Introduction

The coronavirus disease 2019 (COVID-19) pandemic created challenging work conditions across the healthcare system. Healthcare personnel (HCP) have been faced with uncertainty about resources, staffing, risk of SARS-CoV-2 exposure, concern for the well-being of coworkers and family, and providing clinical care amid visitor restrictions. Many studies have documented increases in anxiety, depression, stress disorders, and burnout among HCP since the start of the COVID-19 pandemic [1–8]. In addition to causing significant individual distress, these symptoms are associated with a higher likelihood of leaving the field of medicine [9]. Data from a public health workforce survey in 2021 suggest that just under half (44%) of workers reported that they were considering leaving their jobs within the next 5 years for retirement or other reasons, with the majority (76%) having begun to think about leaving since the start of the COVID-19 pandemic [10]. This is an alarming trend for maintaining health system capacity in the US, and information about how to address stress and burnout is urgently needed. Frontline HCP, particularly those working in emergency departments (EDs) and critical care units, have long been recognized to be at increased risk for stress and burnout and this risk was further exacerbated by the pandemic [11–15]. We previously reported prevalent symptoms of anxiety and burnout across the spectrum of ED staff during the initial wave of the pandemic and here examine support program availability and use in this at-risk population [16, 17].

Support programs for frontline HCP can mitigate the negative effects of the pandemic on HCP well-being and preserve the ability of frontline HCP to provide medical care during this public health emergency [18–26]. Research suggests that programs targeting both individual (HCP-targeted) and structural (workflow-targeted) factors are important [27]. Several types of programs have been evaluated, including education, peer support, mental health treatment, and changes to clinical workflow, with most have being evaluated in single-site studies [28–32]. Data on how hospitals implemented support programs and how they were used in the context of the COVID-19 pandemic has been lacking, however [15]. Studying the COVID-19 experience is important to understand the long-term impact of COVID-19 on health care careers, evaluate the effectiveness of the COVID-19 health system response, and maintain readiness for future public health crises.

The goals of this study were to describe the availability and use of institutional support programs targeting both individual ED staff and clinical workflow during the early COVID-19 pandemic, and to identify HCP and facility characteristics associated with program use.

Materials and methods

Results

Participants

A total of 1,541 HCP completed the survey (response rate 96.0%), including 762 (49%) physicians or advanced practice providers, 396 (26%) nurses, and 383 (25%) non-clinical staff. Their median age was 36 years (IQR 30–45 years). Most were female (63%) and non-Hispanic White (81%). At the time of survey completion, the median ED COVID-19 patient volume was 107 patients (IQR 57–271) per facility per month (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Emergency Department (ED) Health Care Personnel (HCP; N = 1,541) and facility COVID-19 volume for 21 participating COVERED EDs during the COVID-19 pandemic, June–September 2020.

https://doi.org/10.1371/journal.pone.0298807.t001

Support program availability and use

The proportion of sites with each type of program and each program category available and HCP use are shown in Fig 1 and S1 Table. S2 Table presents information about which of the classification criteria resulted in coding each program as available. In 61% of cases the program was coded as available due to site investigator report and in 39% of cases the program was coded as available due to program use by ≥ 2 HCP. For HCP individual support programs, the proportion of sites that had at least one type of program available from each category was 95–100%. Individual support program category use ranged from 6% for emotional support to 13% for work demand mitigation. All sites had at least one type of program in each of the clinical work program categories in place in their ED (i.e., COVID-19 exposure reduction, patient care services, or patient–family communication facilitation and ED team communication). Clinical work support programs with the lowest availability were ethics consultations (47%) and self-administered swabs for COVID-19 patient testing (33%). Use of clinical work support programs ranged from 24% for patient–family communication facilitation to 43% for use of patient care services, which was driven by high social worker and interpreter involvement in care (24% and 36%, respectively) and low use of palliative care (7%) and ethics consultations (1%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Availability and use of health care personnel support programs across 21 COVERED emergency departments during the COVID-19 pandemic, June–September 2020.

