Network analysis of work-family support and career identity and their associations with job burnout among primary healthcare workers: a cross-sectional study

Abstract

Objective: To explore the complex associations between job burnout, career identity, and work-family support among primary healthcare workers from a network perspective.

Methods: Data were sourced from primary healthcare institutions in China's central provinces. We used the Maslach Burnout Inventory Comprehensive Survey, a career identity scale tailored for primary healthcare workers, and the Chinese version of the Work-Family Support Questionnaire. A Gaussian network model was used to identify key factors, with "central nodes" being those that strongly influence others and "bridge nodes" connecting different parts of the network.

Results: Of the 8,135 participants surveyed, 5,120 (62.9%) reported job burnout. Compared to those with burnout, the non-burnout group scored higher in career identity, family support, and work support (54.29 vs. 49.42; 71.58 vs. 61.26; 35.03 vs. 31.20; p < 0.001). Network analysis revealed structural differences in the burnout-support-identity networks between groups after propensity score matching (M = 0.261, p < 0.001). In both groups, "understanding of role, content and requirements" were central nodes, while "work and family support" served as bridge nodes. Burnout was closely related to work support, family support, job suitability, and media criticism.

Conclusion: Targeting central and bridge nodes can help reduce job burnout among primary healthcare workers.

Keywords: career identity; job burnout; network structure; primary healthcare; propensity score matching; scientific perspective; work-family support.

Figure 1

Network structures of the burnout-support-identity network for the burnout and non-burnout groups after PSM. (A) Network structure of non-burnout individuals after PSM. (B) Network structure of burnout individuals after PSM. Positive correlations are represented by blue edges, while negative correlations are shown with red edges. The thickness of an edge indicates the magnitude of the correlation, and the circles around nodes represent their predictability.

Figure 2

Centrality plot of the burnout-support-identity network for the burnout and non-burnout groups after PSM. Centrality plot depicting the strength and expected influence of each symptom in the burnout and non-burnout network (z-score) after PSM. The dash line represents the non-burnout individuals and solid line represent the burnout individuals.

Figure 3

Bridge strength plot of the burnout-support-identity network for the burnout and non-burnout groups after PSM. Left panel: Index of bridge strength for each node of non-burnout group after PSM; Right panel: Index of bridge strength for each node of burnout group after PSM.

Figure 4

Flow network of burnout-support-identity in 2,986 pairs of propensity score matched participants. Blue edges represent positive partial correlations, and red edges represent negative partial correlations. Thicker edges represent stronger correlations.