Drug–Drug Interactions in Outpatient Psychiatry: From Interaction Profiles to Smart Monitoring ^†

Objective. To characterize drug–drug interactions (DDIs) within outpatient psychiatric prescriptions and identify high-risk combinations, optimize therapy, and minimize adverse effects.

Methods. We retrospectively reviewed 662 psychiatric prescriptions (October–December 2023), recording patient sex, age, diagnoses, and medication. Inclusion criteria required ≥2 drugs per drug regimen and informed consent signed by the patient. DDIs were identified via DrugBank 6.0 [1] and then classified as major, moderate, or minor, and no interaction was found.

Results and Discussion. Of the 662 prescriptions, 458 were for female patients, and 204 were for male patients. Regarding the age category, the most represented is the age category over 65 years (n = 311 prescriptions), more susceptible to the development of adverse effects following drug interactions due to comorbidities, polypharmacy, and physiological changes. Of the 662 prescriptions, 327 contained a number greater than or equal to 5 drugs (polypharmacy) [2,3]. Of the total pairs analyzed, 4.84% are major DDIs, 32.18% are moderate DDIs, 19.08% are minor DDIs, and the rest show no interactions. The most prescribed benzodiazepine (BZD) was alprazolam (n = 146 prescriptions), followed by lorazepam (n = 113), zolpidem (n = 77), and zopiclone (n = 69). Among thymic stabilizers, valproic acid was the most used (n = 129 prescriptions), followed by carbamazepine (n = 54). In the case of antipsychotics, quetiapine is the most frequent (n = 120), followed by risperidone (n = 59). The most prescribed antidepressant is trazodone (n = 98 prescriptions), followed by escitalopram (n = 67), tianeptine (n = 63), and mirtazapine (n = 61). The benzodiazepine involved in the most major DDIs was zopiclone (n = 22), followed by lorazepam (n = 8) and alprazolam (n = 7). In our cohort, zopiclone was involved in the most major DDIs (n = 22), mostly with carbamazepine (n = 9) and clopidogrel (n = 6), reflecting CYP450-mediated pharmacokinetic interactions [4,5,6]. Less-frequent major DDIs occur with clarithromycin (n = 2), salmeterol (n = 2), amiodarone (n = 1), azelastine (n = 1), and mometasone furoate (n = 1), underscoring the need to review any concurrent anti-infectives, antiarrhythmics, and even intranasal therapies. By contrast, moderate interactions, driven by additive CNS depression or mild metabolic shifts, were common with sertraline (n = 25), trazodone (n = 22), escitalopram (n = 18), quetiapine (n = 15), and valproic acid (n = 14), underscoring frequent zopiclone co-prescription alongside SSRIs and mood stabilizers and the consequent risk of excessive sedation and cognitive impairment [4,6,7]. To mitigate the overlap of sedative effects when administering SSRIs together with BZD and BZD-related drugs, we piloted a closed-loop Internet of Medical Things (IoMT) system using Empatica’s Embrace Plus wearable [8]. By tracking accelerometry, heart rate, SpO₂, skin temperature, and galvanic skin response—combined with pharmacokinetic modeling—we aim to identify the optimal moment to administer the personalized SSRI dosing.

Conclusions. Major DDIs, primarily metabolic interactions between zopiclone and CYP-modulating agents, require dose adjustments or alternative therapies/monitoring. The high prevalence of moderate CNS–depressant interactions, especially between BZD and SSRIs/mood stabilizers, calls for vigilant monitoring, patient education on sedation risk, and scheduling strategies [8,9]. By integrating wearable physiology monitoring with simple pharmacokinetic modeling, we can conduct a data-driven analysis of the optimal moment for personalized SSRIs administration with BZD.