key: cord-0762269-djbkcw6u
authors: Vranas, Kelly C.; Golden, Sara E.; Nugent, Shannon; Valley, Thomas S.; Schutz, Amanda; Duggal, Abhijit; Seitz, Kevin P.; Chang, Steven Y.; Slatore, Christopher G.; Sullivan, Donald R.; Hough, Catherine L.; Mathews, Kusum S.
title: The Influence of the COVID-19 Pandemic on Intensivists’ Well-Being: A Qualitative Study
date: 2022-05-11
journal: Chest
DOI: 10.1016/j.chest.2022.05.003
sha: b46ff58c19db6cc7462e9d8051ae5d0851573319
doc_id: 762269
cord_uid: djbkcw6u

Background The COVID-19 pandemic has strained healthcare systems and resulted in widespread critical care staffing shortages, negatively impacting quality of care delivered. Research Question How have hospitals’ emergency responses to the pandemic influenced the well-being of frontline intensivists, and are there any potential strategies to improve their well-being and help preserve the critical care workforce? Study Design and Methods We performed semi-structured interviews of intensivists at clusters of tertiary and community hospitals located in six regions across the U.S. between August-November 2020, using the “Four S framework” of acute surge planning (i.e., Space, Staff, Stuff, and System) to organize the interview guide. We then employed inductive thematic analysis to identify themes describing the influence of hospitals’ emergency responses on intensivists’ well-being. Results 33 intensivists from 7 tertiary and 6 community hospitals participated. Intensivists reported experiencing substantial moral distress, particularly due to restricted visitor policies and their perceived negative impacts on patients, families, and staff. Intensivists also frequently reported burnout symptoms due to their experiences with patient death, exhaustion over the pandemic’s duration, and perceived lack of support from colleagues and hospitals. We identified several potentially modifiable factors perceived to improve morale, including the proactive provision of mental health resources, establishment of formal back-up schedules for physicians, and clear actions demonstrating that clinicians are valued by their institutions. Interpretation Restrictive visitation policies contributed to moral distress as reported by intensivists, highlighting the need to reconsider the risks and benefits of these policies. We also identified several interventions as perceived by intensivists that may help mitigate moral distress and improve burnout as part of efforts to preserve the critical care workforce.

Intensivists experience higher rates of burnout compared to other specialties. 1-3 The Coronavirus disease 2019 pandemic has strained existing critical care resources and increased rates of depression, anxiety, post-traumatic stress disorder (PTSD), and burnout syndrome among intensivists. [2] [3] [4] [5] [6] [7] [8] [9] Burnout syndrome has numerous negative consequences for clinicians. Personally, it is associated with increased risks of depression, PTSD, alcohol abuse, and suicidal ideation among physicians. 1, 10 Professionally, it is associated with poor work performance, decreased quality of care, and increased job turnover. 1

Not surprisingly, the pandemic has exacerbated critical care staffing shortages, further compromising the quality of care delivered. 11 It is important to better understand factors influencing clinicians' well-being during the pandemic as part of efforts to lessen the psychological burden they experience and preserve this crucial workforce. However, the impact of hospitals' emergency preparedness and responses on clinicians' well-being has not been well-described. Therefore, we performed semi-structured interviews of intensivists in multiple settings across the United States (U.S.) to a) qualitatively evaluate the influence of hospitals' COVID-19 responses on intensivists' well-being, and b) gain insight into potential strategies to help preserve the critical care workforce.

As part of a larger qualitative study evaluating the influence of the COVID-19 pandemic on ICU organization and care processes, 12 we also sought to better understand the J o u r n a l P r e -p r o o f impact of the pandemic and hospitals' responses to it on clinician well-being to inform efforts to preserve the critical care workforce. As described previously, 12 we purposively selected participants from clusters of tertiary medical centers and community hospitals in six states across the U.S. that experienced surges early in the pandemic based on publicly-available county-level case rates. 13, 14 We included hospitals in Arizona, California, Louisiana, Michigan, New York, and Washington State. Using key informant and snowball sampling methods, 15 we recruited three practicing intensivists (including one medical ICU director) per hospital via email between August 6-November 4, 2020.

We sought to recruit 36 participants across 12 hospitals based on prior studies showing that 20-40 interviews are needed to reach saturation across multi-site qualitative studies. 16 All participants consented and were provided renumeration. We report detailed methods using the Consolidated Criteria for Reporting Qualitative Research guideline 17 (e- Table 1 ).

We used similar interview guides for frontline intensivists and ICU directors, with the latter containing additional structured questions about ICU organization and staffing models (e-Appendix) 12, 18 We used the "Four S" theoretical framework of emergency preparedness (i.e., Space, Staff, Stuff, and System) to organize the semi-structured interview guides and provide context for intensivists' perceptions of their hospitals' responses to the pandemic. 19 The multidisciplinary research team, including intensivists (KCV, KSM, TV, AD, KS, SC, CLH), health services researchers (SN), and sociologists with expertise in qualitative methods (AS, SEG), iteratively revised the interview guide J o u r n a l P r e -p r o o f during six pilot interviews, which were used to generate a preliminary codebook but otherwise not included in analyses. AS conducted one-on-one interviews virtually over a secure web-based platform. Interviews were digitally recorded, transcribed, deidentified, and verified for accuracy.

