Suicide as an incident of severe patient harm: a retrospective cohort study of investigations after suicide in Swedish healthcare in a 13-year perspective

Objectives To explore how mandatory reporting to the supervisory authority of suicides among recipients of healthcare services has influenced associated investigations conducted by the healthcare services, the lessons obtained and whether any suicide-prevention-related improvements in terms of patient safety had followed. Design and settings Retrospective study of reports from Swedish primary and secondary healthcare to the supervisory authority after suicide. Participants Cohort 1: the cases reported to the supervisory authority in 2006, from the time the reporting of suicides became mandatory, to 2007 (n=279). Cohort 2: the cases reported in 2015, a period of well-established reporting (n=436). Cohort 3: the cases reported from September 2017, which was the time the law regarding reporting was removed, to November 2019 (n=316). Primary and secondary outcome measures Demographic data and received treatment in the months preceding suicide were registered. Reported deficiencies in healthcare and actions were categorised by using a coding scheme, analysed per individual and aggregated per cohort. Separate notes were made when a deficiency or action was related to a healthcare-service routine. Results The investigations largely adopted a microsystem perspective, focusing on final patient contact, throughout the overall study period. Updating existing or developing new routines as well as educational actions were increasingly proposed over time, while sharing conclusions across departments rarely was recommended. Conclusions The mandatory reporting of suicides as potential cases of patient harm was shown to be restricted to information transfer between healthcare providers and the supervisory authority, rather than fostering participative improvement of patient safety for suicidal patients. The similarity in outcomes across the cohorts, regardless of changes in legislation, suggests that the investigations were adapted to suit the structure of the authority’s reports rather than the specific incident type, and that no new service improvements or lessons are being identified.