key: cord-0751653-lgy1mf2x
authors: Sangal, Rohit B.; Bray, Alexandra; Reid, Eleanor; Ulrich, Andrew; Liebhardt, Beth; Venkatesh, Arjun; King, Marissa
title: Leadership communication, stress, and burnout among frontline emergency department staff amid the COVID-19 pandemic: A mixed methods approach
date: 2021-08-13
journal: Healthc (Amst)
DOI: 10.1016/j.hjdsi.2021.100577
sha: 7c0f6d4dd4e9c2a699a0a32d7b355237a9593d34
doc_id: 751653
cord_uid: lgy1mf2x

BACKGROUND: Organizations have a key role to play in supporting healthcare workers (HCWs) and mitigating stress during COVID-19. We aimed to understand whether perceptions of support and communication by local leadership were associated with reduced reports of stress and burnout among frontline HCWs. METHODS: We conducted cross-sectional surveys embedded within emergency department (ED) workflow during the first wave of COVID-19 from April 9, 2020 to June 15th, 2020 within three EDs of a multisite health system in the Northeast United States. All ED HCWs were administered electronic surveys during shift via text message. We simultaneously conducted 64 qualitative interviews to better characterize and validate survey responses. Primary survey outcomes were levels of work stress and burnout. RESULTS: Over 10 week study, 327 of 431 (76%) frontline HCWs responded to at least one round of the survey. More useful communication mediated through higher perception of support was significantly associated with lower work stress (B = −0.33, p < 0.001) and burnout (B = −7.84, p < 0.001). A one-point increase on the communication Likert scale was associated with a 9% reduction in stress and a 19% reduction in burnout. Three themes related to effective crisis communication during COVID-19 emerged in interviews: (1) information consolidation prior to dissemination, (2) consistency of communication, and (3) bi-directional communication. CONCLUSION: This work suggests that effective local leadership communication, characterized by information consolidation, consistency, and bi-directionality, leads to higher perceptions of support and lower stress and burnout among ED frontline workers. As the pandemic continues, these results present an evidence-based framework for leaders to support frontline HCWs through effective crisis communication.

Murray RN, Lori Franson EDT. We also are grateful to Rob Teresi and Leilah Harouni for research assistance.

COVID-19 has placed extraordinary burden and prolonged stress on frontline healthcare workers 25 (HCWs) . The emotional toll that working on the frontline, especially in the emergency 26 department (ED) during COVID-19, has taken is evidenced in elevated rates of stress and stress-27 related conditions. 1-3 This is superimposed on the already high levels of stress and burnout seen 28 in HCWs. 4 29 30 Organizations and leaders have a key role to play in supporting HCWs and mitigating stress. It is 31 not possible to avoid crises with detailed planning alone, and so leaders must reduce the impact 32 of crisis and promote adaptation and flexibility to advance through the crisis. This involves 33 building trust, coordinating multiple groups of stakeholders, and helping teams (managers, staff 34 etc) make sense of new information. 5 Effective crisis communication and information 35 dissemination is critical for stress reduction. Additionally, COVID-19 has created challenges in 36 information access which increases anxiety for HCWs. 6-8 Therefore, a key challenge presented 37 by COVID-19 was providing HCWs with essential information while managing constantly 38 changing guidelines, policies and resources. For example, the World Health Organization 39 guidance on clinical management expanded from 10 to 62 pages between January 2020 and May 40 2020 illustrating the rapid influx of new information. 9,10 41 42

The nature of the crisis necessitated simultaneously managing two paradoxical information 43 pathologies: information anxiety and information overload. 11 The challenge for frontline leaders 44 was to reduce anxiety that arises from uncertainty while not increasing stress from information 45 overload. Managing information effectively is one critical way in which organizations can 46 support employees during times of crises. Previous work has shown that organizational support 47 is an important mechanism for reducing employee burnout during crises or trauma. 12 communication and perceptions of support in this context. To assess communication, we asked 73 on a five point Likert scale "How useful are the communications you're receiving from the ED 74 COVID Task Force." To assess perceptions of support, we asked "How supported do you feel by 75 YNHH right now to do your job effectively." This measure was adapted from the organizational 76 support item in the Trauma-Informed Organizational Culture survey. 12 We focus on 77 organizational support since it has previously been found to have a larger than typical effect on 78 burnout reduction. 79

The questions were first pilot-tested in a six person panel with clinical, leadership, and survey 81 design expertise. In addition, an open response text box allowed frontline staff to provide 82 feedback and suggest changes which were later used for qualitative analyses. The survey items 83 were then pilot tested among a broader group of twelve respondents who provided feedback on 84 item clarity, instructions, and survey length. 85 86

