Ogura Toru, Shiraishi Chihiro
Clinical Research Support Center, Mie University Hospital, Tsu, JPN.
Pharmacy, Mie University Hospital, Tsu, JPN.
Cureus. 2025 Feb 8;17(2):e78746. doi: 10.7759/cureus.78746. eCollection 2025 Feb.
Background The increasing prevalence of polypharmacy has raised concerns about drug-drug interactions (DDIs) and their impact on patient safety. Database-based DDI detection often suffers from insufficient patient background information and missing data, limiting the accuracy and applicability of DDI assessments. A novel model is needed to overcome these limitations and provide a more comprehensive evaluation of DDIs to enhance patient safety in the context of multiple medication use. Objectives This study aims to develop and validate a novel model for evaluating both the beneficial and detrimental effects of DDIs on patient safety. The model is designed to address challenges associated with insufficient patient background information and missing data in database studies while providing a comprehensive assessment of DDIs using statistical inference and hypothesis tests. Methods To address the challenges of insufficient patient background information and missing data often encountered in database studies, the proposed model incorporates an overlap parameter. This parameter represents the degree of commonality in patient profiles susceptible to adverse events from individual drug administrations. The magnitude of DDIs is presented in a 2×2 contingency table constructed by the occurrence or non-occurrence of specific adverse events in observed value and expected value estimated from the model. This tabular format facilitates the assessment of DDIs using statistical inference and hypothesis tests. Results Simulations under various settings confirmed that significance levels for statistical hypothesis tests were strictly observed. Furthermore, applications to real-world databases demonstrated that the proposed model effectively identifies both positive and negative DDIs. Conclusions This research provides healthcare professionals with a robust and practical tool for enhanced DDI detection and management. The presentation of findings in a familiar 2×2 contingency table format improves the accessibility of our results, facilitating straightforward interpretation. The proposed model has the potential to promote a safer healthcare environment for patients on multiple medications, ultimately enhancing patient safety and treatment efficacy.
背景 多重用药的日益普遍引发了人们对药物相互作用（DDIs）及其对患者安全影响的担忧。基于数据库的药物相互作用检测常常因患者背景信息不足和数据缺失而受到影响，限制了药物相互作用评估的准确性和适用性。需要一种新模型来克服这些局限性，并在多种药物使用的背景下对药物相互作用进行更全面的评估，以提高患者安全。目的 本研究旨在开发并验证一种用于评估药物相互作用对患者安全的有益和有害影响的新模型。该模型旨在解决数据库研究中与患者背景信息不足和数据缺失相关的挑战，同时使用统计推断和假设检验对药物相互作用进行全面评估。方法 为解决数据库研究中经常遇到的患者背景信息不足和数据缺失的挑战，所提出的模型纳入了一个重叠参数。该参数表示个体药物给药易发生不良事件的患者特征中的共性程度。药物相互作用的大小以一个2×2列联表呈现，该表由观察值中特定不良事件的发生或未发生以及根据模型估计的预期值构建。这种表格形式便于使用统计推断和假设检验来评估药物相互作用。结果 在各种设置下的模拟证实严格遵守了统计假设检验的显著性水平。此外，在实际数据库中的应用表明，所提出的模型能够有效识别正向和负向药物相互作用。结论 本研究为医疗保健专业人员提供了一种强大而实用的工具，用于加强药物相互作用的检测和管理。以熟悉的2×2列联表格式呈现结果提高了我们结果的可及性，便于直接解释。所提出的模型有可能为使用多种药物的患者促进更安全的医疗环境，最终提高患者安全和治疗效果。
