Sundus Siddique MBBS,MPH ,Doctoral Research Assistant, Center for the Promotion of Health in the New England Workplace (CPH-NEW); Department of Public Health, Zuckerberg College of Health Sciences, University of Massachusetts Lowell; Rebecca Gore PhD, Center for the Promotion of Health in the New England Workplace (CPH-NEW); Department of Biomedical Engineering, Francis College of Engineering, University of Massachusetts Lowell, Lowell, MA 01854; Yuan Zhang, PhD, Solomont School of Nursing, Zuckerberg College of Health Sciences, University of Massachusetts Lowell, Lowell, MA 01854; Laura Punnett, ScD ,Center for the Promotion of Health in the New England Workplace (CPH-NEW); Department of Biomedical Engineering, Francis College of Engineering, University of Massachusetts Lowell, Lowell, MA 01854

Purpose/Objectives
The objectives of this study are: (1) to examine the association between job demands and job resources with emotional exhaustion scores in a population of clinical and non-clinical personnel where unlicensed assistive personnel predominated over licensed staff; and (2) to examine the additional role of job safety hazards in mediating and/or moderating the association between organizational support for safety and emotional exhaustion.

Background
Emotional exhaustion, an early manifestation of burnout, is related to preventable work environment exposures in licensed clinical professionals. This association remains understudied in unlicensed assistive personnel, such as medical assistants and allied healthcare workers, and non-clinical personnel. Most studies focus on demands (workload) and resources but ignore the potential impact of workplace safety – including organizational-level policies and practices as well as job-level hazardous work conditions. We examined these understudied exposures as risk factors for emotional exhaustion in a mixed population of non-clinical and clinical healthcare staff, using a novel mediation analysis approach proposed by Valerie and VanderWeele (2013).

Methods
A self-administered survey was distributed to all employees of five U.S. public sector healthcare facilities in 2018-19. Burnout was defined with the ?Job Demands and Resources? model and measured with an adapted (6-question) version of the Oldenburg Burnout Inventory; three of these items were averaged to generate a continuous scale of emotional exhaustion. Job demands and resources included organizational support for safety (NOSACQ), job safety, psychological demands, physical demands, emotional labor, assault and negative acts (e.g., bullying). Regression, moderation and causal mediation analyses were conducted using SAS. Causal mediation was examined using the VanderWeele method (2014) which decomposed the mediation effect into four components (Figure 2). Data has been collected and analyses have been completed.

Findings
A total of 1,060 surveys were collected from a predominantly female population. Mean emotional exhaustion (~3.0) scores were high and uniform among study sites. Emotional exhaustion was higher in women (p=0.0018), workers in direct patient care jobs (p<0.0001), and younger (p=0.0009) subjects. Spearman correlation coefficients among independent variables ranged from 0.3 to 0.7 (all p-values <0.0001). In multivariable linear regression models, job safety, emotional labor, psychological demands, physical demands, job strain, and occurrence of assault and negative acts were positively associated with emotional exhaustion while organizational support for safety was negatively associated. About 30% of the association between organizational support for safety and emotional exhaustion was mediated by job hazards and 11% of the total effect was due to moderation (p<0.0001). Job hazards served as both mediator and moderator (Figure 1 and 2).(Figures 1 and 2 have been uploaded in supporting documents section since the formatting was affected when uploaded here).

Discussion
Emotional exhaustion was lower among those reporting more organizational support for safety and higher among those reporting more job hazards. The association of organizational support for safety in reducing emotional exhaustion was mediated through job-level safety hazards. The simultaneous moderation demonstrated that organizational support for safety had less benefit on reducing emotional exhaustion when job hazards were high, suggesting that failure to put policy into practice was itself a risk factor. These models demonstrated that two interdependent changes could reduce emotional exhaustion, i.e., reduction of job hazards and improvement of organizational safety support for safety. Feasible ways to improve both safety and staff engagement include providing adequate tools and time for performing tasks, ensuring adequate staffing, and supporting staff who call attention to and refuse to work in potentially harmful situations. To our knowledge this is the first study to examine burnout with this modeling technique and the first to report the effect of job safety on emotional exhaustion.

Conclusions
Organizational commitment to employee safety, which manifests through prevention of day to day job hazards, could improve the quality of life of healthcare workers. Future longitudinal studies are needed to further examine this association.

Tags: Applied research, Basic research, Empirical study, Hazardous Work Environments and Safety, Health care and social assistance, High Risk Jobs; Vulnerable/At-Risk Populations, Job and Task Design, Organization- and Job-Level Environments and Practices, Organizational Practices, Practitioner report or field study, Psychological and Biological Effects of Job Stress, Safety Climate; Safety Management; and Training, Social and Organizational Environment, Theoretical and Conceptual Issues in Job Stress, Traumatic Stress and Resilience, Workplace Injuries and Illnesses, Workplace Stress; Outcomes; and Recovery