A. Individual Employee Support Programs among all HCP participants, B. Clinical Work Support Programs among clinical HCP. The black bars indicate the proportion of participants for whom that support program type and/or category of program was available during the study period. The colored bars indicate the proportion of participants who used that support program type or category. Category pooled totals are represented in all capital letters.

https://doi.org/10.1371/journal.pone.0298807.g001

Emotional support programs, such as stress reduction and mental health hotlines, were reported by site investigators to be available to 95% of participants. Overall, 6% of HCP reported using these programs at some point during the COVID-19 pandemic prior to completing the survey. In our cohort, 57% (n = 787/1393) of HCP reported elevated COVID-19–related stress or anxiety and 12% were at elevated risk of PTSD assessed via the Primary Care PTSD Screen for DSM-5 [PC-PTSD-5]; n = 170/1398). Use of emotional support programs did not significantly differ between HCP who reported elevated stress or anxiety or were at risk of PTSD based on the PC-PTSD-5 and those who were not (see Table 2). Use of emotional support programs was 7.5% among HCP with elevated COVID-19–related stress or anxiety and 9.3% among those who were at elevated risk for PTSD based on the PC-PTSD-5.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic and facility characteristics by use of individual COVERED emergency department health care personnel support program categories during the COVID-19 pandemic, June–September 2020.

In this table, HCP use of each support program category is listed.

https://doi.org/10.1371/journal.pone.0298807.t002

HCP and facility characteristics associated with program use.

Several factors were associated with program use. Instrumental support program use differed by age, with a higher rate of use among HCPs ages 30–39 years than those younger than 30 years (13.8% vs. 8.4%; adjusted odds ratio [aOR] 1.61, 95% CI 1.01–2.56). Instrumental support programs were more likely to be used by physicians and advanced practice providers than nonclinical staff (12.1% vs. 6.8%; aOR 1.84, 95% CI 1.15–2.94). Work demand mitigation program use was most common among physicians/advanced practice providers (16.1% compared with 9.6% among nurses [aOR 1.81, 95% CI 1.23–2.67] and 10.7% among non-clinical staff [aOR 1.59, 95% CI 1.09–2.32]). Staff who worked fewer hours reported more use of work demand mitigation programs (18% for those working 20 hours per week vs. 10.2% working 40 or more hours per week; aOR 1.95, 95% CI 1.17–3.28). Physicians/advanced practice providers reported lower use of financial compensation programs than nurses (7.0% vs 16.4%; aOR 0.35, 95% CI 0.24–0.53) or non-clinical staff (9.9%; aOR 0.62, 95% CI 0.39–0.97). HCP who worked at facilities with the lowest COVID-19 patient volumes reported using financial compensation programs at the highest rate (20.5% in facilities with COVID-19 patient volume <50 in the past month vs. 10.2% in facilities with volume 50–99 [aOR 2.32, 95% CI 1.11–4.88], 5.6% in facilities with volume 100–250 [aOR 4.60, 95% CI 2.12–9.98)] and 7.4% in facilities with volume >250 [aOR 3.38, 95% CI 1.56–7.34]).

Clinical work program use was similar across HCP/facility characteristics, except that non-Hispanic White HCPs reported using programs in the patient care services category (47.8% vs. 30.6%; aOR 1.99, 95% CI 1.55–2.55) and patient-family communication category (25.7% vs. 18.7%; aOR 1.43, 95% CI 1.07–1.92) at a higher rate than HCPs from other racial/ethnic groups.

Discussion

In this study we describe the availability and use of institutional support programs for frontline ED HCP at 21 academic hospitals in the early COVID-19 pandemic before vaccine availability and identified HCP characteristics that were associated with program use. Nearly all sites had at least one program from each category that was included in our survey (95–100%), though there were differences in availability across program types. Programs to mitigate work demands via surge staffing, paid time off, and COVID-19 testing for staff were widely available (90–100%). In contrast, elder care, disability services after SARS-CoV-2 infection, and additional payments for frontline workers were less available. Our primary observation was that despite relatively high availability of many types of individual staff support programs, HCP reported low use (<15% for each program).

