“A slap in the face”: Institutional betrayal, burnout, and career choice regret among frontline health care workers serving COVID-19 patients

Abstract

The COVID-19 pandemic has exacted a physical and mental health toll on health care and hospital workers (HHWs). To provide COVID-19 care, HHWs expected health care institutions to support equipment and resources, ensure safety for patients and providers, and advocate for employees’ needs. Failure to do these acts has been defined as institutional betrayal. Using a mixed-methods approach, this study aimed to explore the experience of institutional betrayal in HHWs serving COVID-19 patients and the associations between self-reported institutional betrayal and both burnout and career choice regret. Between July 2020 and January 2021, HHWs working in an urban U.S. health care system participated in an online survey (n = 1,189) and semistructured interview (n = 67). Among 1,075 quantitative participants, 57.8% endorsed institutional betrayal. Qualitative participants described frustration when the institution did not prioritize their safety while reporting they perceived receiving inadequate compensation from the system and felt that leadership did not sufficiently respond to their needs. Participants who endorsed prolonged breaches of trust reported more burnout and stronger intent to quit their job. Quantitatively, institutional betrayal endorsement was associated with 3-fold higher odds of burnout, aOR = 2.94, 95% CI [2.22, 3.89], and 4-fold higher odds of career choice regret, aOR = 4.31, 95% CI [3.15, 5.89], compared to no endorsement. Developing strategies to prevent, address, and repair institutional betrayal in HHWs may be critical to prevent and reduce burnout and increase motivation to work during and after public health emergencies.

Health care and hospital workers (HHWs), both clinical and nonclinical, often have implicit expectations that the health care institution that employs them will provide the necessary equipment, resources, and policies to protect HHWs from harm while they are working and advocate for their workplace and safety needs (Ahern, 2018). These expectations may be particularly activated during health care crises. During the COVID-19 pandemic, HHWs were praised as “heroes” (Cox, 2020); however, as the pandemic continued, HHWs emerged as a vulnerable group that had to overcome unprecedented challenges (Cohen et al., 2020), including increased burnout, defined as physical and emotional exhaustion due to prolonged exposure to work-related problems (Guseva Canu et al., 2021), and decreased motivation related to prolonged occupational stress. One systematic review found a 52% pooled prevalence of burnout among health care workers (Ghahramani et al., 2021), possibly leading to quiet resignation during the COVID-19 pandemic (American Hospital Association [AHA], 2021). HHWs turned to their institutions to protect and advocate for them; nonetheless, their needs were often left unmet, especially early in the pandemic (Brewer, 2021), breaching the implicit trust between HHWs and their workplace.

“Betrayal trauma theory” emerged from traumatology, a field that has expanded over time to consider the traumatic impacts associated with nonrare experiences like domestic violence, sexual abuse, and infidelity (Freyd, 1996). Betrayal occurs when an individual is being harmed by the intentional commission (e.g., enacted harm) or omission of behavior (e.g., failure to provide support) by a person whom they once trusted or depended upon (Rachman, 2010). Beyond the interpersonal level, betrayal can also take place at the system level through institutional action (i.e., enacted harm) or inaction (i.e., failure to protect or support) by a system upon which one is dependent for care or support. This has been delineated as institutional betrayal (Smith & Freyd, 2013, 2014). Betrayal trauma, whether it is interpersonal or institutional, is known to be associated with a higher prevalence of deleterious mental health outcomes, including posttraumatic stress disorder, dissociation, anxiety, and depression, than traumatic actions without co-occurring betrayal (Andresen et al., 2019; Freyd & Birrell, 2013). However, most existing studies of institutional betrayal have focused on institutions’ reactions following sexual trauma, including covering up, denying, or downplaying experiences of sexual assault (Andresen et al., 2019; Smith & Freyd, 2013). Only recently has institutional betrayal been applied to health care settings.

To date, studies of institutional betrayal in health care systems (i.e., health care institutional betrayal [HIB]) have largely focused on patients’ experiences. A study of 707 U.S. health care seekers found that two thirds of patients reported at least one act of HIB, which was related to higher levels of disengagement with future health care (Smith, 2017). Another study of Gulf War veterans found decreased engagement with Veterans Affairs (VA) or non-VA health care after experiencing HIB within VA health care (Bloeser et al., 2021). Few studies have investigated HHWs’ experiences of HIB by their employer. However, a related review by Ahern (2018) described how nurses reporting unethical behavior at work faced both retaliation and indifference from their institution, leading to traumatization. Other studies conducted during COVID-19 have focused on how health care institutions’ failure to provide safety, personnel, and resources evoked a violation of trust among HHWs, leading to moral injury and burnout (French et al., 2021; Nelson et al., 2022).

