TRENDS IN DELIBERATE SELF POISONING AND SELF INJURY IN OXFORD, 1976-90

Objective - To review trends in deliberate self poisoning and self injury (attempted suicide) over 15 years (1976-90) on the basis of general hospital referrals. Design - Prospective data collection by computerised monitoring system. Setting - Teaching general hospital. Subjects - All patients aged 15 and over (n = 9605) referred to the hospital after episodes (n = 13 340) of deliberate self poisoning or self injury. Main outcome measures - Rates based on population of Oxford city; changes in substances used for self poisoning; history and repetition of attempts; and rates of admission to the hospital and of referral to the psychiatric service. Results - Attempted suicide rates for women declined during the late 1970s and early 1980s but increased again during the late 1980s. Those for men remained relatively steady throughout the period. Highest mean annual rates occurred in women aged 15-19 (711/100 000) and in 20-34 year old men (334/100 000). The proportion of overdoses with paracetamol increased from 14.3% (125/873) in 1976 to 42% (365/869) in 1990 (chi-2 for trend = 481, p < 0.01). Throughout the period the proportions of referred patients admitted to hospital and of those attempting suicide for the first time (over two thirds) did not decrease. Annual rates of repetition of attempts by women declined from 15.1% (257/1700) in 1976-8 to 11.9% (161/1356) in 1987-9 (chi-2 for trend = 7.8, p < 0.01). Lower repetition rates occurred in women admitted to hospital and referred to the psychiatric service (431/4585, 9.4%) than in those not referred (42/235, 17.9%; chi-2 = 17.2, p < 0.0001). Conclusions - Rates of attempted suicide declined in the 1970s and early 1980s, in women, but there are probably at least 100 000 hospital referrals a year in England and Wales because of this problem. Prevention of paracetamol self poisoning requires urgent attention, and psychosocial assessment should be conducted with as many of those who attempt suicide as possible.