Patient perspectives of the climate of diversity, equity, and inclusion in the emergency department.

Objective: Assessing the diversity, equity, and inclusion (DEI) climate of emergency departments (EDs) can inform organizational change to provide equitable, inclusive, and high-quality care to their diverse patient populations. The purpose of this project was to investigate patient perspectives on the climate of DEI in an urban ED.
Methods: This was a cross-sectional survey study conducted in a large-volume, urban ED in Detroit, MI, from November 2018 to January 2019. The survey was developed by an experienced ED DEI committee via an iterative process and broad consensus.
Results: During their care in the ED, 849 patients completed an anonymous survey about their perspectives and experiences of DEI in that ED. Overall, the responses were favorable as most respondents reported that the ED staff treated patients from all races equally (75.8%) and made patients feel accepted (86%). However, some respondents felt that the ED staff's treatment of populations with greater complexity, such as patients who are mentally ill (16.8%) or lower income (14.3%), needs the most improvement. Conclusions: This DEI climate assessment survey of ED patients' perspectives revealed important insights that could guide strategic initiatives to advance DEI in the ED. This assessment may serve as a model for continuous evaluation of DEI over time and in multiple healthcare settings to help guide organizational change efforts.

INTRODUCTION

Background

Emergency departments (EDs) often function as “safety net” healthcare settings for diverse patients, providing critical access to healthcare regardless of patients’ insurance coverage or ability to pay. ED clinicians, similar to clinicians in other specialties and the general population, have implicit biases, such as racial biases, that can influence clinical decision making and patients’ perception of care quality , , , and impede communication and treatment adherence for diverse patients. Compared with other specialties, ED clinicians in particular are expected to set aside their biases to care for all patients, yet they tend to face a greater cognitive load and environmental stress, which are factors that can influence their interpersonal interactions and clinical treatment as well as the overarching climate of care. , , In the ED, cognitive stress may come from competing mental tasks, such as juggling multiple patients, being interrupted while writing orders, or interacting with patients of a minority race; environmental factors, such as overcrowding and insufficient staffing; and clinicians’ personal needs such as fatigue, stress, or hunger. Racial minority patients and those with other structural vulnerabilities are also at increased risk of receiving fragmented and delayed care and often have less satisfaction with and greater mistrust of their clinicians, which can result in greater morbidity and mortality. For populations with persistent and concentrated health inequities such as Black, Latino/a/x, indigenous, and low‐income patients, culturally sensitive, unbiased, and structurally competent care is essential for facilitating affirming patient–clinician cross‐cultural relationships and addressing structural and interpersonal causes of health inequities.

Importance

The unique clinical environment of the ED also presents distinct challenges that may adversely affect equitable and inclusive patient care compared with other care environments. Clinicians face significant environmental stress, physiologic strain, and cognitive loading, which can inadvertently influence patient–clinician interactions, quality of treatment, clinician wellness, and the overarching healthcare climate. These stressors can cause ED clinicians to rely on automatic cognitive processes, such as categorization, stereotyping, and implicit biases, which can contribute to health inequities and disrupt the patient–clinician relationship. Therefore, understanding patient perspectives of the diversity, equity, and inclusion (DEI) climate in the ED can inform organizational change strategies to promote DEI in patient care, which is also related to patient safety and healthcare quality. , ,

The Bottom Line

This survey of >800 emergency department (ED) patients (73% non‐Hispanic Black and 4% Hispanic) found that ED staff treated patients from all races equally and made patients feel accepted. However, some respondents felt that the ED staff's treatment of mentally ill or lower income patients need the most improvement. In addition, 16.8% of respondents reported that they had witnessed discrimination or harassment of ED staff by another patient.

