Associations among the workplace violence, burnout, depressive symptoms, suicidality, and turnover intention in training physicians: a network analysis of nationwide survey.

Depression and anxiety are the most common mental disorders among physicians, who have a greater risk of suicide than those in other professional occupations. Relationships among a demanding workload, workplace violence, burnout, and intention to turnover have also been reported. The current study examined the principal components and propagating patterns of mental health and working environment interactions in training physicians. A total of 1981 training physicians completed online self-report questionnaires during September-October (midpoint of the training year) 2020. Regularized partial correlations in a mixed graphical model (MGM) and joint probability distributions (directed acyclic graph; DAG) were estimated for four subtypes of workplace violence (verbal abuse/physical violence perpetrated by clients/hospital staff), three burnout subdomains (Maslach Burnout Inventory), thoughts about quitting, and nine depressive symptoms, including suicidality, comprising the DSM-5 diagnostic criteria for major depressive disorder (assessed using the Patient Health Questionnaire-9). Thoughts of death/self-harm showed directional dependencies on the joint probability distributions of psychomotor agitation/retardation, concentration difficulty, self-reproach, and sadness in the DAG. In the MGM, a partial correlation with psychomotor agitation/retardation (r = 0.196) accounted for 56.5% of the variance in thoughts of death/self-harm. Partial correlations with concentration difficulties (r = 0.294), self-reproach (r = 0.257), changes in appetite (r = 0.184), and worker-on-worker physical violence (r = 0.240) in the MGM accounted for 54.4% of the variance in psychomotor agitation/retardation. Thoughts about quitting were partially correlated with and dependent upon the joint probability distributions of emotional exhaustion (r = 0.222), fatigue (r = 0.142), anhedonia (r = 0.178), and sadness (r = 0.237). In contrast, worker-on-worker (r = 0.417) and client-on-physician (r = 0.167) verbal abuse had regularized partial correlations with directional dependencies on thoughts about quitting. Organization-level interventions aiming to reduce the worker-on-worker violence and individual-level approaches of clinical screening program and psychiatric counseling clinic are required. Follow-up studies to verify the effectiveness of these interventions for training physicians are needed.