At the system level, having trust among group members is essential for achieving goals, facilitating communication, and encouraging motivation (Okello & Gilson, 2015). In contrast, a breach in the existing bond of trust can not only decrease organizational performance but also cause health problems among the betrayed individuals (Rachman, 2010). Theoretically, employer HIB among HHWs is more likely to occur in high-risk and high-stress environments (Brewer, 2021). Despite the importance of exploring the occurrence and correlates of HIB among HHWs during the COVID-19 pandemic, to date, HHWs and their workplace institutions have been relatively neglected. In particular, little research has addressed how HIB is related to burnout and work motivation among HHWs. Using a mixed-methods approach, the primary purpose of this study was to explore the extent of HIB among HHWs serving COVID-19 patients in an urban health care system in the United States. Secondarily, we examined whether higher levels of self-reported workplace HIB were associated with more self-reported burnout and career choice regret.

METHOD

Participants

Data are drawn from a larger study conducted to understand HHWs’ stress and coping during COVID-19. All HHWs working in a large health system located in the eastern United States, including direct patient care providers and support service staff, were eligible. Between July 2020 and January 2021, a total of 1,189 participants completed the survey. Using a complete case analysis, the final quantitative sample included 1,075 participants (90.4% of the full sample).

At the end of the survey, respondents were asked whether they were interested in engaging in a semistructured interview, with 280 individuals expressing interest (Figure 1). Aiming for variation in age, sex, race/ethnicity, and job categories (see Supplementary Table S1), we contacted 142 people from the pool and scheduled 71 interviews. Five individuals did not complete their interviews due to time conflicts, and we were unsuccessful in our attempts to schedule alternative interview times; thus, 66 survey participants completed semistructured interviews. At the later stage, we also used snowball sampling to recruit three nonsurvey participants whom interviewees identified as providing exceptional support, thereby helping them cope with stress. These three participants’ accounts were substantially included in the current study, as they described advocating for other HHWs, reporting problems to leadership and upper management, and not getting appropriate responses from their institution. Between August and December 2020, 69 HHWs were interviewed; however, 67 are included in this analysis because two recordings were lost due to technical difficulties.

FIGURE 1.

Qualitative participant recruitment flow chart

Procedure

We shared a weblink to an online quantitative survey, administered via Qualtrics (Provo, UT, USA), with key members of six hospitals within an urban academic health system; the link was then disseminated to HHWs via email. The survey was also accessible through a central hospital system website containing information about COVID-19 research opportunities. The survey was available in English and Spanish. A team of eight interviewers, all public health researchers with graduate-level training in qualitative research, conducted the qualitative interviews in English. Interview topics included social contexts and the stressors and support in the work and home environments during the early months of the COVID-19 pandemic. The interview question, “Are there any physical or emotional challenges in providing care for your patients these days?” elicited many responses related to workplace HIB. We continued recruiting until data saturation was achieved in each of the included job categories. The qualitative team had regular discussions to decide the extent of data saturation. All interviews were conducted remotely using Zoom (San Jose, CA, USA).

Mixed-methods approaches highlight the integration of qualitative and quantitative data (Onwuegbuzie & Teddlie, 2003), which we employed at multiple points. First, quantitative variables that assessed workplace HIB were purposefully selected for the study analysis to correspond with the qualitative descriptions. For example, burnout and career choice regret were selected as outcomes in the current study because they were frequently reported in relation to HIB. Second, we constructed a joint table relating responses from HIB items from the quantitative survey to matching quotes from the qualitative interviews (Gutterman et al., 2015; see Supplementary Table S2). Lastly, we examined whether the quantitative results supported the qualitative findings.

This study was approved by the Johns Hopkins School of Medicine Institutional Review Board. Quantitative survey respondents could opt-in to receive $10 (USD) gift cards. Qualitative participants provided oral consent and were given $50 gift cards for their time.

Measures

Sociodemographic characteristics

Demographic variables included age, sex, race/ethnicity, marital status, and educational attainment. Race/ethnicity was categorized into White, Black, Asian, Hispanic, and “other.” Marital status was categorized into married, including living with a partner, versus not married. Educational attainment was categorized as less than a 4-year degree, 4-year degree, and graduate degree. Work-related variables included job type, years of occupation, concern about being infected with COVID-19, and whether coworkers had been infected with or had a suspected case of COVID-19. Participants were categorized into five job types: clinical, administrative, management, service, and “other.” Responses to a single question about concerns related to getting COVID-19 (i.e., “How concerned are you about getting infected yourself?”) were dichotomized from a 5-point Likert-scale into “moderately–extremely,” comprising the highest three levels, and “none–slightly,” comprising the lowest two levels.