Goals of this investigation

Despite leading medical organizations’ more recent focus on DEI, such as the American College of Emergency Physicians, the Society of Academic Emergency Medicine, and the American Medical Association, research on the highest yield strategies to improve the culture of DEI in healthcare has remained sparse, , particularly in the ED. Recognizing this gap, which may be attributed to a historical lack of funding or focus, this study aimed to investigate patient perspectives of the DEI climate in a large‐volume, urban ED. This provides the foundation for future repeat measurements to iteratively develop, modify, and assess the organizational DEI climate in this ED over time.

METHODS

Study design and setting

This cross‐sectional survey was administered in the ED of a tertiary care teaching hospital in Detroit, MI, between November 2018 and January 2019. This ED had 100,000 patient visits in 2018, representing >60,000 unique patients. We surveyed a convenience sample of the ED patient population. This hospital's institutional review board approved this study.

Survey development

The survey was developed by members of the Department of Emergency Medicine DEI committee of this hospital, composed of approximately 40 interdisciplinary ED staff members, including physicians, nurses, social workers, and so on, and based on existing DEI literature, committee members’ expert experiences, and several members’ survey development experiences. The survey was discussed iteratively until consensus was obtained. The larger DEI committee reviewed the survey, and all revisions were incorporated. The assessment was found to be at a Flesh–Kincaid grade level of 5.3 to ensure readability. It was translated into Spanish by 2 authors of this work, 1 a native Spanish speaker and member of the DEI committee, and then certified to be accurate by LanguageLine Solutions. It consisted of 41 questions largely organized into matrices and divided into the following 4 sections: care for specific patient populations, patient's care experiences and values, ED compared to other hospital system clinics and departments, and demographics (Appendix S1). We intentionally combined positive and negative valence questions and 5‐point Likert‐type scale responses to minimize acquiesce bias. The survey collected demographics, including race, age, gender, sexuality, religion, nationality, education, usual place of care, and the number of ED visits in the last year. The survey did not collect unique patient identifiers, allowing patient anonymity. If a patient wished to advance to the next page without answering a question, they were prompted to but not required to respond to these questions.

Participant recruitment

Trained research assistants (RAs) collected surveys between 6 and 12 am in 2‐hour to 4‐hour shifts in the non‐intensive care unit (ICU) areas of the ED. RAs screened patients using the electronic health record (EHR) according to the inclusion criteria before approaching them to further screen for exclusion criteria.

Inclusion criteria included the following: (1) respondent aged 18 years or older (respondent may have been the patient or an adult accompanying the patient if the patient was a minor or had significant limitations) and (2) patient had an emergency severity index (ESI) of 2, 3, 4, or 5 to avoid disrupting care for patients who were overtly critically ill. RAs excluded patients if patients (1) were physically or emotionally distressed, (2) had an acute psychiatric issue or were intoxicated, (3) were in police custody, (4) were unable to take the survey in English or Spanish, (5) were unavailable after 3 recruitment attempts, or (6) had previously taken the survey per their own disclosure. If a patient met all these criteria, RAs would introduce the study verbally. They would also emphasize that the patient's participation and responses were anonymous and would not influence their care in the ED. RAs used telephone translator services to screen and recruit Spanish‐speaking patients. If the patient expressed interest in participating, the RA would give them an electronic tablet on which they could read the informed consent document and proceed with the survey. If patients declined participation, RAs noted the reason why. To reduce interference with patient care, the patients with the ED longest times were prioritized, and RAs approached patients after their initial evaluation by their ED clinician, as indicated on the EHR. The ED clinicians were not involved in the administration of the survey and directed any questions back to the RA present in the ED at that time.

Data collection

After informed consent was obtained, the survey was self‐administered on an electronic tablet device in English or Spanish via Qualtrics. If an English‐speaking participant requested assistance with the survey because of vision or literacy limitations, the RA read the survey verbatim and inputted the patient's responses as stated without further guidance or interpretation to minimize interviewer effects. This assistance was not offered to Spanish‐speaking participants because of low RA Spanish fluency, the logistical difficulty and possible errors of using the translator phones to read the survey, and the possible interference of patient care if the clinical team required a translator phone. An adult companion could take the survey in lieu of a minor patient or an adult patient who requested assistance. The survey took approximately 10 minutes to complete, and RAs left the patient room while it was completed to provide privacy unless the respondent requested assistance. If the clinical team needed to speak or provide treatment to the patient, RAs were able to pause the survey and return the tablet to the respondent afterward. RAs were also trained to document their observations of patient and clinician interactions in the ED in their shift field notes, regardless of whether the patient observed was a survey respondent.

