Warnings about an increased risk of death in patients with dementia who receive atypical antipsychotics were first issued in 2004-5, after the publication of a meta-analysis of data from placebo controlled randomised controlled trials.1 2 In 2008 analysis of databases led to similar warnings for older typical antipsychotics. At the time, inconclusive evidence suggested that the risk of death was greater for typical than for atypical antipsychotics.3 The linked paper by Huybrechts and colleagues (doi:10.1136/bmj.e977) adds to the evidence on differential risk.4 Using risperidone—the antipsychotic most widely prescribed for patients with dementia—as the comparator, they report an increased risk of mortality in nursing home residents taking haloperidol and a decreased risk in those prescribed quetiapine. Causation is not definitively proved but seems highly probable given the strength of the association that unmeasured confounders would need to have with both the use of haloperidol (or quetiapine) and mortality to account for the result.

Information on risk must be weighed against the potential benefits of …