Institutional betrayal (HIB)

Five items were used to assess HIB as experienced by HHWs. Two items aimed to determine participants’ perceptions of their institution’s responsiveness (i.e., communication and safety-related resources) to the COVID-19 pandemic: “How would you rate your institution’s effort of keeping staff informed in response to the COVID-19 pandemic?” and “How would you rate your institution’s effort of providing staff with the resources (e.g., human, training, supplies) needed to safely perform their duties?” Responses were rated on a 5-point Likert scale and dichotomized into “poor,” which encompassed responses of very poor, poor, and fair, and “good,” encompassing responses of very good and good.

One item from the Moral Injury Events Scale (Nash et al., 2013) was adapted to measure feelings of betrayal by institutional leaders. Participants indicated their level of agreement with the statement, “While working in health care during the COVID-19 pandemic, I felt betrayed by institutional leaders whom I once trusted.” The final two items measured workers’ perceptions of experiencing compromised physical and psychological health due to institutional action or inaction (i.e., “While working in health care during the COVID-19 pandemic, I felt my [physical/psychological] health and safety was compromised without my consent or without care for my well-being”). The 6-point Likert scale for responses to these two questions ranged from 1 (strongly disagree) to 6 (strongly agree). Participants who reported agreeing to the statement (i.e., slightly, moderately, and strongly agree) were classified as “betrayed” or “compromised.” Responses to these five items measuring HIB showed good internal consistency across the sample, Cronbach’s α = .83.

Outcomes: Burnout and career choice regret

Burnout was assessed via one question, “Based on your definition of burnout, how would you rate your level of burnout?” (Dolan et al., 2014). Responses were rated on a 5-point Likert and dichotomized into having one or more symptoms versus none. Career choice regret was measured using one question we created, “How have your experiences during the COVID-19 pandemic impacted how you feel about your decision to work in health care?” We dichotomized responses into “regret” (undecided, regret or strongly regret decision) and “no regret” (very happy or happy with decision).

Data analysis

All qualitative interviews were audio recorded, and 67 recordings were transcribed. We drew on grounded theory approaches to qualitative analysis (Charmaz, 2014). Following an inductive coding process, six team members randomly split up transcripts and read them to gain familiarization with the data. We established 77 initial codes by combining the initial codes created during open coding, and the group met regularly to refine the codes. Similarities and differences in codes, themes, and quotes were constantly compared within and across transcripts. To facilitate linking the data analysis and writing, we wrote memos as thoughts and reflections emerged. We had regular discussions to assess the potential inclusion of emergent themes. Using the 37 final codes, six members of the research team coded the interview transcripts. For the present study, we concentrated on 10 codes directly related to HIB. Additional codes, not reported here, focused on stress and coping strategies. NVivo (QSR International, Melbourne, Australia) was used for data management and analysis.

For the quantitative data, 114 respondents (9.6%) had missingness on any variables of interest. We compared descriptive statistics between the samples of individuals who were included and excluded from the study and evaluated whether missingness was independent of the information collected. There were no significant differences between the analytical sample and participants excluded on any variables except for marital status such that included participants were more likely to be married (58.2%) than those who were excluded (46.4%), p = .017. We conducted a complete case analysis after excluding participants who had missing data on key variables of interest, as the pattern of missingness in the quantitative data indicated data were missing completely at random (Little, 1988).

For the descriptive analysis, Welch’s t tests and chi-squared tests were conducted to compare differences between participants who only completed the quantitative survey (n = 1,012) and those who completed both the survey questionnaire and the qualitative interview (n = 63). We performed 12 multivariate logistic regressions with the entire sample to examine the associations between HIB and the outcomes of interest (i.e., burnout and career regret). We ran five models (Models 1–5) to assess the associations between each of the five betrayal items and the two outcomes. We ran another model regressing the outcomes on a constructed binary variable for any endorsement of workplace HIB (i.e., answering poor or agree to any of the five questions; Model 6). Each model was adjusted for age, sex, race/ethnicity, educational attainment, concern about infection, and COVID-19 diagnosis among coworkers; these were selected based on t tests or chi-squared tests to identify characteristics associated with the explanatory or outcome variables (Raab et al., 2000). We calculated Romano–Wolf stepdown p values with 1,000 bootstrap replications to account for multiple hypothesis testing (Romano & Wolf, 2005). All analyses were conducted using Stata (Version 15), and p values less than .05 were considered statistically significant.

RESULTS

Descriptive statistics

The mean participant age was 38.8 years (SD = 11.9; Table 1). Participants were predominantly female (86.1%), White (62.4%), married (58.2%), and had attained a graduate degree or degrees (42.8%). A total of 61.5% of HHWs in the sample were working in clinical roles, and participants had been working in their occupations for a mean of 11.3 years (SD = 10.5). Approximately 73.2% of participants had concerns about acquiring COVID-19, and 71.4% had coworkers with known or suspected COVID-19.

TABLE 1.

