In reply
In response to our article “Antipsychotic Use and Myocardial Infarction in Older Patients With Treated Dementia,”1 van Hout et al emphasized that patients with dementia treated with antipsychotics (APs) were more likely to receive antidepressants and anxiolytics and hence present with behavioral problems that might be related to myocardial infarction. As we have discussed, this is expected because APs are usually prescribed to patients with more severe dementia, in whom antidepressants and anxiolytics are likely to be prescribed.2 Analyses were adjusted for drugs potentially associated with advanced stages of dementia. Van Hout et al argued that it would have been important to compare classes of APs. Although, in the literature, the difference in the risk of stroke across AP classes is not demonstrated,3 we agree and underlined this limitation in the discussion. The limitation of using drug dispensing as a proxy for drug use is common to all studies based on claims data. We agree that confirmatory field studies would bring additional evidence to this first demonstration.