The study team used deductive analysis methods 20 to independently code data from the six pilot interviews and create a preliminary codebook of themes. 19, 20 Next, KCV and SEG jointly coded the first four study transcripts, utilizing the Framework Method for inductive thematic analysis. 20, 21 We subsequently independently coded the remaining transcripts and created framework matrices to aid interpretation. We met frequently to review data and perform comparative analyses of each set of codes to note similarities and differences, discuss codes that were unclear or coded differently, add new codes or collapse them as appropriate, resolve differences, and achieve consensus. [22] [23] [24] Throughout this process, the research team performed analytic triangulation and rereviewed transcripts to ensure analyses remained well-grounded in data. We used ATLAS.ti 8 (Berlin, Germany) to organize data, code transcripts, and track decisions related to the codebook and analyses. The study was approved by the VA/OHSU Joint IRB.

We recruited 36 intensivists, with 33 (92%) participating (including 20 from seven tertiary hospitals, and 13 from six community hospitals). Interviews ranged between 45-90 minutes. All participants completed fellowship in critical care medicine, and 12 (36%) were female (Table 1) . We reached thematic saturation 25 after reviewing 26 transcripts. Subsequent transcripts were reviewed, but no additional novel themes were identified.

We identified four major themes focused on clinician well-being that were consistent across both tertiary care and community hospital settings and both ICU directors and frontline intensivists. These four themes, each with several subthemes noted in Table 2 , included: 1) contributors to moral distress; 2) contributors to burnout symptoms; 3) longterm impacts of the pandemic on the critical care workforce; and 4) targeted interventions to address clinician well-being and morale as perceived by intensivists.

Importantly, participants were not explicitly asked about moral distress or burnout symptoms in the semi-structured interview guide (e-Appendix); instead, these topics were spontaneously and frequently discussed by all participants in their own words and emerged as prominent themes early on in the coding process. In order to ensure consistency and reliability across coders in subsequent data analyses, we defined moral distress as the inability to act according to one's core values and perceived obligations due to internal or external constraints. 26 We defined burnout syndrome as characterized by three symptoms: (1) emotional exhaustion related to devoting excessive time and effort to tasks not perceived to be beneficial; (2) reduced sense of personal accomplishment, including feelings of helplessness; and (3) depersonalization, including J o u r n a l P r e -p r o o f indifferent or negative attitudes toward work, colleagues or patients. 1,25 A conceptual model of our findings is shown in Figure 1 , with representative quotes in Table 3 .

All participants worked at hospitals that restricted visitors during the pandemic.

Participants described negative consequences of these policies on three stakeholder groups: patients, families, and staff. First, participants felt that restricting visitors potentially worsened patient outcomes such as delirium (quotations (Q) 1-5). Several intensivists also noted that families' absences worsened patients' morale. For example, one highlighted the "devastating" impact of restricting visitors on patients "who are teetering on the edge of intubation…who need all the encouragement they can get...We try, but we're not family. We don't have the same power over our patients that their loved ones have"(Q5). Intensivists also felt that restricting visitors prolonged patients' suffering because families were unable to see the "medical reality" of their loved ones (Q6). For this reason, several participants reported being more assertive in their recommendations to withhold life support among patients at the end of life (EOL) (Q7).

Second, participants described harm that these policies inflicted on families by disrupting communication and trust with care teams (Q8-9). One explained that restricting visitors changed the dynamic of EOL discussions, making it difficult for families to " [understand] what their loved one is enduring [and that] we are all working J o u r n a l P r e -p r o o f really hard to keep your loved one alive…There would be one person who just doesn't seem to trust you because they can't see it with their own eyes"(Q8).

Third, many intensivists were distressed by their own loss of humanity attributed to restricted visitor policies. For example, one stated that such policies "Giving [families] information over the phone that your loved one is dying…should be a personal thing.

[Restricting visitors] has taken out some of that humanity"(Q10). Several participants felt that restricting visitors was "barbaric" and the most challenging part of the pandemic.

One explained "[Restricting visitors] was…[one] of the worst things that I have ever seen…Looking back, that was the real nightmare…Watching that many people die in the ICU alone"(Q11). Given their concerns about patients dying alone, many ICU staff chose to be present with dying patients. A few participants deliberately violated policy to allow families at bedside, further contributing to clinicians' moral distress. One participant explained "I find that very frequently, [physicians and nurses] break the rules…It's kind of the final kick in the crotch for your failure. You couldn't save the person and now they are going to die by themselves. I find a lot of us will just let the family go…in the backdoor"(Q12).

Many intensivists described tension between their fear of becoming infected and their desire to help patients. For example, one participant explained "When this started, my own fear was 200%...I asked people for pep talks…like, inspire me that I can do this because I'm scared…tell me this is what I was meant to do"(Q13). Such fear was J o u r n a l P r e -p r o o f intensified when their own colleagues fell ill. One participant stated "Just the psychological stress [of] knowing that you are very vulnerable; seeing our colleagues who work in the same field getting sick and maybe pass away…[has been] very emotionally stressful"(Q14). Due to concerns about infecting their loved ones, many chose to live separately from their families, further exacerbating their stress.

Intensivists experienced moral distress related to the potential or actual need to allocate scarce resources during surge events. Many hospitals developed strategies to assist clinicians with resource allocation (Q15). However, several intensivists reported having to implicitly triage which patients received limited resources based on perceptions of who would benefit the most (Q16). One explained "We had concerns about whether allocation was being done fairly…We were bumping up against the principle of distributive justice"(Q17).

Finally, several intensivists experienced moral distress due to the use of experimental therapies outside of clinical trials for patients with COVID-19. For example, one participant described the conflict that existed between physicians' desire to "do something" and the lack of evidence-based treatments early in the pandemic (Q18).

