An increased risk of choking associated with antipsychotic medication has been repeatedly postulated.

To examine this association in a large number of cases of choking deaths.

Cases of individuals who had died because of choking were linked with a case register recording contacts with public mental health services. The actual and expected rates of psychiatric disorder and the presence of psychotropic medication in post-mortem blood samples were compared.

The 70 people who had choked to death were over 20 times more likely to have been treated previously for schizophrenia. They were also more likely to have had a prior organic psychiatric syndrome. The risk for those receiving thioridazine or lithium was, respectively, 92 times and 30 times greater than expected. Other antipsychotic and psychotropic drugs were not over-represented.

The increased risk of death in people with schizophrenia may be a combination of inherent predispositions and the use of specific antipsychotic drugs. The increased risk of choking in those with organic psychiatric syndromes is consistent with the consequences of compromised neurological competence.

A relationship exists between mental disorder and offending behaviours but the nature and extent of the association remains in doubt.

Those convicted in the higher courts of Victoria between 1993 and 1995 had their pyschiatric history explored by case linkage to a register listing virtually all contacts with the public psychiatric services.

Prior psychiatric contact was found in 25% of offenders, but the personality disorder and substance misuse accounted for much of this relationship. Schizophrenia and affective disorders were also over-represented, particularly those with coexisting substance misuse.

The increased offending in schizophrenia and affective illness is modest and may often be mediated by coexisting substance misuse. The risk of a serious crime being committed by someone with a major mental illness is small and does not justify subjecting them, as a group, to either increased institutional containment or greater coercion.

The present study investigated histories of prior psychiatric treatment in cases of sudden death reported to the coroner.

A matching survey linked the register of deaths reported to the coroner with a comprehensive statewide psychiatric case register covering both inpatient and community-based services.

Sudden death was five times higher in people with histories of psychiatric contact. Suicide accounted for part of this excess mortality but deaths from natural causes and accidents were also elevated. Schizophrenic and affective disorders had similar suicide rates. Comorbid substance misuse doubled the risk of sudden death in affective and schizophrenic disorders.

The rates of sudden death are sufficiently elevated to raise questions about current priorities in mental health care. There is a need both for greater attention to suicide risk, most notably among young people with schizophrenia, to the early detection of cardiovascular disorders and to the vigorous management of comorbid substance misuse.