Analysis

Qualitative data were collected in the form of RA observations but were not coded or analyzed because of the lack of standardization in qualitative data collection methodology. We analyzed the survey data using SAS version 9.4 (SAS Institute). We conducted descriptive statistics for all variables. We report continuous data as means with standard deviations (SDs) and categorical data as counts and percentages. To create consistency in the analysis and data presentation, responses with a negative valence were flipped to positive, and participant responses were reverse categorized accordingly. For example, responses to the question “This ER's employees judge obese patients” were flipped to correspond to “This ER's employees DO NOT judge obese patients.”

Measurements

Item responses were measured using a 5‐point Likert‐type scale to minimize acquiescence bias. We condensed agreement responses into 3 categories to characterize responses as favorable or not in analysis: strongly/somewhat agree, strongly/somewhat disagree, and neither agree nor disagree/no opinion/no experience based on the primary valence (agree, disagree, or neutral) of the response given. We categorized frequency responses into never/rarely, some/most/all of the time, and no experience/prefer not to answer. Demographics were self‐reported. We did not impute any demographic variables.

RESULTS

Of the 1691 patients screened, 849 respondents were sampled (Table 1). The response rates for participant‐reported demographic variables of interest ranged from 60.9% to 86%, whereas the response rates for respondent perspective variables ranged from 90% to 95.5%. Of the surveys, 0.82% (n = 7) were completed in Spanish.

TABLE 1

Participant recruitment flow

Patients screened for participation	Patients excluded from participation	Reason for exclusion from participation
1691	8	Did not meet all inclusion criteria (respondent of adult age and non‐acute ESI)
1683	20	Does not read and understand English or Spanish
1663	123	In physical, mental, or emotional distress
1540	5	Psychiatric or alcohol intoxication issue
1535	248	Unavailable after 3 attempts
1287	48	Had already taken the survey
1239	38	Other reason for exclusion
1201	352	Declined participation
Final sample: 849

Abbreviation: ESI, emergency severity index.

Characteristics of study patients

Survey respondents’ demographics are displayed in Table 2. Response rates ranged from 46.2% to 87.2% for these questions. Most study participants were Black (72.6%), which is comparable with the overall ED population (75%). The majority of participants had lived in the United States for most of their lives (96.4%) and took the survey in English (99.2%). Survey participants reported that they had sought care in any ED on average 3.4 times in the past 12 months (SD = 3.44, range = 1–25) and at this ED 2.3 times during the same time (SD 1.9). This was the first ED visit at that particular hospital in 12 months for 47.1% (n = 400) of survey respondents.

TABLE 2

Characteristics of survey respondents

Characteristic	Sample size proportion, mean (SD)	Sample size, n	ED patient proportion, % a
Age, years	38.4 (15.6)	517	43.6 (20.6)
ESI level	2.7 (0.5)	849	2.6 (0.9)

Abbreviations: ED, emergency department; ESI, emergency severity index.

ED patient population data are from 2017 internal reports.

Main results

Respondent perspectives are presented in Table 3 and Figure 1. Response rates ranged from 86.6% to 95.5% to these questions. There was a large range (8.1%–49.4%) of neutral responses to questions regarding patient perspectives. Responses were classified as “favorable” if participants responded with agreement to statements regarding a positive DEI climate in the ED or with disagreement to statements regarding a negative DEI climate. Correspondingly, responses were categorized “unfavorable” when participants disagreed and agreed with statements on a positive and negative DEI climate, respectively. The statements where most participants responded favorably were for ED staff's equal treatment of patients of all races (75.8% agree, n = 811) and for patients feeling accepted in the ED (86% agree, n = 790). Questions that were responded to least favorably, where the least participants responded favorably or the most participants responded unfavorably, included the statement querying whether ED staff are sensitive to the needs of transgender patients (32.1% agree, n = 808). Participants disagreed most with statements about ED staff treating patients equitably regardless of income or educational status (14.3% disagree, n = 807) and ED staff judging those with mental health needs (16.8% disagree, n = 809).

