Development of machine-learning models using pharmacy inquiry database for predicting dose-related inquiries in a tertiary teaching hospital.

BACKGROUND: Drug-related problems (DRPs) are a significant concern in healthcare. Pharmacists play a vital role in detecting and resolving DRPs to improve patient safety. A pharmacy inquiry program was established in a tertiary teaching hospital to document inquiries about physicians' orders, aimed at preventing potential DRPs or providing medication information during order reviews. OBJECTIVE: We aimed to develop machine-learning models using a pharmacy inquiry database to predict dose-related inquiries based on prescriptions and patient information. METHODS: This retrospective study analyzed 20,393 pharmacy inquiries collected between January 2018 and February 2023. Data included prescription information (drug ingredient, dose, unit, and frequency), patient characteristics (age, sex, weight, and department), and renal function. The inquiries were categorized into two classes: dose-related inquiries (e.g., wrong dose and inappropriate regimen) and non-dose-related inquiries (e.g., inappropriate drug form and administration route). Six machine-learning models were developed: logistic regression, support vector classifier, decision tree, random forest, extreme gradient boosting, and categorical boosting. To evaluate the performance of the models, the area under the receiver operating characteristic curve and the accuracy were compared. RESULTS: The CatBoost model achieved the highest performance (sensitivity: 0.92; accuracy: 0.79). The SHapley Additive exPlanations values highlighted the importance of features in the model predictions, drug ingredients, units, and renal function, in that order. Notably, lower renal function positively contributed to the prediction of dose-related inquiries. Additionally, the subsequent feature importance among drug ingredients showed that drugs such as acetylsalicylic acid, famotidine, metformin, and spironolactone strongly influenced the prediction of dose-related inquiries. CONCLUSION: Machine-learning models that use pharmacy inquiry data can effectively predict dose-related inquiries. Further external validation and refinement of the models are required for broader applications in healthcare settings. These findings provide valuable guidance for healthcare professionals and highlight the potential of machine learning in pharmacists' decision-making.

Evaluation of Pharmacy Inquiries in Physician Order Reviews for Medication Safety: A Cross-Sectional Study.

BACKGROUND AND OBJECTIVES: Despite the effort to prevent drug-related problems (DRPs) in healthcare settings, prescribing errors are common in the medication use process. In a Korean teaching hospital, pharmacists verify prescription orders during their routine order review process and document the details in a homegrown health information system (HIS). The objectives of this study were to identify the annual trends in pharmacy inquiries and to evaluate the prevalence of the inquiries by drug ingredients, including a description of the "pharmacy inquiry" screen in the HIS. MATERIALS AND METHODS: A retrospective cross-sectional study was conducted to describe pharmacy inquiries related to preventing potential DRPs during order reviews and to evaluate the associated factors for discontinuation of prescription orders on medication among inquiries using data from January 2008 to December 2021. A descriptive analysis was performed using 128,188 inquiries, documented by 245 pharmacists for 14 years. RESULTS: The frequency of inquiry steadily increased annually. The most frequent cause was "inappropriate dose or regimen" (49.1%) and "piperacillin and beta-lactamase inhibitor" was the most mentioned drug ingredient in the inquiries (3.4%). The overall acceptance rate of the pharmacists' recommendation was 82.4%, and the cause of the highest acceptance was "inappropriate mix solution" (96.5%). Hospitalization and certain inquiry topics were significantly associated with discontinuation of prescription orders on inquired medications by clinicians. CONCLUSIONS: The findings indicate that pharmacy inquiries with integrated HIS could resolve inaccuracy during physicians' order reviews and ensure safe patient care. As a tool for preventing prescribing errors, the pharmacy inquiry data can help maximize consistent improvement and optimize the medication use process in healthcare settings.