Another physician described how they "went along" with prescribing hydroxychloroquine early on, but when data showed the drug was either "ineffective or maybe associated J o u r n a l P r e -p r o o f with harm…now you're thinking, 'great. How many people did I harm as a result of using this medication which I never thought there was good evidence for?'"(Q19).

Many participants explained that the sheer number of dying patients caused substantial emotional exhaustion and distinguished the pandemic from other experiences caring for critically ill patients (Q20-23). For example, one physician explained how intensivists are used to dealing with death among their patients "in a way that doesn't affect you in the long run…[But] during COVID, it was that times 100…You…do the best job you can and a third still die"(Q20). The duration of the pandemic also impacted physicians' level of exhaustion and morale (Q24-27). One intensivist explained "My concern about staff well-being is that this…'fight or flight' response can only last for so long, but the chronic stress of having to do this over and over again without a solution at the end of the tunnel…that [adds] to the strain"(Q24).

Intensivists reported a reduced sense of personal accomplishment, often manifested as helplessness that they could not prevent patients with COVID-19 from dying (Q28-29).

One stated "It was very traumatic…in terms of mental health, burnout, trauma, and at times depression, because you knew you couldn't do anything about people who were dying"(Q28).

Many intensivists expressed negative attitudes towards hospital staff who were perceived as unwilling to see patients with . Participants also noted that hesitancy among consultants to see patients led to inappropriate treatment delays. One described an instance of ordering an ultrasound for an ICU patient, noting that "None of the radiology techs or radiologists…who are all wearing 'Healthcare Hero' t-shirts…[would] come into the ICU to do that procedure because they don't want to be exposed to COVID"(Q31).

Concerns about the long-term impact of the pandemic on ICU staff were common.

Several intensivists lost trust in and felt undervalued by their institutions due to their perception that hospital administrators did not prioritize clinicians' safety when it came to availability and use of personal protective equipment (PPE). For example, one participant stated "I noticed the hospital claimed that we never ran out…of PPE. At the same time, it was not available plenty of times"(Q34). Another explained that, at a time that their hospital had an "ample supply of N95s," intensivists were told to use surgical masks in the rooms of patients with COVID-19-recommendations which they considered "preposterous" and chose to ignore (Q35). Several hospitals ended up cutting benefits or salaries of physicians, which further compromised their morale (Q36).

Additionally, intensivists expressed concerns about the long-term impact of the pandemic on their mental health and the critical care workforce writ large (Q37-40). One explained: "We were all messes…Things that normally would not make you cry, would make me cry. One of my partners…realized recently how she was laughing at things again-and that it had been a really long time since she laughed that easily"(Q37).

Many also witnessed signs of PTSD and anxiety in their colleagues with a noticeable increase in those seeking counseling and pharmacologic treatment (Q38).

Interventions to Improve Intensivist Well-Being and Morale Participants described several factors at the individual, departmental, hospital, and community levels that improved morale and well-being (Figure 2) . At the individual level, many participants found it helpful if mental health providers proactively reached out to frontline clinicians rather than relying on them to seek help. Additionally, several noted that before the pandemic, physicians would commonly work even when sick themselves. In the setting of COVID-19, however, many departments created formal backup systems for physicians to call out sick, which represented a welcomed change (Q41). One participant noted that establishing formal backup systems for physicians helped create a culture of "psychological safety [in which] it's okay to say I cannot work"(Q42). Others found it helpful to reduce the number of consecutive days worked.

At the hospital level, several participants noted that the use of electronic devices and web-based platforms were helpful to improve communication with families in the setting of restrictive visitor policies. Additionally, intensivists appreciated having protocols in J o u r n a l P r e -p r o o f place that explicitly excluded the treatment team from making decisions about allocation of scarce resources (Q43). Many participants also explained how helpful it was for hospitals to provide basic resources including free parking and food. In some cases, hospitals provided childcare or alternative housing to enable providers to isolate from their families. Many hospitals repurposed waiting rooms within hospitals to create "serenity spaces" in which staff could spend time away from the bedside (Q44-46).

Physicians rarely received hazard pay, but when they did, they felt it boosted morale.

Finally, some participants noted how much they appreciated expressions of gratitude from their communities. One explained "We had letters and cards…Restaurants would feed us! It was unbelievable support from the community…One group even bought us warm, fuzzy socks, so…we were kind of hugged by our feet"(Q47). Another noted how rewarding it was when family members expressed gratitude, explaining "It's nice when the weight of what you're doing is really appreciated"(Q48).

This qualitative study provides an in-depth understanding of the influence of hospitals' emergency responses to COVID-19 on the well-being of intensivists working at geographically diverse tertiary and community U.S. hospitals during the pandemic.

Intensivists reported experiencing moral distress, particularly due to restricted visitor policies and their perceived negative impacts on patients, families, and ICU staff.

Intensivists also frequently reported burnout symptoms due to their experiences with patient death, exhaustion over the pandemic's duration, and perceived lack of unified J o u r n a l P r e -p r o o f support from colleagues and hospitals. Participants identified several potentially modifiable factors at the levels of the individual clinician, department, and hospital that were perceived to improve morale, build trust, and reduce burnout. These include the proactive provision of mental health resources for frontline staff, creation of formal backup schedules for physicians, and clear actions demonstrating that clinicians' well-being is valued by their institutions.

