Perceptions of Factors Associated With Inclusive Work and Learning Environments in Health Care Organizations: A Qualitative Narrative Analysis

Key Points español 中文 (chinese)

Question  What factors matter for creating inclusive clinical and learning environments within health care organizations?

Findings  This qualitative analysis identifies 6 modifiable contributors to an inclusive culture and the implications of these factors on the well-being and performance of members of the health care workforce.

Meaning  Understanding these factors provides health care systems a way to improve inclusivity and facilitate the retention of a diverse health professional student body and health care workforce to ultimately improve care delivery.

Importance  Diversifying the health care workforce remains a critical goal for health care organizations focused on reducing disparities in care. However, it remains unknown what factors create inclusive health system environments and help organizations retain a diverse workforce.

Objective  To understand from members of the health care workforce what factors contribute to inclusive work and learning environments and what can be done to improve inclusion within health care organizations.

Design, Setting, and Participants  A qualitative narrative analysis of responses to a weekly email call for narratives within health care organizations sent June 1, 8, 15, and 22, 2016. The email contained an anonymous link to 2 open-ended stimulus questions asking for stories reflecting inclusion or lack thereof within participants’ work environments as well as demographic questions. The study took place at 6 hospitals, including a free-standing children’s hospital and a Veterans Affairs medical center, 4 health sciences schools (Medicine, Nursing, Dental, and Social Policy and Practice), and outpatient facilities within a university-based health care system in Pennsylvania. There were 315 completed narratives submitted from health care system executives (n = 3), staff (n = 113), academic faculty (n = 97), trainees or students (n = 99), and 3 who declined to specify their positions.

Main Outcomes and Measures  Workplace experiences with inclusivity, implications of these experiences, and recommendations to improve inclusion within environments.

Results  Of 315 narratives submitted from members of the health care system, in 188 (59.7%) the writer self-identified as female; in 10 (3.2%), as transgender/queer; in 38 (12.1%), as non-Hispanic black; in 152 (48.3%), as non-Christian; in 31 (9.8%), as having a language other than English as their primary language; and in 14 (4.4%), as having a disability. Analysis of the narratives revealed 6 broad factors that affected inclusion within health care organizations: (1) the presence of discrimination; (2) the silent witness; (3) the interplay of hierarchy, recognition, and civility; (4) the effectiveness of organizational leadership and mentors; (5) support for work-life balance; and (6) perceptions of exclusion from inclusion efforts. Challenges with inclusion had negative effects on job performance and well-being, with reports of stress, anxiety, and feelings of hopelessness. Most respondents referenced a systemic culture that influenced their interpersonal dynamics and provided specific strategies to improve organizational culture that focused on leadership training and expanding collegial networks.

Conclusions and Relevance  This narrative analysis provides a taxonomy of factors that health care organizations can use to assess inclusion within their learning and work environments as well as strategies to improve inclusion and retain a diverse health care workforce.

Diversifying the health care workforce remains a critical goal for many health care organizations focused on improving clinician education and reducing disparities in access and care.^1-7 Populations with equal access to care experience disparities in treatment despite well-intentioned efforts because of the structural bias ingrained in our health care systems and health care professionals’ implicit biases.^8-11 Prior evidence demonstrates that the engagement of a diverse workforce reduces such biases and improves the cultural competencies of nonminority and minority health care professionals alike.^1,8,12 Enhancing diversity brings together distinct minds with varying perspectives, backgrounds, and experiences, improving the way we generate medical knowledge and deliver care.^8,13-15 Prior solutions for a diverse workforce have centered on recruitment and not retention,^2,16 despite evidence of greater attrition among women and minorities.^17-20 Organizational efforts that focus on creating an inclusive environment may promote greater retention of a diverse workforce, reduce the costs related to attrition, and ultimately affect patient satisfaction and care quality.²¹ However, there is a paucity of research on how health care organizations create a culture that promotes inclusive environments to achieve these goals. Prior work has conceptualized inclusion as a set of social processes that influence an individual’s access to information, sense of belonging and job security, and social support system.²² Health organizations lack sufficient understanding of the operational definition of inclusion to guide their efforts to improve culture. To address this gap, we conducted a narrative analysis to understand from members of the health care workforce what factors affect perceptions of inclusion in a health care system.