Several medical professional organizations have published recommendations to improve staff support [39]. A systematic review, which included data from multiple prior virus outbreaks worldwide (i.e., severe acute respiratory syndrome [SARS], influenza caused by H1N1, Middle East respiratory syndrome [MERS], and Ebola), demonstrated that staff support protocols, clear communication, psychosocial interventions, and adequate infection protection were all important components of programming to improve the psychological wellbeing of medical staff during virus outbreaks [40]. Digital learning packages have been developed to aid in resiliency and coping in some medical centers [28]. Data regarding effective interventions to promote resilience in military settings and/or during natural disasters are relevant to inform programming during pandemics. For example, peer-support interventions developed initially for the military have also been used in the context of the COVID-19 pandemic [29], as peer-support has the advantage of being rapidly deployable and promoting interpersonal relationships. Indeed, in a qualitative study of HCP participating in the COVID-19 Pandemic and Emotional Well-Being Study, participants highlighted personal strategies for maintaining social connectedness that were important in coping with uncertainty from their work [41].

Our primary analyses examined rates of program use across ED HCP, although different HCP are likely to have greater needs for certain interventions than others. For this reason, additional research to target support programs optimally for those that are most likely to benefit is essential. In our study, we examined support program use comparing HCP with and without elevated stress or anxiety. Among the 57% of ED HCP who reported elevated stress or anxiety in our sample, less than 10% had engaged with an institution-provided emotional support program since the beginning of the pandemic, a rate of engagement that was similar to HCP without elevated stress or anxiety (8% vs 5%, respectively). The rate of support program use <10% should be considered in the context of the societal turmoil and work-related demands placed on ED HCP during the early period of the COVID-19 pandemic, and it suggests that employer-provided emotional support services may not reach the population of HCP most likely to be affected. The low rate of support program use highlights the importance of work to better identify social and practical barriers to institutional program use to support frontline HCP during future widespread health crises. In addition to work demands, other potential barriers may include HCP concerns about the efficacy of programs and/or concern about stigmatization by coworkers for utilizing psychosocial assistance programs.

We observed some HCP and facility characteristics that were associated with differential rates of support program use, which should be viewed as hypothesis-generating given that HCP were not asked about their perceived needs or the applicability of each program to their individual circumstances. First, HCP working over 40 hours per week reported the lowest use of individual staff support programs. Full-time staff working extra hours may not have time to engage with institutional programming, which could be one significant barrier to participation. Institutional support programming may have required additional time that was not perceived as an effective countermeasure for stress, anxiety, burnout, and PTSD by all. It is also possible that HCP who worked more hours became more familiar with use of PPE and might not have experienced the same level of concern about risk. Second, facilities with the highest COVID-19 patient volumes had the lowest proportion of HCP receiving extra financial compensation. This observation raises the possibility that in summer 2020, those facilities that were most burdened with pandemic-related demands may have had less capacity to roll out new financial support programs. This period was during the first wave of COVID-19 in most US locations and travel continued to be limited.

