I welcome any contribution that might improve the shocking mortality disadvantage associated with severe mental illness. Yeomans et al ^{Reference Yeomans, Dale and Beedle1} should, however, be aware that this issue was identified long before the 1990s. William Farr reported in 1841 that the mortality of lunatics in England and Wales varied between three and fourteen times from that of the general population. ^{Reference Farr2} He concluded that some of the excess deaths ‘may be fairly ascribed to insanity. The excess above this must be attributed to the diseases generated by the limited space in which the unhappy lunatics are confined – to the collection of large numbers under the same roof – the impurity of atmosphere – the want of exercise and warmth – the poor unvaried diet – and the deficiency of medical attendance.’ Plusça change.

While it is positive that the excess natural mortality of mental illness is attracting greater attention, we should be wary of blindly jumping on the cardiovascular screening bandwagon. There is no evidence that health checks improve the mortality of the general population. ^{Reference Gotzsche, Jorgensen and Krogsboll3} However, people with severe mental illness are a high-risk group for cardiovascular disease and thus their risk/benefit balance may be different. Still, screening takes resources that could be used for other interventions and has harmful consequences for significant numbers of patients.

Rather than setting up huge cardiovascular screening programmes, we might do better to put far more effort into smoking cessation work and to review our prescribing practice and avoid those drugs with particularly bad metabolic profiles. At the very least, we need far better audit of these programmes – audit that measures whether they improve mortality.