To conduct this narrative analysis, we used a culturally diverse study team that included 3 student researchers, a research coordinator, and 3 faculty members. We solicited anonymous narratives from employees, faculty, and students about their experiences with inclusion at 6 hospitals, including a free-standing children’s hospital and a Veterans Affairs medical center, and 4 health schools (Medicine, Nursing, Dental, and Social Policy and Practice) within a university-based health system. Administrators sent a weekly email call for narratives on June 1, 8, 15, and 22, 2016, to listservs representing different constituents within and across organizations affiliated with the academic health system. This email contained an anonymous link to an inquiry using REDCap,²³ consisting of demographic questions and 2 open-ended stimulus questions designed to evaluate participant interpretations of inclusion:

1. Think about a time when you witnessed or participated in a situation where you or a colleague/member of [this organization] was treated in a manner that made you/them feel either included, valued, and welcome OR excluded, devalued, and unwelcome as a member of this [organization].

2. Please comment on your perception of the general climate at the [organization] with regards to inclusion and respect.

Respondents were instructed to avoid use of personal identifiers and were promised anonymity. We flagged and redacted any narratives with identifiable information. Some initiated a narrative but did not submit (n = 1270), and others submitted incomplete narratives (n = 47). All data from the responses were stored in REDCap with access limited to the study team. 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 Standards for Reporting Qualitative Research (SRQR) reporting guidelines in reporting this study.

We analyzed data from all submitted narratives with completed responses to both open-ended questions (n = 315). We evaluated the narratives with a focus on both structure and content, using the Labov and Waletzky model for narrative analysis.²⁴ We coded core features of each narrative (introduction, presenting problem, complicating factors, resolution, moral or meaning, characteristics of persons involved) to facilitate comparisons across narratives. We jointly analyzed a subset of narratives (n = 30) to identify emerging patterns in the data and develop a codebook, with explicit definitions and examples to ensure coding accuracy and facilitate intercoder reliability.²⁵ Three research assistants independently coded the remaining narratives (n = 285), with each member coding an unique sample (n = 84) as well as a shared sample (n = 33) to assess intercoder reliability. All coding discrepancies were resolved by group consensus. We used NVivo 11²⁶ for all data management and coding. Our intercoder reliability using the κ coefficient²⁷ revealed acceptable agreement among coders (mean [range] κ, 0.93 [0.76-1.00]). We organized codes into themes and subcoded further to characterize dimensions of themes by participant attributes and level of interaction (individual and/or interpersonal, group, and system).

Of the 315 completed narratives submitted, 3 (1.0%) were from health care system executives, 113 (35.9%) from staff, 97 (30.8%) from academic faculty, 99 (31.4%) from trainees or students, and 3 (1.0%) from participants who declined to specify their positions. Only 48 participants (15.2%) reported being at the institution for less than 1 year, compared with 91 (28.9%) who reported being at the institution for 1 to 5 years, 64 (20.3%) for 5 to 10 years, and 107 (34%) for more than 10 years. More than 90% of participants provided their demographic information, and 188 (59.7%) self-identified as female, 10 (3.2%) as transgender/queer, 38 (12.1%) as non-Hispanic black, and 152 (48.3%) as non-Christian. Also, 31 (9.8%) reported a primary language other than English and 14 (4.4%) reported having a disability (Table 1).

The submitted narratives varied in content but were similar in structure. Most responses detailed a presenting event, reactions, and conclusions. The median (interquartile range) character count of narratives submitted in response to stimulus question 1 was 377 (600); for question 2 it was 66 (90). Despite an online process with assurances of anonymity, some responded that they feared sharing their story, while others often sandwiched their negative experiences with positive statements. Both positive and negative examples of inclusion and lack of inclusion supported the themes we present in the following sections. We aimed to provide quotes most representative of the theme, irrespective of valence.

Six broad factors emerged from our analysis to form the basis of a taxonomy that characterizes perceptions of inclusion or lack thereof within health care systems: (1) presence of discrimination; (2) silent witness; (3) interplay among hierarchy, recognition, and civility; (4) effectiveness of leadership and mentors; (5) support for work-life balance; and (6) perceptions of exclusion from inclusion efforts. Regardless of the factor, the underlying thread among these 6 factors was the need to belong and feel recognized and valued. Table 2 summarizes the key factors of an inclusive environment. These factors are also detailed in the following paragraphs with representative quotes.

Reported discrimination cut across all demographic characteristics and ranged from harassment and bullying to nepotism. Two commonly cited manifestations included microaggressions, defined as casual degradations of any marginalized group, and unequal performance expectations, with males and nonminority groups, often referred to as “the old white boys club,” reportedly benefiting from this inequality. Minorities and women consistently reported being held to stricter standards and needing to work harder to advance within the organization.