Even under normal circumstances, ICU clinicians are a limited resource. 1,27,28 Our findings suggest that the COVID-19 pandemic may accelerate shortages of the critical care physician workforce by exacerbating moral distress, a key factor associated with burnout syndrome. 2 Among healthcare workers, moral distress and burnout syndrome are often related to the organizational environment; therefore, addressing the interaction between the healthcare worker and their work environment is fundamental in efforts to mitigate moral distress. 1,29 Our use of the "Four S" framework to organize the interview guide provided context for intensivists' perceptions of their hospitals' responses to the pandemic and on clinician well-being. Moreover, our findings were consistent across both tertiary and community hospital settings, and reinforce the need for a coordinated, national response to combat burnout and improve clinician well-being. 30 Participants identified restricted visitor policies as the major contributor to moral distress during the pandemic because of threats they posed to patient-and family-centered care. The absence of family at bedside disrupted the care team's communication with family members and further complicated complex decision-making processes, J o u r n a l P r e -p r o o f particularly among patients at the EOL. 31 Although restricting visitors during the pandemic was common 32,33 and has substantial face validity to minimize virus transmission and conserve PPE, our findings question whether the benefits of restricting visitors outweigh the risks. More research to better quantify the risks and benefits of restrictive visitor policies is urgently needed to inform ongoing and future pandemic responses.

While many factors contributing to burnout symptoms were beyond the control of clinicians and hospitals, participants identified several factors that were potentially modifiable. For example, ensuring adequate access to PPE may help reduce clinicians' fear of becoming infected while simultaneously building trust in their institutions.

Additionally, some participants noted that proactive involvement of mental health professionals with frontline staff was helpful in efforts to mitigate psychological burden and potentially reduce burnout symptoms. Our findings build upon recent literature by focusing on moral distress experienced by ICU physicians and outlining potential strategies to proactively address clinicians' mental health during and after the pandemic. 3,6,34,35 These include actively screening staff to evaluate the influence of the pandemic on their mental health, as well as proactively educating clinicians of the possibility of moral distress and burnout. 36 Such actions are particularly important in light of evidence that clinicians are often reticent to speak about their mental health due to stigma. 35

Finally, our study adds to the literature by providing an in-depth understanding of how strategies implemented at the departmental-and hospital-levels may reduce distress and improve morale among intensivists. For example, physicians miss work for illness much less often than other hospital staff. 37,38 Furthermore, intensivists' often work up to 14 consecutive days; 39 such schedules are associated with burnout. 40,41 A recent study found that limiting the number of consecutive days worked by intensivists was associated with shorter ICU length of stay, without differences in ICU readmissions or hospital mortality. 42 Taken together, these findings suggest that both intensivists and patients may benefit from implementation of staffing models at the departmental-level that limit the number of consecutive days worked and/or provide more formalized systems for backup coverage.

At the hospital-level, intensivists reported the importance of ensuring that basic needs of frontline staff (e.g., food, childcare) are met during surge events. Such actions by a healthcare system represent tangible demonstrations of support for healthcare workers, which reinforce institutional compassion and help clinicians overcome distress and fear to continue to provide care under difficult circumstances. 35, 43, 44 Strengths of our study include its rich perspectives from intensivists at tertiary and community hospitals across the U.S. Additionally, we identified novel contributors to moral distress within the context of COVID-19. Our study also has limitations. First, the experiences and strategies described in this study do not reflect the multidisciplinary perspectives of all ICU team members. Second, women comprised approximately one-J o u r n a l P r e -p r o o f third of study participants, although this proportion generally reflects the representation of women in the critical care workforce nationwide. 45 Third, we did not collect data on age, race, or ethnicity of participants in order to protect confidentiality of participants.

Fourth, given the qualitative methodology, we were unable to capture granular details about hospitals' visitor policies over time. Fifth, perceptions of intensivists may have evolved over time, and given time constraints, transcripts were not returned to participants for their review. Finally, we did not quantitatively measure moral distress or burnout.

This study provides an in-depth understanding of specific factors contributing to moral distress and burnout symptoms as perceived by intensivists working in U.S. hospitals during the COVID-19 pandemic. We found that restrictive visitor policies contributed to moral distress among intensivists, highlighting the need to more carefully consider the risks and benefits of visitor policies prior to their widespread implementation.

Additionally, several potentially modifiable factors were identified to improve morale and mitigate burnout symptoms, including the proactive provision of mental health resources for frontline staff, creation of formal back-up schedules for intensivists, and clear actions demonstrating that clinician well-being is valued. Implementation of these interventions may help preserve the critical care workforce during current and future public health crises.

J o u r n a l P r e -p r o o f Acknowledgments Authors would like to thank Matt Howard for his help with recruitment of study participants, as well as all of the physicians who participated in the study. Results: In this qualitative study of 33 intensivists at tertiary and community hospitals across the U.S., physicians reported experiencing substantial moral distress, particularly due to restricted visitor policies and their perceived negative impacts on patients, families, and staff. Intensivists also frequently reported burnout symptoms due to their experiences with patient death, exhaustion over the pandemic's duration, and perceived lack of support from colleagues and hospitals. We identified several potentially modifiable factors perceived to improve morale, including the proactive provision of mental health resources, establishment of formal back-up schedules for physicians, and clear actions demonstrating that clinicians are valued by their institutions.

