BACKGROUND: Most of the research conducted on people who harm themselves has been undertaken in secondary healthcare settings. Little is known about the frequency of self-harm in primary care patient populations. This is the first study to describe the epidemiology of self-harm presentations to primary care using broadly representative national data from across the United Kingdom (UK). METHODS: Using the Clinical Practice Research Datalink (CPRD), we calculated directly standardised rates of incidence and annual presentation during 2001-2013. Rates were compared by gender and age and across the nations of the UK, and also by degree of socioeconomic deprivation measured ecologically at general practice level. RESULTS: We found significantly elevated rates in females vs. males for incidence (rate ratio - RR, 1.45, 95 % confidence interval - CI, 1.42-1.47) and for annual presentation (RR 1.56, CI 1.54-1.58). An increasing trend over time in incidence was apparent for males (P < 0.001) but not females (P = 0.08), and both genders exhibited rising temporal trends in presentation rates (P < 0.001). We observed a decreasing gradient of risk with increasing age and markedly elevated risk for females in the youngest age group (aged 15-24 years vs. all other females: RR 3.75, CI 3.67-3.83). Increasing presentation rates over time were observed for males across all age bands (P < 0.001). We found higher rates when comparing Northern Ireland, Scotland, and Wales with England, and increasing rates of presentation over time for all four nations. We also observed higher rates with increasing levels of deprivation - most vs. least deprived male patients: RR 2.17, CI 2.10-2.25. CONCLUSIONS: Incorporating data from primary care yields a more comprehensive quantification of the health burden of self-harm. These novel findings may be useful in informing public health programmes and the targeting of high-risk groups toward the ultimate goal of lowering risk of self-harm repetition and premature death in this population.
Stored only when both papers are in the corpus — 95 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.