Abstract
Background
In the European Union, over 2000 suicides on railway premises were reported in 2020. Identifying individuals’ behavioral and location patterns just before they die or attempt to die by train suicide (ITS) is critical for effective prevention of suicides by train. We conducted a naturalistic study using a newly developed instrument for the assessment of fatalities in rail traffic that used information from on-site video cameras.
Methods
A total of 56 case files and surveillance recordings of ITS prior to their suicide or suicide attempt were compared to 46 surveillance recordings of matched regular train passengers (RTP) before they boarded their trains. Groups were compared on individuals’ behavior as well as location and contextual parameters.
Results
ITS performed unusual movement patterns more frequently, carried luggage less often, stayed on the platform longer, and let more trains pass relative to RTP.
Conclusions
If this study is replicated with a larger sample, artificial intelligence could be used to detect suspicious/unusual (movement) patterns in order to prevent train suicide. Social awareness campaigns that foster the identification of people in distress at train stations in combination with lower thresholds for the use of emergency devices on platforms may help to detect potential train suicides and reduce their incidence.