Interpretation: Restrictive visitation policies contributed to moral distress as reported by intensivists, highlighting the need to reconsider the risks and benefits of these policies. We also identified several interventions as perceived by intensivists that may help mitigate moral distress and improve symptoms of burnout as part of efforts to 23. J o u r n a l P r e -p r o o f Restrictive Visitor Policies: Impact on Patients "Part of the recovery is recovering your mental state, especially when you are sick but not being able to have your family around or having been there for limited hours also affects that."

Restrictive Visitor Policies: Impact on Patients "When you have elderly individuals at high-risk for delirium, taking out contact and communication with people that are normally a part of their life is very bad, and only exacerbates that problem."

Restrictive Visitor Policies: Impact on Patients "Initially when COVID hit…the hospital instituted a no visitor policy. It was all well-intentioned... we didn't want family members exposed, we didn't want nurses exposed to potentially infected families. I think something that I haven't seen discussed is the law of unintended consequences. You don't have family in the room, so redirection is difficult. You are keeping the nurse out of the room, so redirection is more difficult… These patients definitely were difficult to sedate and control on the ventilator which led to more delirium."

Restrictive Visitor Policies: Impact on Patients "Not having someone to advocate for [patients] may have affected their care because a lot of times sick patients can't ask for stuff and do stuff and it's their family members that kind of advocate for them. So, I do think to a certain extent it may have affected the quality of care that they got."

Restrictive Visitor Policies: Impact on Patients "The morale that having family around gives to a patient, them not having that has been devastating. Especially with these patients who are kind of teetering on the edge of intubation on max high flow, who need all the encouragement they can get, and there's no one there to give it except for the nurses and ourselves, and we try, but we're not family. We don't have the same power over our patients that their loved ones have."

Restrictive Visitor Policies: Impact on Patients "We really struggled with what was a very unique challenge to try to engage family members in surrogate decision-making since so many of these patients were dying and you really needed to make some decisions about what the goals of care were. It was a tremendous impediment to not have that family member able to visit with the family, to actually see the medical reality of their loved one. That was a tremendous impediment to the dying process."

Restrictive Visitor Policies: Impact on Patients "I'm a little bit more adamant and a little bit more assertive in my withholding of critical care, of life support, when there are situations when it is not going to help in the long run, not going to make them survive. And I know that if they are on that, they are going to die on a ventilator in a room by themselves. I will frequently tell patient's family that, "I will not offer it, it's not going to help you."" Quotation 8 Community Hospital 305

Restrictive Visitor Policies: Impact on Families "[Restricted visitor policies] really changed the dynamic of end of life discussions, understanding what their loved one is sort of enduring, as well as we are all working really, really hard to try and keep your loved one alive and trying to do better and I would say that the J o u r n a l P r e -p r o o f vast, vast majority of people trusted us and were thankful for what we did, and felt like we were doing a good thing in the world. But very, very rarely there would be the one person who just doesn't seem to trust you because they can't see it with their own eyes."

Restrictive Visitor Policies: Impact on Families "One unintended consequence [of restricted visitor policies] is that [families] did not fully grasp the gravity of the situation the patients were in. To me that's the one big downside, is that you can't really call them and have a discussion that involves informed consent if you can't inform them. Part of being informed is seeing how bad things are, and if they can't see it, then how can they really be informed?"

Quotation 10 Tertiary Hospital 409

Restrictive Visitor Policies: Impact on Clinicians "[Restricted visitor policies] made it incredibly difficult to interact with patients' families, especially for patients who have COVID, and giving them that information over the phone that your loved one is dying, I think should be a personal thing and it has taken out some of that humanity."

Restrictive Visitor Policies: Impact on Clinicians "Everyone in the ICU has the same disease so on a given day, there may be half the patients in your ICU [who] just die. They die alone because their families can't come visit them because they have COVID. Watching that many people die alone is probably something that is going to give a lot of us critical care providers… a lot of us PTSD at the end of this. That was probably some of the worst things that I have ever seen... Looking back, that was the real nightmare that occurred, and it occurred at every hospital in the country. Every hospital just had to do this. But watching that many people die in the ICU, aloneand they had us with them, but that's not what they would want. That's not the ideal way that people want to die. That, understandably, affected mental health of everyone."

Restrictive Visitor Policies: Impact on Clinicians "I find that very, very, very frequently myself and nurses, and other physicians, break the rules. Because when you talk about things that are the biggest strain on physicians, that's the biggest strain, is watching that. Because it's kind of the final kick in the crotch for your failure. You couldn't save the person now they are going to die by themselves. I find a lot of us will just let the family go in the room, go in the back door." Quotation 13 Tertiary Hospital 513 Use of Experimental Therapies "We felt like it was important to go along and give hydroxychloroquine for the sake of avoiding a lot of variation in care between providers... And then of course, all the data comes out showing that they're either ineffective or maybe associated with harm, and now you're sitting there thinking well, 'great. How many people did I harm as a result of using these medications which I never thought there was any good evidence for?'"

Emotional Exhaustion "There is, on a good day in the intensive care unit, some of the things that staff witness with the human condition and with patients dying in the ICU. On a good day it's not something you can see over and over again and be completely comfortable with it and adapt to it and process it in a way that it doesn't affect you in the long run, emotionally, mentally. And that's on a good day. During COVID, it was just that times 100. I mean, now you have a patient population in the ICU where a third of them are going to die, that you are going to do the best job you can and a third will still die because there isn't a cure for this disease." Quotation 21 Community Hospital 623

Emotional Exhaustion "It's certainly the most amount of death that any of us have seen compacted into this short of a time period."