This study has several limitations. First, participants were ED HCP from primarily large academic and urban centers in the US with demographic characteristics generally consistent with the US hospital workforce [42]; however, their experiences may not reflect those of community and rural ED HCP. Second, information about institutional program use was collected only at one time point, and thus the cross-sectional nature of the data precludes an assessment of changes in program availability and use over time. It is possible that HCP with anxiety, stress, burnout, and/or PTSD symptoms used program resources later in the pandemic. However, information regarding HCP needs early in the pandemic are important to prepare for the future, and the current data build on prior research from early responses during other virus outbreaks. Third, our survey asked only about availability and use of institution-based programs and did not gather information about other potential sources of support for HCP outside of work, HCP’s perception of unmet program needs, the perceived or actual effectiveness of these program types, or satisfaction with programming. Fourth, our definition of program availability relied on site investigators and HCP’s awareness of existing programing at their hospital and this approach likely missed some support programs that were available but unknown to either study investigators or HCP participants. HCP were not asked about their awareness of programs and instead reported only about their individual program use. We classified a program as available at a site if the site investigator indicated that it was available or if 2 or more HCP participants at the site had used it. In 39% of cases programs were coded as being available based on HCP use (i.e., instances where the site investigator had not identified the program as present). Using this flexible definition allowed us to utilize all information collected in the study regarding program availability, but the data suggest that lack of awareness of support programming is an important issue for facilities and ED departments to address. Fifth, the small sample prevented substantive analysis stratified by race/ethnicity. Future efforts can focus on improving data, including by race/ethnicity, to reflect diversity of the workforce. Finally, these data are from the initial wave of the COVID-19 pandemic from June to September 2020, a time of uncertainty about COVID-19 transmission, prior to availability of the COVID-19 vaccine, when personal protective equipment was limited at many sites in the U.S. Indeed, our data do not reflect changes in hospital programming, increasing HCP burnout, and changes in clinical practice as the pandemic has proceeded. Nonetheless, our data from frontline HCP can be used to inform our preparedness for future widespread health disasters and the impact of those events on a capable HCP workforce.

Conclusions

In conclusion, institutional support programs for frontline ED HCP were widely available early in the COVID-19 pandemic. Despite availability and the high prevalence of symptoms of stress, anxiety, and burnout, use of individual employee support programs was low, highlighting the importance of future work to identify and address barriers to accessing current services and the need to develop more effective interventions to protect and improve ED HCP well-being during times of crisis.

Supporting information

Acknowledgments

The authors thank the COVERED participants, the participating institutions/emergency departments, and individuals. The authors acknowledge the following participating Project COVERED emergency departments: Allegheny General Hospital, Pittsburgh, PA; Baystate Medical Center, Springfield, MA; Denver Health Medical Center, Denver, CO; Detroit Receiving Hospital/Sinai–Grace Hospital, Detroit, MI; Hennepin County Medical Center, Minneapolis, MN; Jackson Memorial Hospital, Miami, FL; Johns Hopkins Medical Institute, Baltimore, MD; University Medical Center, New Orleans, LA; Mount Sinai Hospital East/Elmhurst Hospital Center, New York, NY; Orlando Regional Medical Center, Orlando, FL; University of Alabama at Birmingham Hospital, Birmingham, AL; Ronald Reagan–UCLA Medical Center/Olive View–UCLA Medical Center, Los Angeles, CA; University of Iowa, Iowa City, IA; University of Massachusetts Memorial Medical Center, Worcester, MA; University of Mississippi Medical Center, Jackson, MS; UCSF Zuckerberg San Francisco General, San Francisco, CA; UT Southwestern Medical Center/Parkland Memorial Hospital, Dallas, TX; Truman Medical Center, Kansas City, MO; Thomas Jefferson University, Philadelphia, PA; and Washington University Barnes–Jewish Hospital, St. Louis, MO.