TABLE 3

Participant perception of the ED staff background, training, and treatment of diverse populations

	Participant response frequency, n (%)
	Strongly/somewhat disagree	Neither/no opinion	Strongly/somewhat agree	Participant response number, n
Q20: Please share how much you agree or disagree with the following statements about the employees in this ER. This ER's employees…
Treat patients of all races equally	62 (7.7)	134 (16.5)	615 (75.8)	811
Are sensitive to the needs of immigrant patients	65 (8.1)	397 (49.4)	342 (42.5)	804
Respect all religions and religious preferences	39 (4.8)	226 (27.9)	546 (67.3)	811
Try hard to accommodate a patient's request for providers of a preferred gender	61 (7.5)	294 (36.3)	454 (56.1)	809
Provide the same quality of care to patients who do not speak English as English‐speaking patients	46 (5.7)	270 (33.5)	490 (60.8)	806
Treat disabled patients well	46 (5.7)	200 (24.8)	560 (69.5)	806
DO NOT treat patients with mental health problems poorly a	136 (16.8)	322 (40)	351 (43.4)	809
DO NOT treat homeless patients poorly a	106 (13.1)	340 (42.1)	361 (44.7)	807
Treat patients with lower income and education the same as patients with higher income and education	115 (14.3)	242 (30)	450 (55.8)	807
DO NOT judge obese patients a	90 (11.2)	341 (42.5)	372 (46.3)	803
DO NOT judge people with reproductive or sexual health problems a	112 (13.9)	357 (44.3)	337 (41.8)	806
Are sensitive to the needs of transgender patients	55 (6.8)	494 (61.1)	259 (32.1)	808
Q21: Please share how much you agree or disagree with the following statements based on the interactions you have had or seen between employees and patients in this ER.

Abbreviations: DEI, diversity, equity, and inclusion; ED, emergency department; ER, emergency room.

Question and response valency were flipped.

FIGURE 1

Patient level of agreement with perspective statements. DEI, diversity, equity, and inclusion

Notable to the environment of care in the ED, 16.8% of respondents reported that they had witnessed discrimination or harassment of ED staff by another patient, although fewer participants reported experiencing discrimination or harassment themselves by other patients (7.2%) or by ED staff (9.7%). Patients also infrequently reported observing discrimination by ED staff against other patients (11.4%) or other ED staff (8.6%).

In comparing the DEI climate of the ED to other clinical contexts, most respondents who had experience in other clinics and departments in the same hospital felt that the ED had similar or greater diversity (56.4%, n = 417), equity (65.2%, n = 482), and inclusion (58.3%, n = 429) comparatively. Neutral opinions to this question, because of patients’ lack of experience with other departments within the hospital or preferences not to answer, ranged from 29.8% to 38.0%.