Emotional Exhaustion "Just the sheer volume of patients and the hours they had to work, and everybody was physically very exhausted by the end of it, so I think that's one thing that happened. The other thing that happened is just the psychologic stress, you know... And yeah, just seeing the sheer amount of death in generally, in terms of the patients just not doing well. In our experience you did all you could, but a lot of patients just didn't do well." Quotation 23 Community Hospital 407

Emotional Exhaustion "After our initial surge wound down, when we saw [another wave] coming back…then you saw morale really dip because you think you are through it and then you see it coming again and people don't want to do it again after the first run through." Quotation 24 Tertiary Hospital 325

"My concern about staff well-being is that this sort of fight or flight response that people had initially can only last for so long but the chronic stress of having to do this over and over and over again without a solution at the end of the tunnel right now…that [adds] to the strain."

Quotation 25 Tertiary Hospital 102

Emotional Exhaustion "As the pandemic has progressed, we have this double wave that has now become a very prolonged course of being engaged, you know, and this has caused people to be very tired.

We are talking about six, seven months of just being on the alert and, you know, just that level of physical and emotional exhaustion that I could tell among our staff." Quotation 26 Community Hospital Emotional Exhaustion "Our nursing staff have reached burnout several times. I think over the last six months, I think it was especially emotionally difficult, psychologically difficult to have our peak get so much J o u r n a l P r e -p r o o f 114 better and then have what was worse than the first go around. And so, I think that was mentally and emotionally difficult for the staff…physician burnout is something we are struggling with as well."

Emotional Exhaustion "Fatigue is a big deal and also the morale and the emotional state of the staff has been pretty negatively impacted in our ICU… Some of these patients are on this high flow oxygen support for weeks and weeks, and our nurses and staff in general get to know these patients during that time and they still even weeks later deteriorate and pass away. And I think that's the biggest sort of impact on the staff is sort of the emotional state and being able to continue on because it's that and then being overworked, working more often than you would normally do is kind of a bad mix for mental health."

Reduced Personal Accomplishment "I know a couple of colleagues that burned out also to the point of I took a week off and just went off to a quiet lake and needed time to recover. It was traumatic. It was very, very traumatic and the colleagues that I have spoken to who are willing to talk about it seem to have suffered that trauma also. In terms of mental health, burnout, trauma, at times depression, because you knew you couldn't do anything about the people who were dying." Quotation 29 Community Hospital 222

Reduced Personal Accomplishment "I cried with some of my colleagues just about some of the losses, feeling helpless in certain situations. It still feels very raw."

Quotation 30 Community Hospital 623

Depersonalization "Consultants have actually been the most challenging part of it… Delaying procedures that should have been done same day, waiting for a COVID test to come back. Like a GI bleeder coming into the ICU for hemorrhagic shock and the endoscopy gets postponed waiting for a COVID test, because we don't have, or we didn't at the time, have any in-house testing, so it took 48, 72 hours to get a test back. We're like, a GI bleeder with varices, you know, why? That's not okay… It's hard to advocate for your patient in that setting."

Depersonalization "Let me give you an example of some tension here. I am on service in the COVID ICU and I need an ultrasound for one of my patients done. And none of the radiology techs nor radiologists -who were all wearing "Healthcare Hero" t-shirts that were providednone of them will come to the ICU to do that procedure because they don't want to be exposed to COVID. That kind of tension exists in the hospital." Quotation 32 Tertiary Hospital 624

Depersonalization "There were consultants who refused to see patients… patients were being ruled out, they wouldn't see until they were ruled out. The consultants I'm talking about are like surgeons, or ENTs, ear, nose, and throat doctors. I just was really angry, honestly, about all that."

Quotation 33 Tertiary Hospital 517

Depersonalization "Getting diagnostic tests was very challenging. It was just a real challenge actually, to provide what I think we consider standard of care for many of our patients. So, that was quite frustrating, and I think from a personal perspective, we often felt like we're in the hospital here for 12-plus hours a day in the patient's room trying to take care of them and here are our colleagues who are at home, scared to come in to sort of help us."

Quotation 34 Community Hospital Compromised trust in institution "I noticed the hospital has claimed that we never ran out or been low PPE. At the same time, it was not available at plenty of times." Compromised trust in institution "Even during the phase of the pandemic where we were trying to reuse our N95 masks, there was still an ample supply of N95 masks… I remember being in the ICU and becoming aware that the infection control standards had been lessened. We all thought it was preposterous. We were all like, 'whoa, I'm not going to go in that patient's room with a surgical mask. I'm going to wear an N95 mask no matter what.' So, we, at the unit level, chose to ignore them." Quotation 36 Tertiary Hospital 513

Feeling undervalued "Benefits started being cut. Salaries were cut... There were a number of layoffs…There are ways to break morale and that's unfortunately one of them."

Quotation 37 Community Hospital 331

Concern for development of mental health disorders "We were all messes, in retrospect, from March through at least June… things that normally would not make you cry, would make me cry. One of my partners actually was saying to me how she realized recently how she was laughing at things again, and that it had been a really long time since she laughed that easily. So, I think it's more in retrospect people recognize how depressed and stressed they were." Tangible gestures of appreciation "We had letters and cards from churches and girl scout troops and individuals. And restaurants would feed us multiple times a day! It was unbelievable support from the community! I cannot get over it. It was just our whole breakroom was covered. There was not a wall to be found. It's just covered in letters and notes and posters and kindness! The food! It was unbelievably amazing, the community support! One group even bought us socks, warm, fuzzy socks, right. So, like when we were home, we were kind of hugged by our feet."

