Evaluating the capability of ChatGPT in predicting drug-drug interactions: Real-world evidence using hospitalized patient data.

Drug-drug interactions (DDIs) present a significant health burden, compounded by clinician time constraints and poor patient health literacy. We assessed the ability of ChatGPT (generative artificial intelligence-based large language model) to predict DDIs in a real-world setting. Demographics, diagnoses and prescribed medicines for 120 hospitalized patients were input through three standardized prompts to ChatGPT version 3.5 and compared against pharmacist DDI evaluation to estimate diagnostic accuracy. Area under receiver operating characteristic and inter-rater reliability (Cohen's and Fleiss' kappa coefficients) were calculated. ChatGPT's responses differed based on prompt wording style, with higher sensitivity for prompts mentioning 'drug interaction'. Confusion matrices displayed low true positive and high true negative rates, and there was minimal agreement between ChatGPT and pharmacists (Cohen's kappa values 0.077-0.143). Low sensitivity values suggest a lack of success in identifying DDIs by ChatGPT, and further development is required before it can reliably assess potential DDIs in real-world scenarios.

On- and off-label utilisation of antipsychotics in Australia (2000-2021): Retrospective analysis of two medication datasets.

OBJECTIVE: To determine antipsychotic utilisation patterns in Australian adults from 2005 to 2021, with a focus on on-label and off-label prescriptions. METHODS: We examined antipsychotic dispensing trends in adults from 2005 to 2021 using a 10% sample of the Pharmaceutical Benefits Scheme (PBS) dataset, which contains patient-level information on medicines dispensed throughout Australia. The lack of diagnostic information in PBS was substituted by analysing BEACH (Bettering the Evaluation And Care of Health) dataset, a cross-sectional national survey from 2000 to 2016, consisting of data from general practitioner-patient encounters. RESULTS: There were 5.6 million dispensings for 164,993 patients in PBS throughout this period; 69% patients had >1 dispensing, with a median of 6 per patient. Calculating the estimated period of exposure gave a total of 693,562 treatment episodes, with a median duration of 80 days. There were steady increases in both the incidence and prevalence of antipsychotic dispensings, mainly due to oral second-generation antipsychotics. The most commonly prescribed antipsychotics were quetiapine, olanzapine and risperidone, with a significant portion of patients receiving low-dose quetiapine without dose titration. Analysis of diagnostic indications from BEACH indicated that 27% of antipsychotic prescriptions were off-label for indications such as depression, dementia, anxiety and insomnia, at much lower prescribed daily dosages. CONCLUSION: The increasing prescribing and off-label use highlights concerns about chronic adverse effects caused by antipsychotics. The combined analysis of medication dispensings and the diagnostic indications for which they are prescribed is a novel approach and throws a spotlight on the need for additional monitoring of antipsychotics.