Taylor et al (Reference Taylor, Young and Esop2004) report on the differences in testing for diabetes among 606 patients receiving antipsychotics, observing that patients receiving atypical antipsychotics were more likely to have been tested than those receiving older agents. Moreover, this appeared to be significant specifically for clozapine, olanzapine, and antipsychotic polypharmacy.
It is noteworthy that very similar results were found by our group when examining hospitalised patients in New York State (Citrome et al, Reference Citrome, Jaffe and Levine2003, Reference Citrome, Jaffe and Levine2004). Among 1154 patients in 2000–2002 with no known prior history of receiving antidiabetic medications, those receiving clozapine, olanzapine, or more than one atypical antipsychotic had a significantly higher frequency of blood glucose testing than those receiving only typical antipsychotics (Reference Citrome, Jaffe and LevineCitrome et al, 2004). Moreover, those receiving risperidone had a frequency of testing similar to those receiving only older agents, resulting in the conclusion that there are clear differences in surveillance for diabetes mellitus among even the newer agents.
Investigators performing pharmaco-epidemiological studies examining the risk of association between antipsychotics and diabetes mellitus need to be mindful of this surveillance bias.
eLetters
No eLetters have been published for this article.