Tangible gestures of appreciation "When I first started in the COVID ICU it was really rewarding because family members were so thankful and they would be saying things like, "God bless you for being there and taking care of my loved one when I can't really be there," you know. So, it's nice when weight of what you're doing is really appreciated." 

Was there consistency between the data presented and the findings?

Results/Discussion 31. Clarity of major themes Were major themes clearly presented in the findings?

"We are conducting this study because we want to better understand changes to ICU organization and processes of care that have taken place across hospitals in response to the COVID-19 pandemic. We expect that this interview will last around 45 minutes, but it may take more time than that depending on the duration of your responses. If that's the case, does your schedule permit us to go beyond 45 minutes if needed?

"In addition, we are interested in learning from your experiences as a clinician during the pandemic. For this reason, it is possible that some questions may be difficult to answer or trigger an emotional response. You do not have to answer any questions you do not wish to discuss, and you can also stop the interview at any time.

At the end of the interview, we will provide you with resources for mental health support in case they are helpful to you.

"This interview will be audio-recorded for accuracy. As explained during informed consent, all responses are confidential. Your name will not be linked to this interview; once you consent to be recorded, I will only refer to you by a study ID number. Your participation is voluntary and you may choose to end the interview at any time. Do you have any questions before we get started?" FRAMING FOR DOMAIN-BASED SECTIONS "As I mentioned, we are interested in how the COVID-19 pandemic has impacted ICU organization and processes of care at your hospital. We are also interested in learning from your experiences as a clinician during the pandemic. We've found that it is often useful to talk about these issues in the context of specific experiences, but we also invite you to think more broadly if other things come to mind related to these outcomes while we are asking these questions."

"This interview will be divided into several key domains called the "4 S's." These include Space, Staff, Stuff, and System, and are important components of surge capacity planning for hospitals. We'll finish with some questions on end of life care in the setting of the COVID pandemic. Do you have any questions before we begin?"

Let's start with some baseline information about your role and practice:

• What is your professional background?

• What is your current role at your institution? For how many years have you been in your current position? • Do you have any other responsibilities for research, teaching, or administration? How much time do you commit to these activities? Has this changed since COVID started?

We'd also like to understand where you are in this pandemic.

• Have you experienced a surge of patients yet? o If yes, are you past the surge or in the middle of it now? o If yes: most of the questions below will refer to the time when your ICU was experiencing the peak surge, although we recognize that this has been very dynamic situation and welcome any comments about how things have changed over time as well.

The first domain I'd like to discuss with you is the first "S", or "Space." This refers to Hospital Capacity in the Setting of the COVID-19 Pandemic.

• Could you describe some of the strategies your hospital has used to accommodate an anticipated or actual increase in patient volume due to the COVID-19 pandemic?

• Probe: cohorting of COVID-19 patients in specific units, changes in number of ICU beds, canceling elective surgeries/admissions, repurposing non-ICU beds for critically ill patients (e.g.,

• How has the COVID-19 pandemic changed your interactions with patients and their families?

• Are palliative care services utilized in your ICU? If so, how? o Has this changed with the COVID pandemic? • What kinds of ethical dilemmas have you experienced as a result of the COVID-19 pandemic?

o Probe: use of experimental therapies, allocation of scarce resources o How have these ethical situations affected your current practice? • Do you know of any ethical guidelines or policies being implemented at your hospital to help guide your treatment strategies? • Has your hospital placed any limits on whether CPR is offered, or how long to attempt resuscitation, in the case of a COVID-19 patient experiencing cardiac arrest?

Wrap Up: "That is all of our questions for today. Is there anything I haven't asked you about that you were hoping to discuss?

"Thanks so much for your time. As we wrap up, I want to take a moment to remind you that we have several resources available that can help provide mental health support during these challenging times which I will send you following conclusion of the interview." [STOP RECORDER] J o u r n a l P r e -p r o o f INTRODUCTION "We're conducting this study because we want to better understand changes to the structure, organization, and processes of care that have taken place in your ICU in the setting of the COVID-19 pandemic. We are particularly interested in learning from you given your role in ICU leadership. This interview will take approximately 40 minutes and will be audio-recorded for accuracy. As explained during informed consent, all responses are confidential and you can choose not to answer any question. Your name will not be linked to this interview; once you consent to be recorded, I will only refer to you by a study ID number. Your participation is voluntary and you may choose to end the interview at any time. Do you have any questions before we get started?" "Thank you for your time. [START AUDIO RECORDER] . This is [NAME OF INTERVIEWER]. Today's date is [DATE] . I am with participant [STUDY ID#]. Please confirm that you consent to be recorded."

"Once again, we're interested in how the COVID-19 pandemic has impacted the structure, organization, and processes of care in your ICU. We are also interested in learning from your experiences as an ICU director during the pandemic.

Variation in Initial U.S. Hospital Responses to the Coronavirus Disease 2019 Pandemic

Snowball versus respondent-driven sampling

How many interviews are enough to identify metathemes in multisited and cross-culture research? Another perspective on Guest, Bunce, and Johnson's (2006) Landmark Study

Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups

The impact of hospital and ICU organizational factors on outcome in critically ill patients: results from the Extended Prevalence of Infection in Intensive Care study

A Conceptual and Adaptable Approach to Hospital Preparedness for Acute Surge Events Due to Emerging Infectious Diseases

The qualitative content analysis process

Using the framework method for the analysis of qualitative data in multi-disciplinary health research

Article 72. step down units, operating rooms, procedure suites); opening beds in nonclinical areas (e.g., lobbies, multipurpose and educational spaces, surge tents), diverting patients to other hospitals • What do you think about these strategies? Are there any that you felt worked particularly well, or that didn't work well? The next

• Does your ICU have medical students, residents, or fellows? If so, how has their clinical and educational experience been impacted by COVID? • Probes: removing housestaff from code teams, avoiding rooms of patients who are confirmed positive or PUI, asking them to take on new clinical responsibilities • How is your hospital handling healthcare workers who are exposed or contract the virus?

