Large highlights two issues in relation to suicide prevention: (a) the differing terminology used internationally in relation to models of suicide prevention; and (b) the difficulties inherent in assessing suicide risk following discharge from psychiatric hospital.

Whereas in the UK the terms high-risk (or targeted) approach and population (or mass) approach are used commonly, ^{Reference Rose1} terminology in the USA and elsewhere differs, referring to universal, selective and indicated interventions. ^{Reference Knox, Conwell and Caine2} A universal intervention corresponds to the population approach, in that it is applied to a broad population irrespective of the risk of individual members, in order to change norms and values, to influence unidentified members of the population who may carry more risk and, ultimately, to shift the risk of the entire population. At the other end of the spectrum, an indicated intervention corresponds to a high-risk approach, in that it is applied to identified symptomatic individuals. It is much the same as a clinical intervention except that public health approaches proactively reach into communities and diverse settings to engage such persons, whether or not they present in clinical settings.

Selective interventions equate to a form of high-risk approach, but one which addresses groups with a significantly higher-than-average risk of developing mental disorders or adverse outcomes. ^{Reference Knox, Conwell and Caine2} Such groups are described in the 2012 suicide prevention strategy for England as those ‘with particular vulnerabilities or problems with access to services’ (p. 21). ³ The groups listed include children and young people; people with a history of childhood abuse; minority ethnic groups and asylum seekers; and people with untreated depression. These are distinguished from groups regarded as high risk for completed suicide on the basis of clear epidemiological evidence, which in the English strategy include people under the care of mental health services; people with a history of self-harm; people in contact with the criminal justice system; adult men under 50; and specific occupational groups. Whereas effectiveness studies tend to concentrate on proximal interventions for these highest-risk groups, less evidence describes the effectiveness of selective interventions, but this situation is likely to evolve.

In relation to the second issue that Large raises, also highlighted in his recent letter to The Psychiatrist, ^{Reference Large, Ryan and Callaghan4} it would be fair to say that anyone admitted to hospital for a major mental disorder, or a substance use disorder, has a greater degree of risk for suicide than non-hospitalised individuals with mental disorders or the general population. However, people in contact with mental health services in the year prior to death account for 27% of general population suicides in England. ^{Reference Appleby, Kapur, Shaw, Hunt, Flynn and While5} Gunnell et al's study ^{Reference Gunnell, Metcalfe, While, Hawton, Ho and Appleby6} found that 10% of all suicides in England occurred within the year following psychiatric discharge. Applying the term ‘high risk’ to this group of patients describes their overall risk in relation to the general population, ignoring the wide degree of variation in risk between individuals within this group. One could argue that integrated aftercare constitutes high-quality care for all but, on the basis of the above taxonomies, we would not regard this as universal because it is indicated for all such discharged patients.