Application of systems thinking accident analysis methods: A review for railways

•This article presents a comprehensive literature review and bibliometrics of systems thinking accident analysis in railways.•HFACS, AcciMap, STAMP, and FRAM methods are compared regarding systems theory criteria and application characteristics.•The review has shown the popularity of HFACS and STAMP in railway accident studies, with recent attention to FRAM.•The application of systemic methods in railway accident modelling has usually been associated with qualitative and quantitative improvements. Accident analysis methods are used to identify contributory factors and the entire cause-effect process leading to an accident. The safety science literature has emphasized the suitability and popularity of systems thinking accident analysis methods, particularly Human Factors Analysis and Classification System (HFACS), AcciMap, Systems Theoretic Accident Model and Processes (STAMP), and Functional Resonance Analysis Method (FRAM), for analyzing accidents in sociotechnical systems. However, a comprehensive literature review and bibliometric analysis of systemic accident modelling in railways are missing. This paper aims to fill this gap and analyze the state of the art of systems thinking accident analysis with application to the railway from 2000 to November 2022. Moreover, the systemic methods are compared regarding the systems theory and practical characteristics. By employing a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) approach, 32 relevant studies were identified from Scopus and Web of Science databases and a backward search. The review highlights the popularity of HFACS and STAMP in railway accident studies, with recent attention on FRAM. The research direction in applying systemic accident modelling for railways focused on analyzing railway accidents along with qualitative and quantitative improvements of the methods. The qualitative improvements were mainly related to making the techniques more compatible with the systems theory criteria, and the qualitative advancements comprised an estimation of relationship strengths. We gleaned additional findings and several future research directions, including suggested developments and further applications of the systems approach in rail accident analysis.