Drug Prescription and Delirium in Older Inpatients: Results From the Nationwide Multicenter Italian Delirium Day 2015-2016.

Objective: This study aimed to evaluate the association between polypharmacy and delirium, the association of specific drug categories with delirium, and the differences in drug-delirium association between medical and surgical units and according to dementia diagnosis.

Methods: Data were collected during 2 waves of Delirium Day, a multicenter delirium prevalence study including patients (aged 65 years or older) admitted to acute and long-term care wards in Italy (2015-2016); in this study, only patients enrolled in acute hospital wards were selected (n = 4,133). Delirium was assessed according to score on the 4 "A's" Test. Prescriptions were classified by main drug categories; polypharmacy was defined as a prescription of drugs from 5 or more classes.

Results: Of 4,133 participants, 969 (23.4%) had delirium. The general prevalence of polypharmacy was higher in patients with delirium (67.6% vs 63.0%, P = .009) but varied according to clinical settings. After adjustment for confounders, polypharmacy was associated with delirium only in patients admitted to surgical units (OR = 2.9; 95% CI, 1.4-6.1). Insulin, antibiotics, antiepileptics, antipsychotics, and atypical antidepressants were associated with delirium, whereas statins and angiotensin receptor blockers exhibited an inverse association. A stronger association was seen between typical and atypical antipsychotics and delirium in subjects free from dementia compared to individuals with dementia (typical: OR = 4.31; 95% CI, 2.94-6.31 without dementia vs OR = 1.64; 95% CI, 1.19-2.26 with dementia; atypical: OR = 5.32; 95% CI, 3.44-8.22 without dementia vs OR = 1.74; 95% CI, 1.26-2.40 with dementia). The absence of antipsychotics among the prescribed drugs was inversely associated with delirium in the whole sample and in both of the hospital settings, but only in patients without dementia.

Conclusions: Polypharmacy is significantly associated with delirium only in surgical units, raising the issue of the relevance of medication review in different clinical settings. Specific drug classes are associated with delirium depending on the clinical setting and dementia diagnosis, suggesting the need to further explore this relationship.