Benzodiazepine Prescribing in Delirium Diagnosis: National Data on Clinical Practice.

BACKGROUND: Delirium is a common and serious neuropsychiatric condition, yet current patterns of benzodiazepine use around delirium diagnoses across various clinical settings remain poorly delineated. This retrospective cohort study aims to characterize clinicians' benzodiazepine prescribing patterns in chart-based delirium diagnosis in various clinical settings within a large national health record network, focusing on medication types, other psychotropic use, and temporal trends. METHODS: Data were obtained from the TriNetX U.S. Collaborative Network [132 million patients across 72 health care organizations], including both academic and non-academic sites with inpatient and critical-care services. Between 1,313,619 and 3,821,400 patient records with an ICD-10 clinically-recorded diagnosis of delirium were identified across four cohorts using differing delirium definitions, with inclusion irrespective of etiology, comorbidity burden, or medications. Benzodiazepine prescribing patterns before and after delirium diagnosis were examined at five-day intervals across clinical settings, focusing on lorazepam, midazolam, and diazepam, and compared with patterns of antipsychotic medications and dexmedetomidine. Benzodiazepine prescribing patterns over the 2015-2024 period were also evaluated to assess longitudinal trends. RESULTS: Benzodiazepine prescription within 5 days after delirium diagnosis ranged from 37.1%-57.8% across settings, with higher proportions in critical care compared to inpatient environments. Non-academic institutions displayed slightly higher use than academic centers, although differences were not consistent across the groups. Benzodiazepine prescription increased progressively in the days preceding the delirium diagnosis and peaked immediately afterward. Lorazepam was the most frequently used agent, followed by midazolam and diazepam. Antipsychotic use ranged from 30.1%-34.4% of which haloperidol was the most commonly used antipsychotic, followed by quetiapine, olanzapine, and risperidone. Dexmedetomidine use within 5 days after delirium ranged 18.3-21.7% in critical care settings. Over the past decade, benzodiazepine prescribing in clinically-diagnosed delirium has shown minimal variation, ranging 33.9%-41.2%. CONCLUSIONS AND RELEVANCE: Benzodiazepines are frequently prescribed around the time of delirium diagnoses across diverse inpatient settings. Overall, the data highlight a discrepancy between guidelines discouraging benzodiazepines and actual clinical practice. The data suggest the need for further research that may identify clinical approaches to enhance adherence to delirium management guidelines.