Reciprocal Effects of Psychological Distress and Burnout Among Child Welfare Workers

Background and purpose

Burnout among child welfare caseworkers is a persistent problem that undermines service delivery and contributes to turnover. Prior studies have also documented an association between caseworkers’ mental health and burnout, but the direction of this relationship remains unclear because much of the existing evidence is based on cross-sectional data. To address this limitation, our study examined the bidirectional relationship between psychological distress and burnout using a random-intercept cross-lagged panel model. Our study addresses two research questions: 1) Do current levels of burnout predict subsequent levels of psychological distress over time? 2) Do current levels of psychological distress predict subsequent of burnout over time?

Methods

Data were drawn from The Florida Study of Professionals for Safe Families (FSPSF), a longitudinal study of child welfare workforce retention. Caseworkers entering the child welfare workforce between September 2015 and December 2016 were surveyed every six months for 4 years. The analytic sample included 694 caseworkers. Psychological distress was measured with the Kessler Psychological Distress Scale (K6), and burnout was measured with the combined work- and client-related subscales of the Copenhagen Burnout Inventory. Age, gender, race/ethnicity, and social work degree were included as time-invariant covariates. A random-intercept cross-lagged panel model—which tests for associations between variables over time while controlling for present levels of each variable—was estimated in R using lavaan, MLR, and FIML. A series of nested models were run to assess whether the scales used to measure burnout and psychological distress were reliable over time, and whether the paths (i.e., relationships) between variables should be free to vary or time-invariant (i.e., constrained) over time. The models were compared using Satorra–Bentler scaled difference tests and multiple fit statistics (CFI, TLI, RMSEA, SRMR).

Results

The scaled difference tests and fit statistics indicated our measurements of burnout and psychological distress performed reliably across all waves. These tests also indicated that a time-invariant model was the most parsimonious model that fit the data best (CFI = .976, TLI = .974, RMSEA = .026, SRMR = .040), which indicates that the relationship between burnout and psychological stress was stable over time. Both cross-lagged paths were statistically significant and positive: greater psychological distress predicted higher subsequent levels of burnout (b = 0.12, p = .004) over time, and higher levels of burnout predicted greater subsequent psychological distress (b = 0.08, p = .005) over time. Overall, these findings suggest a bidirectional reinforcing cycle between burnout and psychological distress among child welfare caseworkers.

Conclusions and Implications

Breaking the distress–burnout cycle is critical not only for workforce well-being but also for child safety, given its implications for turnover and service quality. These findings suggest that interventions targeting distress or burnout alone may be insufficient. Child welfare agencies should adopt comprehensive and integrated approaches that combine supports for mental well-being (e.g., self-care plans and access to mental health services) with strategies that prevent and address burnout (e.g., manageable workloads and accessible and high quality supervision)