• Has your employer taken any actions to help ease the burden of being a frontline provider during a pandemic? • Probe: childcare, parking, food, hazard pay, opt out of work policies, alternative living arrangements, occupational/employee health resources • How has the pandemic impacted the well-being of you and your colleagues? The third "S" has to do with "stuff," meaning the physical equipment required to deliver care and support care delivery

• Did you and/or your colleagues receive any training how to don/doff PPE safely? If yes

Since the pandemic started, have you experienced any shortages in personal protective equipment? • If yes: have any PPE shortages affected your ICU or treatment decisions for patients? If so, how? • How have your employer's recommendations about PPE use for healthcare workers changed over time during the pandemic? • Probe: use of airborne precautions, universal masking, reuse or extended use of masks

Systems refers to the planning and leadership to operationalize and optimize a response effort. • How has the COVID-19 pandemic changed usual clinical practices such as daily rounds, the availability of consultants, or bedside tests/procedures? • Can you describe any guidelines or policies put in place that limit the number of hospital visitors during the pandemic? • If yes

Each section with begin with a few structured, closedended questions in order to provide us with a better understanding of the structure and organization of your ICU. We'll finish each section with more open-ended questions about the experiences of you and your staff providing critical care for patients in the setting of the COVID-19 pandemic. Do you have any questions before we begin?" First, I would like to understand more about your role and practice with a few background questions: • What is your profession? • What is your current role at your institution? • How many years have you been in your current position? • What, if any, is your role in COVID planning at your hospital? The first domain I'

the surge or in the middle of it now? • Are patients with suspected or confirmed COVID-19 cohorted together in your hospital? Yes/No • If yes, does this apply to both ICU and non-ICU patients with COVID? • Have there been changes in the number of ICU beds in your hospital in response to the COVID-19 pandemic? • If yes, how has your hospital increased the number of available ICU beds? • Probe: Repurposing non-ICU beds in clinical areas; placing new ICU beds in nonclinical areas (e.g., lobbies, multipurpose and educational spaces, surge tents), canceling elective surgeries/admissions, diverting/transferring patients to other hospitals or accepting patients from other hospitals • Could you describe some of the policies or strategies your ICU developed or adapted to accommodate an anticipated or actual increase in patient volume due to the COVID-19 pandemic? • Probe: For example

• The next few questions are going to focus first on your unit's staffing model before the pandemic started. • What was the staffing model in your ICU? Open, closed, or other. • If other: please describe the staffing model used in your ICU

Since the pandemic started, has the staffing model in your ICU changed? If so, how? • Probe: Have you used tiered staffing models like the one recommended by SCCM? What's the role of APPs? Have non-intensivists been providing critical care? Extending shifts, expanding workforce through volunteerism, graduating medical students early • In your ICU, are any providers without critical training providing critical care? Yes/No • If yes, what is the background of those providers? Do they receive any additional training in critical care before stepping into this role? • How has the role of housestaff/trainees in direct patient care changed over the course of the pandemic? • Has the number or role of nurses, respiratory therapists, or other staff changed in response to the pandemic? Yes/No • If yes, how? • How is your hospital handling healthcare workers who are exposed to COVID-19? How has this impacted staffing your ICU? • Probe: did you have systems in place for jeopardy or sick call pre-COVID? How have these changed? • From your perspective as an administrator/ICU director, how has the pandemic impacted the well-being of your staff? • Have you had any conversations to debrief with staff about what's working or not working in your ICU? If yes, what have you learned from the staff during these debriefs? The third "S" has to do with "stuff," meaning the physical equipment required to deliver care and support care delivery. • Let's start by discussing Personal Protective Equipment (PPE) • Did your hospital provide any training for how to use PPE: Yes/No • If yes, can you describe the format for the training? • Probe: watch a video, in person training, PPE buddies • Since the pandemic started, have you experienced any shortages in personal protective equipment? • What strategies have been used in your ICU to deal with PPE shortages? What has gone well?

• How have your employer's recommendations about PPE use for healthcare workers changed over time during the pandemic? If so, how has this been communicated to employees? • Probe: Have there been recommendations regarding the use of airborne precautions, universal masking, reuse or extended us of masks

• Let's move onto Patient Care Supplies • Since the pandemic started, have you experienced any shortages in any essential equipment or medications?• How have any equipment shortages been addressed in your ICU? J o u r n a l P r e -p r o o f • Probe: Have you been increasing supply orders, utilizing local stockpiles, conservation (e.g., reducing inefficient or nonessential use); substitution or adaptation (e.g., repurposing related supplies)?Now we're going to move onto the 4 th "S," which is "Systems." Systems refers to the planning and leadership to operationalize and optimize a response effort.• To the best of your knowledge, around what date did COVID planning begin at your hospital?• How has the COVID-19 pandemic changed your ICU's usual practice or workflow in any way that we haven't already discussed? J o u r n a l P r e -p r o o f