Self-harm in people experiencing homelessness: investigation of incidence, characteristics and outcomes using data from the Multicentre Study of Self-Harm in England.
Clements C., Farooq B., Hawton K., Geulayov G., Casey D., Waters K., Ness J., Patel A., Townsend E., Appleby L., Kapur N.
BACKGROUND: People who experience homelessness are thought to be at high risk of suicide, but little is known about self-harm in this population. AIMS: To examine characteristics and outcomes in people experiencing homelessness who presented to hospital following self-harm. METHOD: Data were collected via specialist assessments and/or hospital patient records from emergency departments in Manchester, Oxford and Derby, UK. Data were collected from 1 January 2000 to 31 December 2016, with mortality follow-up via data linkage with NHS Digital to 31 December 2019. Trend tests estimated change in self-harm over time; descriptive statistics described characteristics associated with self-harm. Twelve-month repetition and long-term mortality were analysed using Cox proportional hazards models and controlled for age and gender. RESULTS: There were 4841 self-harm presentations by 3270 people identified as homeless during the study period. Presentations increased after 2010 (IRR = 1.09, 95% CI 1.04-1.14, P < 0.001). People who experienced homelessness were more often men, White, aged under 54 years, with a history of previous self-harm and contact with psychiatric services. Risk of repetition was higher than in domiciled people (HR = 2.05, 95% CI 1.94-2.17, P < 0.001), as were all-cause mortality (HR = 1.45, 95% CI 1.32-1.59. P < 0.001) and mortality due to accidental causes (HR = 2.93, 95% CI 2.41-3.57, P < 0.001). CONCLUSIONS: People who self-harm and experience homelessness have more complex needs and worse outcomes than those who are domiciled. Emergency department contact presents an opportunity to engage people experiencing homelessness with mental health, drug and alcohol, medical and housing services, as well as other sources of support.