Understanding and detecting behaviours prior to a suicide attempt: A mixed-methods study

OBJECTIVE

Prior research suggests there are observable behaviours preceding suicide attempts in public places. However, there are currently no ways to continually monitor such sites, limiting the potential to intervene. In this mixed-methods study, we examined the acceptability and feasibility of using an automated computer system to identify crisis behaviours.

METHODS

First, we conducted a large-scale acceptability survey to assess public perceptions on research using closed-circuit television and artificial intelligence for suicide prevention. Second, we identified crisis behaviours at a frequently used cliff location by manual structured analysis of closed-circuit television footage. Third, we configured a computer vision algorithm to identify crisis behaviours and evaluated its sensitivity and specificity using test footage.

RESULTS

Overall, attitudes were positive towards research using closed-circuit television and artificial intelligence for suicide prevention, including among those with lived experience. The second study revealed that there are identifiable behaviours, including repetitive pacing and an extended stay. Finally, the automated behaviour recognition algorithm was able to correctly identify 80% of acted crisis clips and correctly reject 90% of acted non-crisis clips.

CONCLUSION

The results suggest that using computer vision to detect behaviours preceding suicide is feasible and well accepted by the community and may be a feasible method of initiating human contact during a crisis.

Characteristics of Train Stations Where Railway Suicides Have Occurred and Locations Within the Stations

Background: To devise effective railway suicide countermeasures, it is necessary to identify stations where suicide is likely to occur. Aim: We explored the characteristics of stations where railway suicides have occurred and locations within the stations. Method: (Study 1) Using suicide data from between April 2014 and September 2019 provided by a major railway company in Japan, station-specific suicide was modeled as an outcome variable in a multivariate Poisson regression model. (Study 2) With railway company staff, we visited stations where suicide frequently occurs and conducted fieldwork. Results: (Study 1) Our estimation using a Poisson regression model revealed that railway suicides were more frequent when stations were serviced by passing trains, had a large number of passengers, and were located near psychiatric hospitals. (Study 2) Of 50 suicides, 48.0% occurred in front of benches or waiting rooms, 26.0% occurred at the front end of the platform, 24.0% occurred at the entrance to the platform, and 22.0% occurred at a blind spot for the train driver. Limitations: All data were provided by one railway company in Japan, limiting the generalizability of the results. Conclusion: Stations where suicide occurs frequently have distinct characteristics. Focusing on suicide hotspots may aid suicide prevention.

Interventions to reduce suicides at suicide hotspots: a systematic review

BACKGROUND

'Suicide hotspots' include tall structures (for example, bridges and cliffs), railway tracks, and isolated locations (for example, rural car parks) which offer direct means for suicide or seclusion that prevents intervention.

METHODS

We searched Medline for studies that could inform the following question: 'What interventions are available to reduce suicides at hotspots, and are they effective?'

RESULTS

There are four main approaches: (a) restricting access to means (through installation of physical barriers); (b) encouraging help-seeking (by placement of signs and telephones); (c) increasing the likelihood of intervention by a third party (through surveillance and staff training); and (d) encouraging responsible media reporting of suicide (through guidelines for journalists). There is relatively strong evidence that reducing access to means can avert suicides at hotspots without substitution effects. The evidence is weaker for the other approaches, although they show promise.

CONCLUSIONS

More well-designed intervention studies are needed to strengthen this evidence base.