Background Deliberate self-harm (DSH) is one of the strongest risk factors for suicide and a significant public health problem in its own right. Nationally representative data on the incidence, repetition and determinants of DSH are sparse.

Objective To establish the incidence and risk of repetition of hospital-treated DSH in Ireland.

Setting The Republic of Ireland.

Design Surveillance of presentations to hospital emergency departments as part of the Irish National Registry of Deliberate Self Harm.

Participants Individuals who presented with DSH to one of the forty hospital emergency departments that operated in Ireland in 2003–2008. Case ascertainment and data collection were performed by trained data registration officers working independently of the hospitals and following standard operating procedures.

Main outcome measures Annual age-specific and age-adjusted rates of persons presenting to hospital per 100 000 population. Kaplan-Meier cumulative incidence curves and corresponding life-tables showing the risk of repeat DSH presentation at various time intervals and hazard ratios arising from multivariate Cox regression.

Results For 2003–2008, 63 154 DSH presentations were recorded involving 41 205 individuals. The average annual total, male and female rate of persons presenting with DSH were 202, 172 and 234 per 100 000, respectively. Overall, the female rate was 36% higher than the male rate. There was a clear peak in the female rate in 15–19 year-olds (651/100 000), twice the equivalent rate in men (323/100 000). In men, the highest rate was in the 20–24 year age group (426/100 000). For both genders, rates fell with increasing age. Of the 41 205 DSH patients, 8755 (21.2%) presented on at least two occasions. The risk of repetition was highest soon after a DSH presentation. Half (50%) of all repeat events occurred within three months. Risk of repetition within 12 months was almost 40% in subjects who self-cut compared to approximately 25% for those using other methods of DSH. Repetition rates were similar in men and women. Repetition rates rose rapidly with the number of previous DSH presentations. Respectively, twelve-month repetition rates of 13.7%, 36.1%, 47.7%, 59.7%, and 70.9% were observed for those with one, two, three, four and five previous presentations.

Conclusion Population-based data on the incidence and repetition of hospital-treated DSH represent an important indicator of the burden of mental illness and the pattern and distribution of suicide risk in the community. The development and implementation of effective methods to reduce repetition rates is an important challenge for health systems.