LIMITATIONS

Limitations of this study include social desirability and sponsor bias, as patients were waiting for healthcare while sharing their perspectives on the clinician providing that care; thus, it is possible that there is underreporting of negative experiences. This bias may have contributed to our lower response rate to individual prompts and a higher percentage of “neutral” responses. We attempted to mitigate these biases by making surveys anonymous, emphasizing that RAs were not a part of the clinical team, seeking out participants after their care had been initiated, and stating that participants’ care would not be affected by survey responses. Conversely, taking the survey during patient care provided the advantage of real‐time evaluation of the ED DEI climate, minimizing recall bias on relevant survey items. The surveys were available in only English and Spanish, and Spanish‐speaking patients were not offered reading assistance; thus, perspectives from speakers of other languages and associated immigrant populations could be underrepresented, and their unique insights were not collected. However, only a low number of Spanish‐speaking people completed the survey, so our data are likely not applicable to those who do not speak English. Our survey was not formally pilot tested or validated before administration. Without cognitive interviewing of potential respondents, it is possible that the questions were not interpreted in the manner that we intended. , However, RAs did provide robust field notes, and the team will use field notes for continuous improvement and results from this inaugural survey to conduct future validity and reliability analyses. The survey was developed by a committee of volunteers who were all ED clinicians. Because this study was performed in a single ED, the generalizability of the findings to other clinical environments, such as non‐urban or non‐ED environments, is likely low. However, results are meant to illuminate local climate, so populations of interest will need to be adapted to in other settings.

DISCUSSION

Our survey study of ED DEI climate showed that overall patient perspectives from a small subset of the ED population at a single urban ED regarding DEI were generally favorable while still illuminating opportunities for improvement. Most participants felt that ED staff treat patients across various races, religions, ability, education levels, and income statuses equally or well. A high proportion of participants also reported that ED staff made patients feel accepted, cared for the community served, and valued patients’ differences. However, the data also showed that there is still room for improvement regarding DEI climate, more so in some areas than others. Information on the relative areas of DEI success can help inform the creation of adaptable strategies and processes to address areas that need the most DEI improvement. Notably, the response rates for questions on demographics compared with those for questions assessing participants’ perspectives were different. Perspective questions tended to have higher response rates and a lower variance in response rates compared with questions on participant demographics. This could be attributed to the fact that the demographic section of the survey was the last section, and participants may have felt survey fatigue. In addition, the divergence in response rates could be attributable to participants' perception of the importance of providing their demographic data to our analyses compared with their perspective data. Finally, participants could fear loss of anonymity by answering demographic questions despite our RAs emphasizing that surveys were anonymous.

Patients identified as requiring improved care

Patients perceived by our respondents to be immigrants or transgender or with sexual health, mental health, and housing problems were noted to have the lowest scores for equitable and inclusive care. These groups identified by our patients, including immigrants, , sexual minorities, , and those experiencing mental health, sexual health, or homelessness issues, are documented to be subject to implicit biases from healthcare clinicians. Moreover, these populations may face greater systemic disparities that may precipitate more contact with, and bias within, the ED. This is particularly notable as clinicians are most susceptible to acting on implicit biases, such as the biases they may have toward these communities, when their work involves multitasking, limited information, high‐pressure circumstances, negative emotional states, or care delivery outside of an established patient–clinician relationship—characteristics typical of ED care. , , Consequently, it is important to implement mechanisms to reduce bias and promote DEI in a climate that inherently exacerbated clinician biases against populations that may rely on the ED more heavily.

Improving clinician wellness for patient wellness

Policies and practices that address operational issues may not only improve quality of care directly but also may improve clinician wellness and the DEI climate because cognitive stressors posed by ED phenomenon such as overcrowding that can then lead to clinician reliance on implicit biases in their behavior and decision making and perpetuate biases that inhibit high‐quality, equitable care. , ED clinicians likely have greater vulnerability to empathy fatigue and rely on biases when interacting with these groups because of the environmental and psychological stressors typically characteristic of ED work. , , , , Potential system‐level solutions to this challenge include increasing ED social work services, optimizing chronic disease management integration within the ED, creating processes and forums for patients and clinicians to anonymously express and collectively address concerns, and evaluating health system policies from a DEI lens regularly. , Thus, teams could consider comeasuring DEI climate and clinician wellness as outcomes of quality improvement initiatives.

