Exploratory prescribing risk profiles and drug drug interaction burden in older adults with depressive episodes

Abstract

Introduction Older adults with depressive episodes are frequently treated in the context of psychiatric multimorbidity, somatic co-treatment, and polypharmacy, increasing their vulnerability to drug-drug interactions (DDIs) and cumulative medication-related risk. This study characterized DDI burden and exploratory prescribing risk profiles among psychogeriatric outpatients, focusing on patients with depressive episodes. Methods We conducted a retrospective observational study of psychiatric outpatient visits in 2023. The primary subgroup consisted of those diagnosed with a depressive episode. Patients with at least one diagnosis of a depressive episode formed the primary depression-focused subgroup. Potential DDIs were assessed using DrugBank and classified as major, moderate, or minor. These potential DDIs were database-identified and were not clinically confirmed adverse drug events. Visit-and patient-level analyses described potential DDI burden, medication risk burden, psychiatric co-diagnoses, and diagnosis-specific potential DDI pairs. Multivariable logistic and negative binomial regression models were fitted among patients with depressive episodes. Patient-level exploratory prescribing risk profiles were derived in the full cohort using DDI and medication risk burden features. Results The cohort included 309 patients and 1242 eligible visits; 242 patients had at least one depressive episode diagnosis. Anxiety disorders were the predominant co-diagnosis among patients with depressive episodes, being recorded in 193 patients (79.8%) and co-occurring with depressive episode at the same visit in 177 patients (73.1%). Nearly half (50.4%) of visits from depressive-episode patients included at least one database-identified major potential DDI, while 93.7% had at least one moderate potential DDI. Leading moderate DDI pairs mainly involved benzodiazepine/Z-drug and benzodiazepine/antidepressant combinations. Leading major DDI pairs included cardiovascular, carbamazepine-related, and trazodone-related interactions. Medication load was the variable most consistently associated with potential DDI burden, while diagnostic complexity contributed to cumulative annual major DDI burden. A higher-burden exploratory prescribing risk profile identified 86 patients with depressive episodes (35.5%), highlighting a clinically important subgroup characterized by greater polypharmacy, potential DDI burden, and geriatric medication risk scores. Conclusion Database-identified potential DDI burden was substantial among older adults with depressive episodes and was driven mainly by medication load. Anxiety overlap, recurrent sedative-related moderate potential DDIs, and higher-burden prescribing risk profiles highlight the need for structured, regimen-level medication review.