that standards are vulnerable to being mobilized with a much greater degree of strictness among marginalized groups. 22 The concept of professionalism itself can center norms associated with a singular majority culture and inadvertently marginalize minority groups. The findings of the present study reveal how medical professionalism can be displayed through the unequal application of standards, also referred to as being "weaponized." This weaponization, as noted by Lee, 46 can damage individuality and identity and cause conformity, shrink efforts to increase the diversity of the workforce, create harmful environments, and force marginalized populations to assimilate in an attempt to be included. 40,41 Health care professionals, organizations, and medical institutions can ensure that the standards collectively set to govern conduct with patients and each other are not weaponized to pressure members of our profession to assimilate.
An increasing number of studies report that women and minority groups have higher attrition rates than majority groups. 30,[47][48][49][50][51][52][53][54][55][56][57][58][59][60][61] The findings of the present study build on this work by revealing why marginalized groups may seek employment elsewhere. Departments and divisions have their own unspoken rules, norms, policies, and leadership styles that can create different daily workplace cultures within an institution. 62 This study supports the need for institutional policy adjustments pertaining to professionalism. There remains a need to create environments where individuals are comfortable to be themselves and express a range of cultural practices and traditions. Such inclusive environments may help mitigate factors, such as avoidance or the persistent need to conform, that if left unchecked can lead to attrition, disillusionment, and less organizational commitment, in addition to the personal trauma of discrimination. 22,[63][64][65][66]

Strengths and Limitations
A notable strength of this study is the application of a mixed-methods approach to further understand the intersection of professionalism and inclusion. We acknowledge some limitations to this study. The findings may not be generalizable to other study settings. However, our data are from a diverse regional set of hospitals and health science schools in New Jersey, Pennsylvania, and Delaware. The quantitative survey assessment of respondents' belief system or religion was limited to 2 categories (Judeo-Christian or not); however, the qualitative assessment captured a broader array of categories. In addition, the survey administration method allowed for individuals present on more than 1 listserve to be counted more than once in our sampling population and may have