Many narratives were submitted by witnesses to discrimination. Witness narratives displayed the impact of discrimination on all parties involved, including fostering anxiety and hindering job performance. Most bystanders disclosed worrying about their own job security and well-being. As one stated, “Some of us whispered about how the [program leader] has done it to other people of color. ‘But do you know how powerful s/he is at [this institution].’ We learned to be silent.”

The integral role of leadership and mentorship in fostering discriminatory practices vs promoting inclusion was a common theme. A common perception was that “leadership, the faculty, all hold on to positions of power and promote and protect their own.” Many lamented the lack of mentorship, stating, “people hire and mentor those that have similar demographics.”

Respondents reported differences in treatment and their perceived value based on their status within the organization. Examples of differential treatment varied in degree, from not holding a door or being greeted to perceptions of inadequate compensation, evaluation, or promotion. Narratives characterized status in many ways, including type of position (eg, faculty vs staff, tenure vs nontenure, clinician vs clinician investigator, specialist vs generalist, postdoctoral student vs principal investigator), level of education, gender, and seniority and/or rank.

Respondents highlighted how unwritten rules often overshadowed policies put in place to support employees and faculty. Many reported inappropriate comments and/or devaluation of their work on returning from maternity leave or leave for medical reasons. Others praised the supportive actions of their supervisors as examples of an inclusive culture. As one wrote, “He went above and beyond including dedicating space in our center for lactation as well as flexibility to work from home when needed.”

There were many who reported feeling excluded specifically because they were not female or a minority and many others who assumed asking about inclusion must apply only to select groups. Comments ranged from narrow interpretations of inclusion, such as “I suppose, what you mean is in terms of the race and demographics. Then, my comments would not be what your office is seeking” to sentiments such as, “Oh to be a black female and gay! Being none of those things, not only have I been excluded, but I have been intimately involved in processes that required excluding others who are not those things.”

Respondents noted how the lack of inclusion they experienced or witnessed affected their well-being and caused stress, anxiety, and feelings of hopelessness, social isolation, and expendability. As one bystander expressed about the department’s approach to an injured colleague, “It left me worried about how I would be treated if I were disabled.” Table 3 further summarizes the effects of noninclusive workplace culture with illustrative quotes.

Narratives often described how microaggressions and favoritism eroded participants’ sense of value and thereby limited their engagement and contributions to the organization. As one narrator noted, “Needless to say, I felt exceptionally excluded and no longer want to be engaged in [this health] system. If this is how they treat new people, then it’s not worth my time investing my energy and ideas into the system. I felt very unwelcomed and it still bothers me to this day.”

A common theme that emerged was the lack of effectiveness of formal channels, such as going to human resources, leadership, supervisors, or an ombudsman, to address challenges with workplace culture and interactions. As one witness noted, “The [individual] reported the bullying to HR as well as his/her director’s immediate boss—to no avail. S/He was told s/he was being ‘too sensitive’ when his/her boss made comments about his/her weight.”

Many attempts to confront microaggressions proved futile, as one related, “If I raise a concern [about inappropriate racial/ethnic or sexist comments], they [colleagues] say something sarcastic like ‘sorry I’m not being [politically correct],’ or ‘humor is our way of coping with having to work with very sick people every day.’” Many described self-accommodating behaviors, such as, “my other friends in the office invited [the individual who makes insensitive jokes about ethnicity/race] to lunch, so I started wearing noise canceling headphones and eating alone.” In addition, others referenced resignation and/or transferring as the only solution to their experiences within work environments that lacked inclusion. As one wrote, “Using the words ‘unwelcome’ and ‘devalued’ to describe how I felt [when being pushed off a team after taking family leave] would be understatements. Fortunately, I have found employment elsewhere and will be leaving the [institution’s name] community.”

Strategies for fostering an inclusive organizational culture emerged from the narratives in the form of direct recommendations and positive examples of inclusion. We found most respondents referenced a systemic culture that influenced their group and interpersonal dynamics. Therefore, recommendations centered on system-level interventions. Table 4 details the respondent-proposed solutions described in the following sections with representative quotes.

Narratives raise the importance of examining leadership, noting, for example, that the institution “needs to educate and strengthen its chairman to change this culture.” Many proposed instituting “mandatory education…on how to accept that they have implicit biases and combat these biases,” starting with leaders and mentors within the organization. Many narratives discussed the importance of diversifying leadership as this quote illustrates: “I am confident we could be a stronger, more resilient and ultimately successful institution if there was authentic engagement of more women at the higher levels of leadership.”

