Background: Understanding sex differences in adverse drug reactions to drugs for psychosis could potentially guide clinicians in optimal drug choices.
Aims: By applying a text-mining approach, this study aimed to investigate the relationship between drugs for psychosis and biological sex differences in frequencies and co-occurrences of potential adverse drug events (ADEs).
Methods: Electronic patient records of a psychiatric population (1427 men and 727 women) were text mined for potential ADEs. The relative risk of experiencing specific ADEs and co-occurrence of ADEs were calculated for each sex.
Results: Findings included 55 potential ADEs with significantly different frequencies between the two sexes. Of these, 20 were more frequent in men, with relative risks of 1.10-7.64, and 35 were more frequent in women, with relative risks of 1.19-21.58. Frequent potential ADEs were psychiatric symptoms, including sexual dysfunction and disturbances in men, and gastrointestinal symptoms, suicidal and self-injurious behaviour and hyperprolactinemia-related events in women. Mention of different hyperprolactinemia-related ADEs often co-occurred in female patients but not in male patients.
Conclusion: Several known sex-related ADEs were identified, as well as some previously not reported. When considering the risk-benefit profile of drugs for psychosis, the patient's sex should be considered.
Keywords: Sex differences; adverse drug events; antipsychotic drugs; drugs for psychosis; electronic health records; text mining.