Predictors of Suicidal Behavior in Depression

Background and rationale:

Suicidal behavior (SB) is a major health problem in France, with more than 10,000 suicides (6th UE28) and 220 000 suicide attempts (SA) per year. A large percentage of men (6%) and women (9%) in France have made at least one SA. Moreover, suicide is the second leading cause of death among the youngest (15-44 years). The clinical model currently admitted for understanding SB is a stress vulnerability model. But to date, scientists have not (yet) a clinical application for the research on SB. The management of psychiatric patients, including depressed subjects, faces the impossibility of detecting those at high risk of occurrence of SB. Indeed, although many risk factors for SB are known, their predictive value is very low. In addition, these risk factors are primarily derived from cross-sectional studies. The identification of predictors, using a standardized evaluation would allow clinicians to identify the subject at high risk of SB for which therapeutic interventions and the organization of care should be particularly aggressive and intensive.

Objectives:

Main objective: To identify predictors (clinical, biological, environmental) of the occurrence of SB during 12 months in a population of depressed suicide attempters and lifetime non attempters to optimize a short standardized evaluation applicable to emergency settings.

Secondary objectives: 1) To build a multidimensional score for predicting the risk of recurrence of suicide attempt among suicide attempters ; 2) To evaluate the prospective performance of this score to predict emergence of suicide attempt among depressed patients without history of suicide attempt; 3) To identify predictors of suicidal spectrum (suicidal thoughts, suicidal planification, suicide attempt and completed suicide) in depressed subjects ; 4) To evaluate the feasibility of a multicenter network that will use standardized assessments to evaluate suicidal risk through a secured website.

Methods:

2000 patients with major depressive episode will be recruited in the emergency departments and post-acute care units of 14 French hospitals, and then will be followed for 12 months. For every depressed patient admitted within 48 hours of a suicide attempt (case), a depressed patient without lifetime personal history of SB (psychiatric control) will be included. Cases will be mainly recruited in emergency settings; psychiatric controls will be mainly recruited from the inpatient units (hospital stay <7 days) and outpatient units to limit selection bias.

At inclusion, at 3, 6 and 12 months, clinical and biological assessments will ne realized.

Statistical analyses:

1. Kaplan-Meier and Cox models adjusted for potential confounding factors, 2) Univariate and multivariate analysis, 3) Establishment of a predictive score from multivariate analyses.

Short-term perspectives:

1. Optimization of standardized assessment of suicidal risk in depressed patients applicable in daily practice, and 2) Development of a network of clinical centers involved in suicidology field to promote specific care and research.

Long-term perspectives:

1. Access to a computerized assessment tool via the Internet to improve the management of depressed patients all over France 2) Understanding of the psychological and biological factors underlying SB in the context of mood disorders, and 3) Development of preventive and therapeutic approaches.