Another component of poor clinician wellness that was identified in our survey was discrimination. The most witnessed form of discrimination reported in the survey was patient discrimination against ED staff. Survey respondents likely witnessed this given curtained rooms, hallway care, and open workspaces for clinicians. This observation may also reflect nuances in the expression of discrimination in that rude or unkind patient behavior may be seen as discriminatory or clinician behavior may be subtle. Given the aforementioned ED stressors, ensuring that staff wellness is prioritized is critical. Self‐care among physicians is associated with better patient counseling and screening. Clinician wellness increases their ability to handle the physical and emotional tolls associated with treating high needs ED populations. Promoting clinical well‐being is a complex task that may require major changes in healthcare delivery. Organizational leadership should assess physician well‐being and identify areas where staff need support and may start by facilitating access to and providing time to use tools such as positive psychology exercises, mindfulness, narrative medicine, work‐hour limitations, and resources for maintaining a work–life balance to reduce the burn out and stressors that they may experience in the ED, particularly if they are being discriminated against. , It is vital to develop and enforce policies that protect ED staff and limit abusive patient behavior while protecting patients’ rights. Prior literature even suggests using physician well‐being as an indicator of the health of healthcare organizations. Although reports of ED staff perpetuating discrimination were comparatively lower than for patients, organizations must continue efforts to extinguish any forms of discrimination within the ED.

Optimizing cultural sensitivity

Many participants indicated that ED staff should learn about patients’ cultures. Improving the cross‐cultural understanding and communication of clinicians is associated with better patient outcomes and may reduce inequities in outcomes of and access to care. Increased cultural humility training can also reduce the probability of minority employees experiencing microaggressions and discrimination at work, which can improve job satisfaction and organizational performance, giving the healthcare organization a competitive market advantage. Moreover, the relationship between cultural humility and diversity has been well established: a more diverse staff allows for greater opportunities for positive cross‐cultural interactions, which can increase cultural awareness and knowledge, which help to mitigate biases. This is particularly relevant to emergency practices where clinician demographics are unlikely to mirror those of the community, particularly because emergency physicians’ racial and ethnic diversity lags behind other large specialties, such as family medicine, with 18% underrepresented minorities (URM), internal medicine (15% URM), and obstetrics and gynecology (22% URM). , ,

This survey demonstrated that patients tended to perceive the ED DEI climate positively. More important, the results revealed areas of focus for improving the DEI aspects of patient care, patient–staff interactions, and ED staff support. Through organizational change, particularly with regard to leadership, management, accountability, policies, and partnerships with other healthcare organizations within the ecosystem, hospitals can facilitate equitable care for diverse patient populations. Despite the broad implementation of DEI initiatives within healthcare organizations, the publication of outcomes related to these programs is limited. Thus, data from surveys such as this can add to the sparse literature and inform departmental‐strategic DEI priorities and ED staff understanding of how collective behaviors contribute to the overarching DEI climate in the ED. Moreover, organizational leaders can use these data to explore, create, evaluate, and revise policies for improving the detection and management of healthcare inequities and reduce clinician, operational, and cultural factors that impact the delivery of inclusive care. We encourage other organizations to implement similar assessments in their EDs using our survey as a model to iteratively evaluate DEI climate and changes.

AUTHOR CONTRIBUTIONS

Martina T. Caldwell, Nikhil Goyal, Harold Gomez Acevedo, Jason Folt, Namita Jayaprakash, and Michelle Slezak conceived the study and designed the survey with oversight from the Henry Ford Health Department of Emergency Medicine Diversity, Equity, and Inclusion Committee. Kavya Davuluri and Harold Gomez Acevedo oversaw the translation of the survey into Spanish. Martina T. Caldwell and Kavya Davuluri supervised research assistant onboarding, data collection, and data management, including quality control. Martina T. Caldwell and Kavya Davuluri guided data analysis, which was completed by a statistician. Kavya Davuluri drafted the manuscript and tables. Nikhil Goyal developed the figure and contributed substantially to the manuscript's emendations. All authors participated in manuscript revisions. All authors take responsibility for the accuracy of the manuscript. Martina T. Caldwell takes responsibility for the work as a whole.

CONFLICTS OF INTEREST

The authors declare no conflict of interest.

Supporting information

Click here for additional data file.