Perceived Stress and Compassion Fatigue in Intensive Care Nurses: The Serial Mediating Roles of Psychological Resilience and Burnout

ABSTRACT Background Intensive care nurses face substantial occupational stress and are at high risk of compassion fatigue. While psychological resilience is recognized as a protective personal resource and burnout as a major risk factor, little is known about how these two constructs interact, in sequence, to link perceived stress to compassion fatigue. Aims To examine the direct and indirect effects of perceived stress on compassion fatigue among nurses, while specifically investigating the serial mediating roles of psychological resilience and burnout within a structural equation modelling framework. Methods This descriptive, cross‐sectional study was conducted between March and May 2026 with intensive care nurses from training and research, public and private hospitals in Central Anatolia, Türkiye, recruited through convenience sampling. Perceived stress, burnout, psychological resilience and compassion fatigue were measured with validated scales. The hypothesized serial mediation model was tested using covariance‐based structural equation modelling with bootstrapping (5000 subsamples). Results A total of 768 intensive care nurses participated in the study. Perceived stress correlated positively with burnout and compassion fatigue and negatively with psychological resilience ( p < 0.001). The model showed excellent fit ( χ 2 /df = 1.396, CFI = 0.986, RMSEA = 0.023). Perceived stress was directly associated with compassion fatigue ( β = 0.290) and showed significant indirect associations via psychological resilience ( β = 0.059), burnout ( β = 0.196) and the serial path ( β = 0.036). Indirect effects accounted for 50.1% of the total effect, with the burnout pathway contributing 67.4% of the total indirect effect. The model explained 49.9% of the variance in compassion fatigue ( R 2 = 0.499). Conclusions Perceived stress was associated with compassion fatigue both directly and indirectly through psychological resilience and burnout, in a pattern consistent with the hypothesized serial model. Because the design was cross‐sectional, these indirect associations should not be interpreted as evidence of causal mediation. The stronger burnout pathway suggests that organizational strategies targeting burnout, alongside resilience‐enhancing programmes, may be particularly relevant for protecting nurses' professional well‐being. Relevance to Clinical Practice The findings suggest that nursing administrators should consider multidimensional strategies to address burnout and support psychological resilience. Given the relatively stronger indirect association through burnout, organizational strategies targeting modifiable workplace conditions, such as balanced workloads, adequate staffing and systematic psychological support, may warrant particular attention. At the same time, resilience‐enhancing interventions may serve as a complementary individual‐level approach. Such strategies may support nurses' professional well‐being and capacity for compassionate care.