We conducted repeated cross-sectional surveys of all ED HCWs including attending and resident 88 physicians, advanced practice providers, clinical and administrative nurses, and patient care 89 associates (ED technicians). Given time pressure on the frontline, survey questions were rotated 90 through survey iterations to increase the spread of data desired, surveys were kept brief, and 91 surveys were sent out at times that would be least disruptive to workflow. For example, since 92 J o u r n a l P r e -p r o o f burnout is a result of longer term stress, such questions were asked in later weeks of the study. 93

HCWs were contacted via text message fifteen minutes after their shift ended. All frontline 94

HCWs were eligible to participate in the survey including environmental services, registration 95 and security. Each HCW received a text every six shifts with an opt-out option. Additional 96 recruitment materials were provided in department announcements and bulletin board materials. 97

All communications were done in English. 98

In addition to surveys, we conducted in depth interviews with ED HCWs between April 14, 2020 100 and June 21, 2020 via telephone. Interviews followed a semi-structured interview protocol (See 101 supplementary for protocol), which included questions about communication, support, and 102 burnout. Interviewees were recruited via email snowball sampling. We reached out to 121 103 individuals asking them to participate and completed 64 interviews. Interviews were conducted 104 by a team of three professors and two PhD students led by one of the authors (MK) with 105 expertise in qualitative methods and healthcare and had no prior relationship to interviewees. 106

Interviews were conducted over the telephone and recorded with the permission of interviewees 107 and professionally transcribed. Interview summary notes were completed using a standardized 108 form immediately after the completion of the interview and individual interviewers reviewed the 109 transcriptions. Since the data was collected without identifiers, transcripts and quotes were not 110 returned to participants for comments and/or corrections. Data collection for this study ended in 111 mid-June 2020 as COVID-19 cases in the hospital approached a nadir. 112

Analysis 113 J o u r n a l P r e -p r o o f

We examined survey data for associations between communication and perceptions of 114 organizational support and our primary outcome variables of burnout and work stress. We 115 estimated two sets of regression models: linear regressions and structural equation models for the 116 primary outcome. Regression models were adjusted for COVID-19 volume, the ED total volume, 117 and staff occupation since these are likely to be associated with stress. Standard errors were 118 clustered by day to account for the likely correlation among responses received on the same day. 119

The first set of models relied on cross-sectional data to establish an association between 120 communication and the primary outcomes. Results 150

Of the 431 frontline staff surveyed, 327 (76%) responded at least once to the text message based 152 survey. Table 1 summarizes the characteristics of survey respondents and shows baseline levels 153 of stress and burnout. Frontline staff reported lower levels of stress (coefficient B=-0.32; 154 p<0.001) and burnout (B=-7.8; p<0.001) when communication was perceived to be more 155 effective controlling for COVID-19 caseload, occupational role, and total ED volume ( Table 2) . 156

Based on the regression models in Table 2, a one point increase on the communication scale was  157 associated with a 9.45% reduction in stress and a 18.97% reduction in burnout. 158

We further aimed to disentangle the extent to which communication directly alleviated stress or 160 

Establishing consistency of communication helped alleviate information anxiety. For example, 219 consolidating the pertinent information into a single email, sent at a specific time of day, allowed 220 the messaging to become as predictable as postal mail delivery. This email update was created by 221 department leadership and sent from the medical, advanced practice provider and nursing 222 directors to their respective staff. The email was formatted the same way with important changes 223 at the top, reminders in the middle and operational data at the bottom. One interviewee stated 224 "…It was like the Bible…You would look for that at exactly 6:00 AM in the morning…I know, 225 to me as a leader, that was huge. I felt like that was, kept me informed. I was able to speak to the 226 points. I felt well informed through this whole thing. So then if I feel good about it, I'm able to 227 articulate that to my people." This is consistent with research that suggests that routines and 228 habits can help reduce performance stress. 25 229 230

In order to further facilitate open communication, ED town halls were implemented to allow for 232 adaptability and real-time responsiveness. 26 While the health system implemented broader town 233 halls where questions were submitted in advance, having a department specific town hall allowed 234 more personalized interaction and be more of a conversation. Town halls also provided an open 235 forum for issues or questions that needed clarification or further operational refinement. They 236 provided an opportunity for ED HCWs to raise concerns. Town halls were held weekly but the 237 frequency could be adjusted based on pandemic developments. These were moderated by the 238 medical directors, department Chair/Vice-Chair but could be supplemented as needed with 239 updates from the emergency medical services director or residency director. During town halls at 240 the beginning of the crisis, chats frequently contained 150 to more than 200 comments and 241 questions. As one attending physician described, if a problem surfaced that was generating 242 concern they could approach leadership and say, "Hey listen, I'm hearing a lot of concerns about 243 X that you might not have realized is a problem. But maybe you want to address it in your next 244 town hall." implementing patient care orders placed by physicians and feel more exposed to the virus. 3,31 292