The authors acknowledge the following individuals for their help with the study: Aishat Adeyemi, BA, Lisa Allen, MPH, MPA, Gregory Almonte, Otuwe Anya, MS, Paula Arellano-Cruz, BS, Ruzana Aronov, Danielle Beckham, RN, MSN, CRC, Lauren Buck, BS, Samuel Ceckowski, BS, Maxime Centeno, BS, Virginia, Chan, BS, Anna Marie Chang, MD, MSCE, Melissa Connor, RN, Gabriella Dashler, BS, Jenna Davis, MSN, Cynthia Delgado, BS, Veronica Delgado PA, MS, Brianna DiFronzo, Radhika L Edpuganti, BS, Alyssa Espinera, MD, Fresa Estevez, BA, Shelly Ann Evans, MBA, RN, Cathy Fairfield, BSN, Phillip Fairweather, MD, Theodore Falcon, BS, Brian Fuller, MD, MSCI, David Gallegos, BBA, Samuel Ganier, MD, Stephanie Gravitz, MPH, Jeffrey Harrison, RN, Kyle Herbert MD, Judy Hermans, MPH, Emily Hopkins, MSPH, Alan Jones, MD, Kia M. Jones, DrPH, Momina Khan, BS, Laura Iavicoli, MD, Robin Kemball, MPH, Laurie Kemble, BHSRT, CCRC, Stuart Kessler, MD, Catherine Lind RN, NP, Karina Loayza, LCSW, Carol Lynn Lyle, PA, MPH, Virginia B. Mangolds, MS, RN, Hannah Makarevich-Manilla, MPH, CCRC, Thomas Mazzocco, RN, Sarah Meram, MS, Valerie H. Mika, MS, Reynaldo Padilla, BA, Giacomo Passaglia, BS, Rebekah Peacock, BSN, Danielle Perez, BS, Kye E. Poronsky, MA, Eric Raines, EMT-P, Monica N. Ramage, MSN, RN, Kavita Rampertaap, MSN, RN, Sarah Reineck, BS, Nicole Renzi, RN, Erin P. Ricketts, MSPH, Stephanie Rodriguez, MA, Justin Sabol, BS, Valeria Samame, BA, Katie Schneider, MSN, Robert Sellman, PA, Kristine Sernulka, BSN, CCRP, Edward Sheriff, PhD, MPH, Jennifer Siller, DNP, RN, Colleen Smith, MD, Timothy Smith, BS, Patricia Slev, MD, Kelly Szabo, MPH, CCRC, Meghan Tinetti, BSN, CCRP, Denise Tritt, CIM, CRCP, Julia Vargas, BS, Samuel Vargas, BS, Kavey Vidal, BS, Lori Wilkerson, RN, CRC, Darleen Williams, DNP, APRN-CNS, Sallie Anne Wright, MPH, BSMT, and Isaias Yin, LVN. Additionally, the authors acknowledge Paul Casella, MFA for his editorial assistance.

The Project COVERED Emergency Department Network includes the following: Monica Bahamon, MPH, Jestin N. Carlson MD, MSc, Makini Chisolm-Straker, MD, MPH, Brian Driver, MD, Brett Faine, Pharm D, MS, Brian M. Fuller, MD, James Galbraith, MD, Philip A. Giordano, MD, John P. Haran, MD, PhD, Elisabeth Hesse, MD, MTM&H, Amanda Higgins, MS, Jeremiah Hinson, MD, Stacey House, MD, PhD, Ahamed H. Idris, MD, Efrat Kean, MD, Elizabeth Krebs, MD, MSc, Michael C. Kurz, MD, MS, Preeta K. Kutty, MD, MPH, Lilly Lee SM, MD, Stephen Y. Liang, MD, MPHS, Stephen C. Lim, MD, Clifford McDonald, MD, Gregory Moran, MD, William Mower, MD, PhD; Utsav Nandi, MD, MSCI, Kavitha Pathmarajah, MPH, James H. Paxton MD, Yesenia Perez, BA, Lynne D. Richardson, MD, Richard Rothman, MD, PhD, Walter A. Schrading MD, Jessica Shuck, BA, Patricia Slev, MD, Howard A. Smithline, MD, Michelle St. Romain, MD, Kimberly Souffront, PhD, FNP-BC, RN, Mark T. Steele, MD, Amy Stubbs, MD, Morgan B. Swanson, Josh Tiao, MD, Jesus R. Torres, MD, MPH, Stacy A. Trent MD MPH, Lisandra Uribe, BS, Arvind Venkat, MD, Gregory Volturo, MD, Kurt D. Weber, MD, and James Willey, MD. The lead author of this authorship group is Nicholas Mohr, MD (nicholas-mohr@uiowa.edu).

The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.