Our findings highlighted the strategic need for all health care organizations to examine existing policies, such as family leave or observed religious holidays, to ensure they best meet the needs of a diverse workforce. For example, one respondent noted, “Adoptive parents are not offered any kind of paid parental leave. This makes nontraditional families like mine feel devalued and excluded.”

A common theme centered on creating a culture and a structure that supports advocacy from those who witness discrimination and/or incivility. Personal narratives routinely stressed the need for bystanders to speak up against discriminatory behaviors rather than remaining silent. For example, one participant stated, “it’s not enough to not be racist/sexist yourself—you have to stand up for other colleagues when patients or coworkers make such comments.” Narratives illustrated the success of bystanders speaking up, as in this example: “I told the patient this [patient requested a student from certain race/ethnicity to leave the room] was unacceptable and either all of us were going to take care of him/her or none of us. It was [his/her] choice. [Patient] apologized and moved on.”

Respondents emphasized the importance of interacting with different individuals within the organization to gain new perspectives and foster a sense of community by expanding collegial networks. Many respondents, irrespective of position, described the benefits of mentorship on their confidence and ability to manage stressful situations. Some suggested “having mentors and support groups where they discuss interpersonal issues and how to adapt” and proposed pairing employees with mentors who may serve as advocates, role models, and resources when concerns arise.

When comparing participants grouped by a single attribute, such as race/ethnicity, themes were strikingly consistent across all comparisons. We found minimal differences by position, except for executive leadership, who were underrepresented in our sample. Participants at the institution for less than a year related more positive instances of feeling recognized and fewer events of discrimination, as compared with those who were there longer.

Evaluating and addressing inclusion within health care environments is a new and evolving field. The diversity engagement survey, endorsed by the Association of American Medical Colleges, aims to capture the aspects of institutional culture and social dynamics that sustain an inclusive culture and support the retention of a diverse workforce.²² Our narrative analysis augments this survey’s findings²² with a deeper understanding and taxonomy of what contributes to an inclusive culture within health care organizations.

Our analysis identified 6 concrete contributors to an inclusive culture that guide tangible strategies to improve inclusivity. This taxonomy is consistent with prior investigations. There is evidence that discrimination manifesting most commonly as unequal expectations and microaggressions for both women^28-33 and minorities^34-36 in medical fields leads to social isolation, disengagement, and burnout not just for those who experience discrimination but for bystanders as well.^37,38 We see similar findings in other marginalized populations, including non–US-born international medical graduates³⁹ or health care professionals and workers with disabilities.⁴⁰ Prior studies also reveal that despite policies in place, persistent challenges in work-life balance hinder female faculty advancement and sense of value within the organization.^41-43 However, our narrative approach was able to provide a nuanced understanding of the concept of inclusion and ways that institutional policies fail to adequately address the problems they attempt to mitigate.

Inclusion is defined as the “the action or state of including or of being included within a group or structure.” There is no mention of the act of including some at the expense of excluding others. However, our analysis revealed that some perceived that inclusion efforts threaten gains for those who have traditionally thrived in the organization. This sixth factor in our proposed taxonomy for inclusion aligns with prior work. The word inclusion appeared to invoke the same perceptions of exclusion or reverse discrimination among the majority culture seen in other efforts to promote diversity.^44-46 For health care organizations and affiliated schools to move forward with inclusion efforts, we need ongoing research on how to design programs that affirm the benefits for all members while providing education to dispel the notion that such efforts are zero-sum.^47,48

We found that a lack of inclusive culture within health care organizations relates to job performance and emotional wellness. With ongoing efforts to address the wellness of our health care workforce,^49-51 it is important to understand this interplay between an inclusive environment and the wellness of employees, trainees, and students. How we treat each other and how we allow patients to treat us is as important as how we treat patients. Our wellness efforts should reflect this understanding and consider how policy adjustments and training in topics such as cultural humility and implicit biases may assist health care employees, faculty, and students with how they interact with each other^52,53 as well as address situations where patients may make discriminatory remarks.^54,55

Achieving inclusion starts with effective ways of understanding what predicts it within our health care learning and work environments. Building on surveys that assess culture within health care institutions,^22,56,57 our targeted online narrative analysis of solicited responses from employees and students provides an effective and innovative method of conducting an audit of organizational inclusion. The need to preserve anonymity and individual voice, avoid social desirability bias, and ensure impunity for members of the workforce and student body makes qualitative assessments using focus groups or semistructured interviews challenging.^58-60 Our study demonstrates that targeted online narrative analysis overcomes the challenges seen with other qualitative methods and provides an anonymous and effective method for conducting ongoing assessments of organizational inclusion. Ongoing assessments of inclusion allow for health care organizations to adapt to evolving workplace cultures.

