Predictors of re-attempt in a cohort of suicide attempters: A survival analysis

• Published in: Journal of affective disorders Vol. 247; pp. 20 - 28
• Main Authors: Irigoyen, María, Porras-Segovia, Alejandro, Galván, Leonardo, Puigdevall, Margarita, Giner, Lucas, De Leon, Santiago, Baca-García, Enrique
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15-03-2019
• Subjects: Phsychopharmaology; Prevention; Suicide; Prevention; Phsychopharmaology; Suicide
• ISSN: 0165-0327; 1573-2517
• DOI: 10.1016/j.jad.2018.12.050

•19% of suicide attempters re-attempted suicide in the following 2 years.•Most of suicides re-attempts took place in the first 6 months.•Cluster B personality disorders were strongly associated with suicide re-attempts.•There was no evidence that psychopharmacological treatment prevented suicide.•Poor adherence to treatment increased the risk of suicide.
Suicide prevention is one of the greatest challenges in mental health policies. Since a previous suicide attempt is the main predictor of future suicidal behaviour, clinical management of suicide attempters is vital for lowering mortality. Psychopharmacological interventions are still nonspecific, and their effectiveness have often been questioned. In this study, we aim to identify predictors of suicide re-attempt in a cohort of suicide attempters, with particular focus on different aspects of psychopharmacological treatment.
This is a prospective study. Adults presenting with a suicide attempt were approached to take part in our study, resulting in a final sample of 371 participants. Participants were followed from inclusion to next suicide attempt, death by other causes, loss of the patient, or after a maximum of two years. We conducted Kaplan–Meier survival analyses and a multivariate Cox regression model for several exposure variables.
During the study period, 70 participants (18,9%) re-attempted. 60% of re-attempts occurred within the first 6 months. Three factors were independently associated with risk of re-attempt in the Multivariate Cox regression model: diagnosis of a Cluster B personality disorder, good treatment compliance, and at least one previous suicide attempt prior to the index event.
Indication bias precludes a clear interpretation of our results regarding psychopharmacological treatment. Poor adherence may also be a consequence of relapse rather than just one of its causes.
A correct psychopharmacological treatment is insufficient to prevent re-attempts in populations at risk. Strategies to increase compliance should be taken into account as part of prevention programs.