Trends and Associations of Insomnia Medications with Delirium: A Single- Center Cohort Retrospective Study

Background Clinical guidelines increasingly recommend dual orexin receptor antagonists (DORAs) over benzodiazepine receptor agonists (BZRAs) and non-benzodiazepines (Z-drugs) for insomnia treatment because of their improved safety profiles. Whether these changes in prescribing influence the incidence of delirium or antipsychotic use in hospitalized patients remain unclear, particularly given the multifactorial nature of delirium. Methods We retrospectively analyzed 79,555 inpatients admitted between 2019 and 2024, excluding those in the psychiatry and neurology wards. The monthly proportions of hypnotic use BZRAs, Z-drugs, melatonin receptor agonists (MRA), and DORAs, antipsychotic use, and psychiatrist-confirmed delirium incidence were evaluated using autoregressive integrated moving average (ARIMA) models. Results ARIMA-based time-series analyses showed significantly decreasing trends in monthly BZRAs and Z-drug use and a significantly increasing trend in DORAs use, while MRA prescriptions remained stable. The incidence of delirium did not show any significant temporal changes over the study period. Overall antipsychotic use remained stable; however, agent-specific analyses revealed distinct time-series patterns, including significant first order moving average components for quetiapine and haloperidol and a positive drift for perospirone. Conclusion A substantial shift toward DORAs occurred without a corresponding reduction in delirium or overall antipsychotic use in older adults. Further research in high-risk cohorts, such as older adults with frequent hypnotic exposure, will help clarify the potential patient-level benefits.