Potentially fatal incidents: identification, classification and human factor analysis

Potentially fatal incidents (PFIs) are increasingly used as leading indicators in high-risk industries, yet their definitions, classification criteria, and investigative depth vary widely across organisations, limiting. their preventive value and comparability. Human factors (HF) play a critical role in determining whether incidents escalate into PFIs and must be considered together with technical and organisational barrier performance. This research aims to examine the role of human and organisational factors in PFI identification, analyse misclassification patterns, and propose a human-factors-based model to improve PFI classification consistency and learning value. A retrospective document analysis was conducted using incident reports from a heavy-industry organisation covering the period from 2020 to 2024. The dataset was systematically reviewed and PFI classifications were re-evaluated using a structured framework integrating hazardous energy and exposure assessment, barrier performance evaluation based on Bow-Tie logic, and human and organisational factor coding using an HFACS-based structure. Analysis revealed inconsistency in PFI classification, including overclassification and under-classification linked to limited recognition of human and organisational factors. Number of incidents were labelled as PFIs despite lacking credible fatal energy exposure, while other events with systemic and human-factor contributors associated with fatal risk were not recognised as PFIs. The HF-PFI Model demonstrated improved classification reliability by integrating energy exposure, barrier status, human factor categories, and systemic indicators. Integrating human-factors analysis into PFI identification can strengthen serious injury and fatality prevention in high-risk industrial environments.