Moreover, ensuring broad participation and inclusivity has been demonstrated in a management 293 research as being critical for surfacing information and harnessing collective intelligence during 294 periods of organizational stability. 32 However, during crises, organizations frequently default to 295 command and control and more hierarchical forms of organizing. Our work demonstrates the 296 benefits of inclusivity and opportunities for broad participation in decision making during 297 crises. 33 298 299

Limitations 300

While our sample for a qualitative study is robust, it continues to represent a smaller proportion 301 of the overall workforce. Additionally, the study enrolled those willing to participate and thus 302 may have either a very positive or negative viewpoint to share. Despite this, responses generally 303 aligned to similar themes. It is also important to note that the results are not connected to any 304 specific intervention which limits our ability to causally establish the efficacy of specific 305 interventions. Additionally, while our models control for workplace variables, we did not include 306 non-workplace variables such as housing or financial situation. We hope future research will use 307 field experiments to examine whether the interventions we identify lead to decreased burnout 308 and stress during crises. Finally, perceptions of communication did not consistently increase and 309 began to decrease after the first wave of the crisis passed. Future work is needed to understand 310 how to most effectively communicate as crises wane. 311 312

Effective communication is associated with decreased work stress and burnout which was 314 mediated through increased perception of support across frontline HCWs tackling Effective crisis communication during COVID-19 involve bidirectional communication, 316 consistency of communication, and information consolidation prior to dissemination. While 317 further work is needed to causally identify interventions to improve crisis communication, our 318 work suggest that small changes in organizational communication are associated with 319 significantly lower levels of stress and burnout. As the pandemic continues or a second wave is 320 encountered, we provide a framework for leadership to effectively communicate in a crisis. Table 3 : Major themes and representative quotes for semi-structured interviews 454

Interview Themes Representative Quotes April 10 th , a resident wrote in the open survey item, "Timely overall medical strategy updates would help. What's in the works about awake proning, experimental treatments inpatient?" "Departments need to be better with communication...I don't understand why it is so hard for everyone to be on the same page. It's very frustrating." Nurse April 14h a physician assistant described being "inundated in the beginning," "the first week, two weeks, you're getting like 20 emails a day and trying to understand what's going on."

"I think that there was some growing pains in the beginning and less communication than we'd like... I think consolidating it into this is the one big one you need to pay attention to was really important." "It was a little much at first because things were changing so frequently, but it seems like now we have a system in place…the update is being rolled out as usual per every day and if anything new is changed, then it's highlighted within the update…we've kind of figured out what has been working well and we have fewer of those changes to our daily updates. Things seem to be communicated better."

"It got to the point where it was almost like ... It was like the Bible," according to a nurse manager. They continued, "You would look for that at exactly 6:00 AM in the morning. Even the charge nurses and the staff would look for it. "Are there any updates?" Preparing for the next shift. Making sure everybody has the information. I know, to me as a leader, that was huge. I felt like that was, kept me informed. I was able to speak to the points. I felt well informed through this whole thing. So then if I feel good about it, I'm able to articulate that to my people."

327 Prevalence of depression, anxiety, and insomnia among healthcare workers during the 328 COVID-19 pandemic: A systematic review and meta-analysis

Psychometric Properties 375 of the Copenhagen Burnout Inventory in an Academic Healthcare Institution Sample in 376 the U.S. Evaluation & the health professions. 2020:163278720934165. 377 19. Guest DE, Conway N. Communicating the psychological contract: an employer 378 perspective

An introduction to structural equation modeling

Introduction to mediation analysis with structural 384 equation modeling. Shanghai archives of psychiatry

Grounded theory research: Procedures, canons, and evaluative 386 criteria

Don't stop believing: Rituals improve 391 performance by decreasing anxiety. Organizational Behavior and Human Decision 392 Processes

Relationships and Resilience: Care Provider Responses to Pressures From 396 Managed Care

Compassion satisfaction, 398 burnout, and compassion fatigue among emergency nurses compared with nurses in other 399 selected inpatient specialties

High Prevalence 401 of Burnout Among US Emergency Medicine Residents: Results From the 2017 National 402 Emergency Medicine Wellness Survey. Annals of emergency medicine

Impact of Burnout on Self-405 Reported Patient Care Among Emergency Physicians

Psychological Safety and Learning Behavior in Work Teams Amy 408

Evidence of a Collective 410 Intelligence Factor in the Performance of Human Groups