TY - JOUR T1 - 564 Surveillance of suicide using administrative health databases: a study in Québec and in Belgium JF - Injury Prevention JO - Inj Prev SP - A203 LP - A203 DO - 10.1136/injuryprev-2016-042156.564 VL - 22 IS - Suppl 2 AU - Christelle Senterre AU - Alain Levêque AU - Eric Pelletier AU - Louis Rochette AU - Danielle St-Laurent AU - Elham Rahme Y1 - 2016/09/01 UR - http://injuryprevention.bmj.com/content/22/Suppl_2/A203.1.abstract N2 - Background It is well known that a suicide attempt (SA) is one of the main risk factors for a death from suicide and that some events, as occurrence of injuries can be interpreted as predictive factors of future suicidal behaviours. Surveillance of these types of events would permit to better organise the prevention. The use of health administrative databases may be an interesting source of information that allows a longitudinal surveillance of these events. The aim of this research is to study the occurrence of traumatic (self-inflicted) injuries and (mental) health events over time in the Province of Quebec, Canada and in Belgium.Methods A retrospective cohort study was first conducted in Quebec and is will be reproduced in Belgium. In Québec, the 2006–2014 data from the SISMACQ (Système intégré de surveillance des maladies chroniques du Québec) database was used; and in Belgium, the RCM (Résumé Clinique Minimum) database will be use. The cohort included all individuals with SA in 2006–2014 identified from hospital records in SISMACQ using the international classification of disease 10th edition for self-inflicted injury. For each individual, the first SA during a given year was retained (index event). SAs during the year preceding the index event were assessed. Recurrent SAs on a prospective period of 3 years were also assessed. In addition, health care consumption and mortality (all causes and related to suicide) were examined during the three-year period post the index event.Results The first descriptive analysis of Québec data have identified 11,022 suicide attempts related to 9,755 individuals, meaning that 846 people (8.7%) had more than one hospitalisation for SA, with 2.3% having 3 or more SA during the follow-up period. In 65.6% of cases, the time between 2 SA was equal or lower than 1 year, while for 17.3% it was between 1 and 2 years; and for the last 17.1% it was 3 years or more. Regarding the Belgian data, no result is currently available due to long administrative procedures to obtain them.Conclusions The use of health administrative databases for the study and surveillance of suicidal behaviours, and of their associated events, allows a better understanding of the suicidal continuum (before – after) and therefore may inform organisation of the universal, selective, and indicated prevention. ER -