Participant characteristics

	Total (n = 1,075)		Survey and interview (n = 63)		Survey only (n = 1,012)		p a
Variable	M	SD	M	SD	M	SD
Sociodemographic characteristics
Age (years)	38.82	11.86	36.52	10.38	38.97	11.94	.077
Years of occupation	11.25	10.46	7.49	7.75	11.49	10.57	< .001
	n	%	n	%	n	%
Female sexb	926	86.1	43	68.3	883	87.3	< .001
Race/ethnicity							.001
White	671	62.4	31	49.2	640	63.2
Black	227	21.1	10	15.9	217	21.4
Asian	75	7.0	11	17.5	64	6.3
Hispanic	64	6.0	7	11.1	57	5.6
Other	38	3.5	4	6.4	34	3.4
Marriedc	626	58.2	31	49.2	595	58.8	.134
Educational attainment							.624
< 4-year degree	250	23.3	15	23.8	235	23.2
4-year degree	365	34.0	18	28.6	347	34.3
Graduate degree	460	42.8	30	47.6	430	42.5
Job type							.355
Clinical	661	61.5	41	65.1	620	61.3
Administrative	128	11.9	5	7.9	123	12.2
Management	97	9.0	5	7.9	92	9.1
Service	81	7.5	8	12.7	73	7.2
Other	108	10.1	4	6.4	104	10.3
Moderately-extremely concerned about getting a COVID-19 infectiond	787	73.2	48	76.2	739	73.0	.582
Coworkers had known/suspected COVID-19e	767	71.4	41	65.1	726	71.7	.257
Institutional betrayal
Poorf institutional responsiveness to communication	254	23.6	18	28.6	236	23.3	.341
Poorf institutional responsiveness to resource distribution	381	35.4	27	42.9	354	35.0	.205
Feeling betrayed by institutional leaders: Feeling betrayedg	364	33.9	31	49.2	333	32.9	.008
Compromisedg physical health during the pandemic	389	36.2	30	47.6	359	35.5	.052
Compromisedg psychological health during the pandemic	409	38.1	27	42.9	382	37.8	.418
Any institutional betrayalh	621	57.8	45	65.1	580	57.3	.226

Note: One qualitative participant who had missingness for quantitative data was excluded. Three quantitative participants recruited through snowball sampling did not complete the survey.

^aWelch’s t tests and chi-square tests were conducted to assess differences between participants who completed both the survey questionnaires and the semistructured interview and those who only completed the survey questionnaires.

^bVersus male sex.

^cVersus not married.

^dVersus no concern to slightly concerned.

^eVersus no coworkers with a known or suspected infection.

^fVersus good.

^gVersus not feeling betrayed or compromised.

^hVersus none.

Approximately 23.6% of participants perceived that the institution did a poor job of communication, and 35.4% of participants had a poor perception of the institution’s resource distribution. One third of HHWs in the sample reported feeling betrayed by institutional leaders (33.9%), and 36.2% and 38.1% reported compromised physical and psychological health, respectively, due to institutional actions during the COVID-19 pandemic. Altogether, 57.8% of participants endorsed at least one type of HIB as measured by the five included items.

Compared to participants who only completed the quantitative component, those who participated in both the qualitative interview and quantitative survey questionnaires were more likely to be male (31.8% vs. 12.8%), p <. 001; identify as Asian (17.5% vs. 6.3) or Hispanic (11.1% vs. 5.6%), p = .001; report having been in the occupation for fewer years (7.5 years vs. 11.5 years), p < .001; and feel betrayed by leadership (49.2% vs. 32.9%), p = .008.

Observed themes

HHW safety was not the priority

All qualitative participants described fear, stress, and being overwhelmed at the beginning of the COVID-19 pandemic due to the ambiguity surrounding the virus and their personal safety as well as the potential risk of infecting their family members. Given the unknowns, participants acknowledged that the hospital was running less smoothly than usual. As one participant stated, “When you’re dealing with [the] unknown, there’s always going to be disorganization.”

Participants agreed that the provision of personal protective equipment (PPE) gradually improved as the pandemic continued. At the time we conducted the interviews, participants rarely reported PPE shortages as a current problem. A few clinicians who treated COVID-19 directly addressed this, with one noting that “the institution did a great job of providing us with enough PPE.” However, in the early stages of the pandemic, most participants reported that the PPE supply was short. One environmental service staff member noted, “At one time we couldn’t even get face masks. They were like gold.” Some participants brought in their own face masks, hand sanitizer, or wipes to protect themselves. HHWs who did not work directly with COVID-19 patients tried to save PPE for those who did. One clinician described their guilt about using PPE when it was in short supply, saying, “I was feeling a lot of guilt about using PPE because I am not a bedside nurse. I didn’t want to take resources from people who needed it.” Although PPE supply increased as the pandemic progressed, some participants expressed dissatisfaction with institutional preparedness and HHWs’ infection vulnerability. One physician (White, middle-aged, female) noted:

I think people are more unhappy about the fact that [the hospital] didn’t plan for this in a way I would have imagined that they did. They didn’t have any stockpile of PPE at all…that stuff is expensive, and [it] is something that an institution invests to protect the people that are part of it. That expense apparently was not warranted.