Our study has some limitations. Our findings from a regionally limited set of hospitals and health science schools may not be nationally generalizable. However, prior work in other disciplines supports our results, suggesting that our taxonomy for an inclusive organizational culture may be widely applicable. Our email call for narratives sent out by administrators to their constituencies may be subject to selection bias. We know that executive leadership as well as individuals who self-identified as heteronormative Christian non-Hispanic white males were underrepresented in our sample. However, this narrative analysis by design aims to intentionally capture meaningful qualitative data from individuals within organizations motivated to share their stories.⁶¹

This study also has some notable strengths. While we reached thematic saturation with a substantially smaller sample size, we still analyzed all completed narratives.⁶² This qualitative assessment of more than 300 stories of inclusion within health care organizations captures experiences from health care professionals, staff, administrators, students, and trainees.

Addressing organizational culture is an emerging science in medicine, and we can learn from other disciplines about how to design system-level interventions to achieve inclusion.^63,64 Consistent with prior work in education, our findings support interventions that expand collegial networks to foster a sense of belonging and community, especially among women and minorities.⁶⁵ Our findings echo the importance of effective leaders and mentors.^66,67 In addition to diversifying leadership, existing administrators, leaders, and mentors should receive implicit bias training and inclusive leadership skills that include how to be reflective and responsive to feedback.^68-70 Leaders should foster general civility, encourage everyone to speak up against discriminatory acts, and promote policies that advocate for all members of their organization.⁶⁶ A key factor in creating and sustaining an inclusive environment is to empower bystanders and victims alike to speak up against acts of discrimination or incivility.^71,72 A system for accountability must couple such efforts with policies that support individuals subject to discrimination. Lastly, all health care professionals should possess a working understanding of how unconscious biases may influence daily interactions with colleagues and patients.⁷³

Growing evidence reveals a complex and delicate interplay among how health care professionals treat each other, the wellness and engagement of a diverse workforce, and the care provided to all patients irrespective of their cultural background or personal characteristics.^8,68 This study provides health care organizations with a novel and effective method for assessing inclusion within health care organizations along with a set of key factors to guide their efforts to operationalize inclusivity. Moreover, our findings underscore the implications of inclusion on wellness and retention of the health care workforce and student body. A focus on factors that promote retention and advancement of a diverse workforce only enhances recruitment efforts of groups underrepresented currently in our workforce.⁷⁴ How we approach both evaluating and addressing inclusion within health care learning and work environments will shape the complex dynamics between the diversity of our health care workforce, the wellness of that workforce, and the care we provide to diverse patient populations.

Accepted for Publication: May 2, 2018

Published: August 3, 2018. doi:10.1001/jamanetworkopen.2018.1003

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2018 Aysola J et al. JAMA Network Open.

Corresponding Author: Jaya Aysola, MD, DTM&H, MPH, Perelman School of Medicine, University of Pennsylvania, 423 Guardian Dr, 1229 Blockley Hall, Philadelphia, PA 19104 (jaysola@upenn.edu).

Author Contributions: Dr Aysola had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Aysola, Barg, Kearney, Higginbotham.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Aysola, Kearney, Agesa, Carmona, Higginbotham.

Critical revision of the manuscript for important intellectual content: Aysola, Barg, Bonilla-Martinez, Kearney, Higginbotham.

Statistical analysis: Aysola, Agesa, Carmona.

Administrative, technical, or material support: Bonilla-Martinez, Kearney.

Supervision: Aysola, Barg, Kearney, Higginbotham.

Conflict of Interest Disclosures: None reported.

Additional Contributions: We would like to thank Jeffrey Berns, MD, Designated Institutional Official for Graduate Medical Education at the University of Pennsylvania, PJ Brennan, MD, Chief Medical Officer for the University of Pennsylvania health system, and Larry Jamieson, MD, Dean of the University of Pennsylvania Medical School, for their support of this study. We would like to thank Corrie Fahl and Kya Hertz, from the University of Pennsylvania Office of Inclusion and Diversity, for their administrative support. None of these individuals received compensation for their contributions.