BACKGROUND: Suicide rates are high among prisoners but little is known about the precise weight of each risk factor. METHODS: We collected data on the periods of imprisonment of all adult males incarcerated in France between 1 January, 2006 and 15 July, 2009. We used survival analyses from the incarceration to its end, censored by the observation period. We calculated suicide rates and performed a Cox model to assess the link between prisoners' imprisonment characteristics and suicide risk. RESULTS: Overall, 301,611 periods of imprisonment were observed and 353 suicides were recorded. The suicide rate was 17.9 suicides per 10,000 person-years (95% CI: 16.1-19.9). The hazard ratio (HR) of suicide risk was high for placements in a disciplinary cell (15.7, 95% CI: 10.6-23.5) and varied depending on the main offence (homicide: 7.6, 95% CI: 5.3-10.9, rape: 4.6, 95% CI: 3.2-6.6, other sexual assault: 2.9, 95% CI: 1.9-4.6, other violent offence: 2.1, 95% CI: 1.5-2.8, compared with other offences). HR was lower when visits from relatives were observed (0.4, 95% CI: 0.3-0.5) and higher if a hospitalization was observed (1.6, 95% CI: 1.3-2.0). After conviction, HR halved with respect to the remand period, but there was no difference by sentence length. HR was higher if incarceration occurred after age 30 (1.4, 95% CI: 1.1-1.7). Foreigners tended to have lower risks than French prisoners, with the exception of those incarcerated for rape. CONCLUSIONS: The suicide rate in prison is generally much higher than in the general population. This study has replicated previous international findings, highlighting the impact of the type of offence on suicide risk. Suicide prevention programmes must consider the high risk associated with incarceration for a criminal offence against a person. With regard to the impact of visits from relatives and placements in a disciplinary cell, further work should be conducted from a psychological perspective to examine the effects of physical and social isolation.
BACKGROUND: Suicide rates are high among prisoners but little is known about the precise weight of each risk factor. METHODS: We collected data on the periods of imprisonment of all adult males incarcerated in France between 1 January, 2006 and 15 July, 2009. We used survival analyses from the incarceration to its end, censored by the observation period. We calculated suicide rates and performed a Cox model to assess the link between prisoners' imprisonment characteristics and suicide risk. RESULTS: Overall, 301,611 periods of imprisonment were observed and 353 suicides were recorded. The suicide rate was 17.9 suicides per 10,000 person-years (95% CI: 16.1-19.9). The hazard ratio (HR) of suicide risk was high for placements in a disciplinary cell (15.7, 95% CI: 10.6-23.5) and varied depending on the main offence (homicide: 7.6, 95% CI: 5.3-10.9, rape: 4.6, 95% CI: 3.2-6.6, other sexual assault: 2.9, 95% CI: 1.9-4.6, other violent offence: 2.1, 95% CI: 1.5-2.8, compared with other offences). HR was lower when visits from relatives were observed (0.4, 95% CI: 0.3-0.5) and higher if a hospitalization was observed (1.6, 95% CI: 1.3-2.0). After conviction, HR halved with respect to the remand period, but there was no difference by sentence length. HR was higher if incarceration occurred after age 30 (1.4, 95% CI: 1.1-1.7). Foreigners tended to have lower risks than French prisoners, with the exception of those incarcerated for rape. CONCLUSIONS: The suicide rate in prison is generally much higher than in the general population. This study has replicated previous international findings, highlighting the impact of the type of offence on suicide risk. Suicide prevention programmes must consider the high risk associated with incarceration for a criminal offence against a person. With regard to the impact of visits from relatives and placements in a disciplinary cell, further work should be conducted from a psychological perspective to examine the effects of physical and social isolation.