Early in the pandemic, some participants were disappointed to receive suspicious and accusatory messaging from the institution. Three participants reported receiving an email from the institution warning them not to steal any PPE to take home, one of whom said it gave them “a really bitter taste in the mouth.” Several participants perceived that insufficient PPE provision was driven by budget considerations, concluding that the institution prioritized money over safety. One occupational therapist noted, “It was a monetary game rather than caring for the workers’ lives.”

Not getting enough recognition or compensation from the system

Many clinicians stated that they liked their job. Seeing patients recover was a big motivation, and it was satisfying to successfully treat COVID-19 patients. One clinician described that “it was rewarding to see [patients] leave the hospital and go home to their families.” Participants also reported receiving tangible and emotional support from communities, such as receiving pizzas for lunch or thank-you cards; one participant described this type of support as “so heartfelt and warming.”

However, many participants, particularly those who endorsed HIB, felt they did not experience the appreciation they expected from their health care workplace. They stated that the institution “didn’t show that kind of gratitude at all.” Participants repeatedly reported that not receiving any additional monetary compensation during the pandemic was unfair. All HHWs were potentially exposed to COVID-19 infection at work, but they did not get hazard pay. Instead, the institution halted retirement plan contributions and retained the 2% annual wage increase to compensate for the institution’s lost revenue from canceling elective surgeries during the pandemic. One participant who was infected with COVID-19 due to possible work exposure explained, “[There’s] no way for me to avoid [COVID-19] for 8 hours [a day of work]. I’m here, potentially getting exposed to COVID, and we didn’t get anything.” Similarly, another support staff described that not receiving hazard pay was insulting. One environmental service staff member (Black, middle-aged, male) reported:

I was told by a coworker that [the hospital’s] not giving us hazard pay because they didn’t cut our [work] hours, which to me is a slap in the face because why would you decrease our hours if you need us the most now with the pandemic? That just seems so disrespectful.

One lead nurse noted, “Trying to convince somebody that they’re a valued member is probably the hardest part right now, and it’s unfortunately deeply rooted in compensation.”

Furthermore, some participants reported that verbal recognition from leadership and upper management, such as words of support and encouragement, was lacking. One participant described, “Not that we needed the money to carry things, sometimes just a ‘thank you’ would have been enough, or just to come and say, ‘Job well done and thank you for doing what you do.’” Other participants reported that getting minimal verbal recognition was even more discouraging because of the lack of material recognition. One nurse explained, “I just thought that [it] was not honoring nurses when they have all these signs up saying, ‘Thank you. You’re a health care hero.’ I feel like it’s a slap in the face.” Similarly, one physician described it as “lip service to what we contributed to the pandemic.”

Division with the system and leadership

A strong relationship with institutional leadership was described as essential to feeling valued or supported. Several participants reported having supportive supervisors and/or managers who motivated them to continue to work. Participants described characteristics of supportive leadership, including ensuring that HHWs were protected, advocating for HHWs’ needs, providing tangible support, and joining in HHWs’ struggles. One environmental service staff member stated how supportive her manager was, noting:

My boss got down and dirty with the rest of us. Whatever we would need, he would make it work. I think he provided a lot of support to get us through [the pandemic]. Just being there and rolling up a sleeve and working as hard as we did, that meant a lot.

Nonetheless, several participants who provided direct care to COVID-19 patients expressed disappointment when they realized their leaders were not working in the same types of environments. Participants expected support from their leadership because caring for patients with a novel disease in a biocontainment unit was physically and emotionally taxing. This expectation turned into frustration when HHWs found out that some people in leadership positions were less engaged in clinical work. One nurse (White, male, younger age) stated:

Leadership was not necessarily present when we were in [biocontainment unit] mode…because they’re trying to avoid COVID themselves. They’re working from home more [laughs] instead of going in person, meeting with us, talking with us, and helping with us.

Similarly, a few physical therapists reported that physicians avoided coming to biocontainment units, which made them realize there was a lack of solidarity. Observing some physicians staying out of the unit, one respiratory therapist described, “You didn’t feel like [physicians] were supporting. There was a growing divide between the different disciplines.”

