Background: Monitoring of suicide attempts is recommended as a part of the national strategies for suicide prevention across many countries. Despite this, examples of nationwide surveillance systems for suicide attempts remain rare and the practical aspects remain to be addressed. The objective was to examine whether electronic hospital records can be used to monitor suicide attempt and what to account for during data management.
Data and methods: A cohort study design was applied to nationwide register data on all individuals living in Denmark during 2000-2021 (N= 7,677,637). Suicide attempts and probable suicide attempts were identified across different versions of the electronic National Patient Register, NPR-1 to NPR-3, and the Psychiatric Central Research Register. Records were excluded for individuals who died during hospital admission, were below 8 years of age, and when admission and discharge dates of records overlapped. Incidence rates per 100,000 person-years were calculated. Distribution of events according to registration procedures and percentual changes in rates between years were assessed.
Results: After excluding respectively 34,744 (23.8%) and 86,923 (27.1%) records due to the above reasons, 146,066 suicide attempts and 320,723 probable suicide attempts remained. Over the 21- year period, incidence rates of 103.2 suicide attempts and 216.2 probable suicide attempts were found per 100,000 person-years, respectively. Most events were due to poisoning. When assessing rates of suicide attempt by single age and calendar year, rates were found to be highest among young adults, while a decreasing trend was observed in recent years (Figure 1). Changes in the rate of suicide attempts during transitions from NPR-1 to NPR2 and NPR-2 to NPR-3 were 6.1% and 5.5%, respectively, which was lower than the annual change observed between other years.
Conclusion: Hospital records were found to provide a feasible and reliable data source for monitoring trends in suicide attempts and probable suicide attempts. Rates remained stable over time despite changes in registers and revised registration procedures used over time.