Risk factors for suicide attempts in community alcohol treatment

BackgroundAlcohol Use Disorder is associated with suicide and suicide attempts, and addiction treatment services have a role in suicide prevention. We aimed to identify risk factors for suicide attempt among a cohort of community-based alcohol treatment service users. MethodsLinked data from 4415 adults accessing secondary addiction services for alcohol treatment between 2006 and 2019 in London, UK, were used to identify risk factors for suicide attempt. Cox proportional hazards regression estimated the relative increase or decrease in hazard associated with each risk factor on a composite outcome event; death by suicide or contact with emergency psychiatric care within one year of starting treatment. FindingsThere were 468 (10.5%) crisis care contact events, and <10 suicide deaths. After adjustment, factors associated with increased hazard of crisis care contact or death by suicide were history of suicide attempt (HR 1.83[1.43-2.33]), poor mental health (HR 1.81[1.41-2.32]), current suicidal ideation (HR 1.65[1.18-2.31]), use of drugs other than cocaine, cannabis and opiates (HR 1.41[1.02-1.95]), female sex (HR 1.34[1.10-1.65]) and social isolation (HR 1.24[1.02 - 1.51]). Factors associated with reduced hazard of crisis care contact or death by suicide were alcohol abstinence (HR 0.51[0.31-0.83], ref>30 units), drinking 1-15 units (HR 0.64[0.49-0.85], ref>30 units), Black ethnicity (HR 0.61[0.45-0.83]) and living with children (HR 0.74[0.56-0.99]). InterpretationThe identified risk factors for suicide attempt can help risk formulation and safety planning among patients accessing alcohol treatment services. FundingNational Institute for Health Research Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, Kings College London. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSAlcohol Use Disorder (AUD) is a risk factor for suicide and suicide attempt, and among people with AUD those accessing addiction treatment services are particularly at risk. Effective suicide risk formulation and safety planning requires an understanding of the demographic, clinical and circumstantial factors that are associated with increased or decreased risk of suicide attempt in the population of people accessing treatment for their alcohol use. We searched PubMed using search terms ((("alcohol use disorder") OR ("alcohol depend*") OR ("substance use disorder") AND (("treat*") OR ("service*"))) AND (("suicide attempt") OR ("suicid*"))) as well as Google Scholar and cited reference searching in Web of Science, to identify previous studies of risk factors for suicidal behaviour in cohorts engaged with some form of Substance Use Disorder (SUD) treatment which included alcohol at least to a minimal degree, and which measured a suicide-related outcome after treatment commencement. The impact of the risk factors identified in these studies varied greatly, reflecting heterogeneity in the substance use profiles and settings of the samples used. We found no longitudinal studies which consider risk factors for suicidal behaviour in a purely alcohol-using sample accessing community-based addiction care. The single consistent risk factor for suicidal behaviour across these studies was a previous history of suicide attempt. Added value of this studyOur study uses 14 years worth of structured data from service users accessing Community Drug and Alcohol Team (CDAT) treatment primarily for their alcohol use. A range of risk factors for suicide attempt (measured via contact with crisis care services) or death by suicide in the year following treatment start were identified: predisposing factors included a history of suicide attempt, female sex and White ethnicity; modifiable factors included social isolation, poor mental health, current suicidal ideation or carer concern, and use of drugs other than cocaine, cannabis and opiates; protective factors included abstinence from or relatively low use of alcohol, and children living with the service user. This is the first prospective analysis of risk factors for suicidal behaviour in a purely alcohol-using sample accessing community-based addiction care. This population represent the largest proportion of CDAT service use, with a uniquely elevated suicide risk. Implications of all the available evidenceA wide range of risk factors for suicide and suicide attempt can be identified among people accessing alcohol treatment, providing population-specific contextual knowledge that can aid patient-centred suicide assessment and safety planning, and a potential framework within which potential avenues for intervention can be identified.