HIB also surfaced when participants felt neglected in the decision-making process or were ignored when they voiced their concerns. For example, the institution banned visitors at the beginning of the pandemic. When the institution started to allow one visitor per patient in the summer of 2020, some participants perceived that their own safety could be compromised. However, participants reported that when they expressed their concerns to management, the concerns were left unaddressed; one participant stated simply that upper management “didn’t listen.” HHWs reported being frustrated when the institution was not responsive to the medical staff while being reactive to patients’ needs. One lead nurse reached out to upper management to explain the nurses’ safety concerns and stated that “they just listened to me and said, ‘Yes, we’re working on that and that is going to come to play…[but] it never came to fruition.” One physician who actively mentioned his frustration with the system also described, “We were told to just keep quiet and just work.”

Intertwined problems resulting in burnout and leaving one’s job

Multiple participants stated that their coworkers left their jobs during the pandemic due to insufficient protection, recognition, and support. Reasons for leaving included increased workload, insufficient compensation, burnout, mental distress, and/or childcare issues raised after school closures. Coworkers’ quitting was particularly salient in accounts of nurses and physician assistants (PAs) who provided direct care to COVID-19 patients. Some nurses reported their coworkers quit their job to become traveling nurses for better pay. The vacancies were filled by deploying nurses from other departments or, ironically, hiring traveling nurses. Having new members in the unit increased training responsibilities for the remaining nurses. Moreover, a hiring freeze enacted after the pandemic, as well as a hospital policy requiring a 10day quarantine period for staff exposed to COVID-19, aggravated understaffing. Trapped in a vicious cycle, staff shortages increased the workload among the remaining staff, causing burnout. One PA working in the emergency department stated, “There is no way I can build myself up to six patients. That’s supposed to be a full assignment. If I have less than 10, I’m having a really good day.” One physician fellow described possibilities of exploitation and reported that fellows were “asked to do a lot of extra hours [of work]” in an understaffed environment. This participant felt that the institution “filled gaps [of taking care of more patients per physicians] using the fellows.” Several participants stated that the increased workload not only elevated their stress but also may have compromised patient safety. One PA (White, female, younger age) explained, “I don’t know how safe it is for the patients. We’re supposed to have nine providers and there’s only five, so you better fly. Sometimes, it makes me think about the safety of my patients.”

Participants who noted being mistreated during the pandemic reported decreased motivation to work, eventually only completing the minimum requirements. One physician described, “I’ve realized that the institution doesn’t have my best interests at heart. I would be the person who would not be an enthusiastic employee.” Anotherclinician (White, female, younger age) stated:

I don’t work past 5:00 [p.m.]. That was one change I made [after the pandemic]. Normally, I’d work till 6:00 [or] 7:00 sometimes to make sure that the work was done and that I advocated the best that I could for my patients. Now, I’m not going to do that. If you’re going to cut my hours and benefits, I am absolutely not going to kill myself to make you look good.

Associations between institutional betrayal and both burnout and career choice regret

Quantitatively, 47.2% of participants reported having one or more symptoms of burnout, and 36.7% reported that experiences during the COVID-19 pandemic made them regret their decision to work in health care (Table 2). Participants who perceived that the institution did poorly in communicating with regard to COVID-19 and distributing resources effectively had 3 times higher odds of burnout, communication: adjusted odds ratio (aOR) = 3.00, 95% confidence interval (CI) [2.16, 4.18], resources: aOR = 2.71, 95% CI [2.03, 3.61], compared to those with positive perceptions. Feeling betrayed by institutional leaders was associated with 4-fold increased odds of burnout, aOR = 3.62, 95% CI [2.70, 4.85]. Compromised physical health due to institutional actions increased the odds of burnout by 2 times, aOR = 2.46, 95% CI [1.85, 3.27]; the odds of burnout were increased 3 times for compromised psychological health, aOR = 3.20, 95% CI [2.41, 4.25]. Reporting at least one type of HIB was associated with 3-fold higher odds of burnout compared to not reporting any HIB, aOR = 2.94, 95% CI [2.22. 3.89].

TABLE 2.

Multivariable logistic regressions investigating associations between institutional betrayal and burnout and career choice regret

Outcomes	Burnout: ≥ 1 symptom vs. no symptoms (n = 507, 47.2%)		Career choice regret: Regret vs. no regret (n = 394, 36.7%)
	aOR	95% CIa	aOR	95% CIa
Model 1: Institutional responsiveness to communication
Good	Ref.		Ref.
Poor	3.00	[2.16, 4.18]**	5.23	[3.75, 7.29]**
Model 2: Institutional responsiveness to resource distribution
Good	Ref.		Ref.
Poor	2.71	[2.03, 3.61]**	4.44	[3.32, 5.95]**
Model 3: Feeling betrayed by institutional leaders
Not feeling betrayed	Ref.		Ref.
Feeling betrayed	3.62	[2.70, 4.85]**	3.69	[2.75, 4.94]**
Model 4: Physical health compromised during the pandemic
Not compromised	Ref.		Ref.
Compromised	2.46	[1.85, 3.27]**	2.79	[2.09, 3.73]**
Model 5: Psychological health compromised during the pandemic
Not compromised	Ref.		Ref.
Compromised	3.20	[2.41, 4.25]**	3.24	[2.42, 4.33]**
Model 6: Dichotomized institutional betrayal
None	Ref.		Ref.
Any institutional betrayal	2.94	[2.22, 3.89]**	4.31	[3.15, 5.89]**

Note: n = 1,075. Significance, denoted using asterisks, reflects Romano-Wolf stepdown adjusted p values. Each finding stems from a separate multivariable logistic regression model.

