Causal factors and configuration paths for chemical occupational accidents: grounded theory and fsQCA analysis of 154 cases in China
Abstract
Background: Chemical occupational accidents remain a critical threat to worker safety in China, but existing research has often failed to capture the complex, nonlinear, multi-factor coupling mechanisms underlying accident causation. Methods: This mixed-methods study analyzed 154 Chinese chemical occupational accident cases by integrating grounded theory and fuzzy-set qualitative comparative analysis (fsQCA). Employing the three-stage coding process of grounded theory (open, axial, and selective) to identify key causal factors, the study then used fsQCA to explore the configuration paths leading to both high- and low-severity accidents. Results and discussion: Grounded theory analysis identified 55 preliminary categories, 16 main categories, and 5 overarching core categories: safety culture, safety management system, safety capability, safety behavior, and worksite condition. These categories form the basis of the proposed Latent-Active Accident Causation Model (LA-ACM), which classifies them into latent conditions and active failures. FsQCA revealed four high-severity paths, namely safety culture-driven, safety management-dominated, safety capability-deficient, and multi-triggered, and two low-severity paths, including safety behavior-induced and worksite condition-triggered. Accordingly, a two-tier risk prevention framework was proposed: the "latent condition rectification layer" targets systemic gaps, and the "active failure interception layer" addresses on-site risks. This study underscores the theoretical implications for upstream safety governance and offers practical strategies to reduce the frequency and severity of chemical occupational accidents.
Fields of Study & Subjects
Occupational Health and Safety Research
Qualitative Comparative Analysis Research
Risk and Safety Analysis
Publication details
- Journal / source
- Frontiers in Public Health
- Publisher
- Frontiers Media SA
- Volume
- 13
- Pages
- 1713532–1713532
- Publication date
- 12 January 2026
- Language
- EN