Managing psychosocial hazards in the workplace: how to link frequency and severity using risk matrices

Evaluating and prioritizing psychosocial hazards remains a challenge for both scientists and practitioners. In contrast to traditional chemical or physical risk assessment methods, psychosocial assessments generally do not record health‑related outcomes for prioritization or severity evaluation, but infer associations probabilistically at the population level. The risk-matrix approach (RMA) addresses these limitations by structuring prioritization around exposure levels and the magnitude of associated health-related impairment. However, the RMA has not yet been applied systematically to psychosocial risk assessment.

We developed a risk-matrix-based prioritization framework using the Copenhagen Psychosocial Questionnaire (COPSOQ). Psychosocial hazards were represented on the COPSOQ's exposure scale and compared across several exposure levels from low to very high. Harm was captured with three health-related outcomes ranging from proximal cognitive stress responses to more distal indicators of burnout and general health. Using data from a German steel company (N = 7,242), we estimated the magnitude of outcome differences across exposure levels for each hazard while accounting for key demographic and job-related characteristics. The risk matrix then translates these associations into expected outcome differences across exposure levels, alongside an indication of uncertainty.

Across outcomes, higher hazard exposure was associated with greater cognitive stress symptoms, higher burnout, and poorer general health. Associations were strongest for hazards such as detrimental environmental conditions, work-privacy conflict, and emotional demands, yielding substantial model-implied outcome differences at moderate to high exposure levels.

For subsequent risk-mitigation interventions, psychosocial hazards can be prioritised based on their model-implied health impact across exposure levels rather than single cut-offs. This work advances the RMA framework by integrating health-related outcomes directly into psychosocial risk prioritisation in applied settings and underscores.