^aAdjusted for age, sex, race/ethnicity, marital status, educational attainment, job type, years of occupation, concern about infection, and COVID-19 diagnosis among coworkers.

^*adjusted p < .05.

^**adjusted p < .01.

^***adjusted p < .001.

Likewise, perceived poor institutional responsiveness to communication, aOR = 5.23, 95% CI [3.75, 7.29], and resource distribution, aOR = 4.44, 95% CI [3.32, 5.95], increased the odds of career choice regret by 5 and 4 times, respectively, compared to perceived good responsiveness. Feeling betrayed by institutional leaders was associated with 4-fold increased odds of career choice regret compared to not feeling betrayed, aOR = 3.69, 95% CI [2.75, 4.94]. Compromised physical and psychological health due to the institution’s actions or inaction increased the odds of job regret by 3 times, compared to no reported compromised health, physical: aOR = 2.79, 95% CI [2.09, 3.73], psychological: aOR = 3.24, 95% CI [2.42, 4.33]. Reporting any HIB was associated with 4-fold greater odds of career choice regret compared to no HIB, aOR = 4.31, 95% CI [3.15, 5.89].

DISCUSSION

Building and maintaining individuals’ trust in the institutions where they work is crucial in creating a positive work environment. However, this trust can be easily breached in times of crisis. Trust breaches and their effects at the organizational level are currently understudied in health care (Mayo et al., 2021) and are rarely considered among HHWs during system-wide health care crises. Instead, studies of HIB often focus on trust violations between health care institutions and the patients they serve (Gigler et al., 2022). It is noteworthy that our findings suggest that early in the COVID-19 pandemic, more than half of a quantitative sample of HHWs indicated their trust was breached in some way by their own workplace (i.e., health care institution). These breaches were strongly associated with increased burnout and career choice regret.

Only a handful of previous studies have investigated employee perceptions of HIB during the pandemic. One qualitative study conducted among 16 medical staff members in the United Kingdom indicated participants felt abandoned by their leadership or organization, which resulted in burnout (French et al., 2021). Our findings are consistent with this small study, and by expanding and including HHWs with various job responsibilities, we confirm that HIB can be prevalent across work roles. Further, many prior studies have focused on institutional betrayal in relation to moral injury (French et al., 2021; Nelson et al., 2022), a concept that originated within a military context. By definition, perpetration-based moral injury arises when one’s actions (e.g., killing others) violate their inherent morality (Litz et al., 2009). However, similar to HIB, betrayal-based moral injury also acknowledges that being betrayed by others, often by an authority figure, can injure one’s morality (Shay, 2014). In betrayal-based moral injury, violations that result in a loss of integrity or the disruption of the moral compass play a critical role. Furthermore, there is no stipulation that the perpetrator of betrayal is related to or acting on behalf of the institution. In contrast, for institutional betrayal to occur, the perpetration has to occur on behalf of a trusted institution (Smith & Freyd, 2014). Institutional betrayal can occur even when the victim remains blind to the betrayal and/or the action by the institution is one of omission (e.g., failure to keep employees safe) rather than commission (e.g., deliberately putting employees in harm’s way). HIB also does not require nor suggest that the recipient act in ways that are inconsistent with their own morals or values; instead, the person is not cared for in the expected way based on their trust of the institution. Given that both these experiences can generate similar emotions (e.g., betrayal, disappointment, disgust), further study is warranted to explore and differentiate between these concepts as they occur among HHWs during public health emergencies.

Previous studies have shown that inadequate PPE provision during the COVID-19 pandemic evoked HIB among health care workers, possibly leading to adverse mental health outcomes (Gold, 2020; Klest et al., 2020). Expanding upon these concerns and using a mixed-methods approach, our study suggests that PPE shortages and wage cuts were intertwined with and interpreted as a disrespectful “monetary game.” Participants also highlighted the navigation of understaffed work environments as another critical factor that compromised the safety of both HHWs and patients. HHWs’ perception of poor working conditions is nothing new, and it existed prior to the pandemic. Working in a hospital with a healthier work environment is known to be associated with lower odds of burnout and job dissatisfaction as well as better quality-of-care outcomes (Aiken et al., 2011). These results suggest that the problem related to poor work environments may have been amplified during the COVID-19 crisis.

