UC Berkeley

Objective: Peer support specialists (PSs) have personal experience in mental health recovery and are hired by service agencies to assist others struggling with similar challenges. Stress impacts work performance among helping professionals, and work-related stress is thought to be particularly detrimental to PSs’ performance given their history of mental illness. However, this topic has received little empirical attention. Furthermore, different forms of stress may be closely related, casting doubt on the utility of work-specific stress measurement. This dissertation examines severity of work-related and general stress in a large sample of active PSs (N=738), and, while accounting for psychiatric symptom severity, addresses their potential differential susceptibility to stress by comparing their responses to samples of clinicians and non-referred adults. The structure of stress-related phenomena is also examined to reveal shared variance among stress measures and their differential impact on work satisfaction and performance.

Method: This is cross-sectional, online survey of adult PSs in mental health service settings nationwide. Participants completed stress-related measures validated for use with health professionals and the general public: the Maslach Burnout Inventory-Human Services Survey, the Secondary Traumatic Stress Scale, and the Perceived Stress Scale. Confirmatory factor analyses were applied to assess internal structure and scores were summarized and compared with samples of mental health service providers and non-referred adults. Comparisons were then stratified by level of psychological distress. Participants answered questions about work absences, job satisfaction, turnover likelihood and self-rated performance. Correlations between these work outcomes and original stress measures are compared to correlations between outcomes and randomized measures created from the stress instruments. Regression analyses revealed incremental utility of specialized stress measures on outcomes above the contribution of general stress.

Results: PSs experience low to moderate levels of work-related and general stress, though not all measures were suitable for use. Secondary trauma affects less than 15% of the sample, while 51.2% report a moderate or high level of emotional exhaustion. General stress scores were normally distributed. Though effect sizes were small, compared with other mental health providers, peer support specialists experienced less secondary trauma (d = -0.15) and general stress (d = .15), and more emotional exhaustion (d = .20). However, the subset of specialists with active psychological distress experienced much greater stress than comparison groups. Work-specific stress measures strongly associate with general stress and one another. General stress accounts for 43.4% of the variance in secondary trauma, and 38.7% of emotional exhaustion. Secondary trauma explains 47.2% of the variance in emotional exhaustion. Correlations between stress measures and work outcomes were weak to moderate. When compared to original measures, randomized stress measures often showed no difference or stronger associations with outcomes. Regression analyses revealed that turnover likelihood, job satisfaction, and absenteeism were best predicted by general stress and emotional exhaustion. Self-rated performance was impacted by general stress only.

Conclusion: The findings, while supporting links between stress and psychiatric symptoms, help allay concerns about peer support specialists’ “unique” susceptibility to different forms of stress. Despite their commonalities, specific forms of stress somewhat differentially associate with adverse employment outcomes and these measures may demonstrate some utility in planning the focus of remedial efforts.