Our study participants appreciated that the local community sought to support “health care heroes.” Some researchers have cautioned against using the language of heroism and evoking military metaphors because these metaphors prevent HHWs from addressing and prioritizing their own needs or concerns related to safety. Cox (2020) suggested that HHWs expect reciprocal social obligation in return for taking risks in fulfilling their duty. Our participants, though, didn’t object to the “hero” label itself nor to the appreciation they got from the broader community—rather, many contrasted it with what they saw as a lack of appreciation from their own institution.

Our findings indicate that endorsing any type of HIB was associated with 3- and 4-fold increased odds of burnout and career choice regret, respectively. These results are noteworthy given the documented links between higher levels of HHW burnout and reduced patient care and safety. Prior research has shown that HHW burnout is associated with increased medical errors and hospital-acquired infections among patients, reduced time spent between providers and patients, and staffing shortages (Cimiotti et al., 2012). Similarly, health care workers’ job dissatisfaction is known to be related to low levels of patient satisfaction (McHugh et al., 2011). Recent studies have underscored the high prevalence of burnout among health care workers and the loss of workforce during the COVID-19 pandemic (AHA, 2021; Ghahramani et al., 2021). Highlighting the associations between HIB and both burnout and career choice regret, we postulate future health care interventions that seek to build and maintain trust between the workplace/institution and its staff members would ultimately improve patient safety by preventing and repairing institutional betrayal and reducing HHW burnout and career choice regret. This is consistent with work by Spence Laschinger et al. (2012), who found that a workplace intervention among Canadian nurses significantly improved structural empowerment, workplace civility, and trust in management. Our study highlights how management behavior can mitigate or exacerbate HIB during workplace crises.

Several study strengths should be noted. The study provides empirical evidence of how HIB is related both to burnout and to career choice regret. The study sample consisted of a wide range of HHWs who interacted with COVID-19 patients by providing care and support. We included relatively less-researched HHW populations, such as service and administrative staff, who have essential roles in serving COVID-19 patients. We also used a mixed-methods design, which allowed us to comprehensively understand HIB.

Despite these strengths, some limitations should be considered. First, the participants are not representative of all HHWs in either the health system we studied or in the United States. Future research needs to further explore HIB among HHWs in other health care settings. Moreover, the sample was limited to HHWs who opted into the online survey, and from this select sample, only those who wanted to participate in the interviews were included in the qualitative study. As presented in Table 1, the 63 qualitative participants were more likely to report feeling betrayed by institutional leaders than the larger sample. It is possible that HHWs with negative experiences were more likely to opt in to participate in the qualitative portion of the study. Further, this study was funded by the parent educational institution of the health system. We tried our best to assure participants of anonymity and confidentiality, but they may have been reluctant to report negative perceptions of their health system employer, resulting in an underestimation of HIB. Second, HIB emerged as a salient theme at the stage of qualitative data analysis; thus, due to the study sequencing, we were not able to utilize a comprehensive measure of HIB (e.g., the Health Care Institutional Betrayal Questionnaire; Smith, 2017). However, the five items we presented in this study do represent key aspects of the HIBQ specifically and institutional betrayal more broadly (e.g., an institution not taking proactive steps to prevent negative experiences, an institution not creating a safe environment, employees not feeling like the institution is protecting staff members’ health and well-being). Although our study did not quantitatively assess all aspects covered in the HIBQ, many accounts in the qualitative portion of the study were clearly related to these additional components of HIB (e.g., covering up negative experiences, treating the experience as no big deal, making it difficult to report concerns, responding inadequately when concerns are reported). Moreover, due to the need to keep the survey length reasonable for respondents who were under stress during the pandemic, some constructs were measured using single items. Future research using more robust measures is warranted. Finally, the qualitative findings are descriptive, and our quantitative data are cross-sectional; thus, causal associations cannot be inferred. However, the use of a mixed-methods approach enabled us to triangulate quantitative and qualitative findings, strengthening the validity of the findings.

In this mixed-methods study, we found that HIB was common and strongly associated with increased burnout and career choice regret among HHWs serving COVID-19 patients. Because breaches of trust within health care institutions are more likely to occur during crises, having a disaster plan, systematically building and maintaining trust between the institution and its staff members, and addressing and repairing ruptures of this trust are essential. The findings suggest that prioritizing institutional staff members’ safety while providing regular, heartfelt emotional and tangible acknowledgment of their work; transparent communication; and supportive leadership are imperative to minimizing HIB. These institutional acts, specifically directed at mitigating HIB, may also serve to facilitate HHWs’ ability to overcome challenges during public health emergencies and provide high-quality patient care. Future health care institutional policy and research should develop and deploy strategies to reduce institutional betrayal among HHWs, as preventing HIB may reduce workforce attrition and increase the quality of service delivery.