Document resource
Background and Importance Drug–drug interactions (DDIs) are a major safety concern in oncology. These interactions may arise from pharmacokinetic or pharmacodynamic mechanisms, potentially compromising the efficacy and safety of both anticancer and concomitant therapies. Although several studies have compared drug interaction databases or focused on oral antineoplastic agents, there is still no universal or standardised tool to comprehensively assess DDIs.Aim and Objectives The aim of this study is to compare the identification and classification of potential DDIs between oncohematological treatments and patients’ chronic medications across four commonly used databases.Material and Methods A prospective study with retrospective data collection was conducted. A total of 386 oral and intravenous oncohematological prescriptions were recorded in August 2025, and the 40 most frequently prescribed active ingredients (AIs) were selected. Concomitant chronic medications were obtained from electronic prescription records, identifying the 20 most commonly prescribed AIs. Each home medication was systematically compared with the 40 oncohematological agents in four databases (DrugBank , OncoAssist , UpToDate and Micromedex ). All identified interactions were recorded and classified according to severity (major, moderate, minor) and clinical consequence (reduced effectiveness or increased risk of toxicity). The Fleiss’ kappa coefficient was used to assess the strength of agreement in DDI severity classification among the four databases and it was interpreted according to Landis and Koch’s criteria.Results A summary table including the 40 oncohematological and 20 chronic medication AIs was created. DrugBank detected the largest number of DDIs (261: 86 minor, 85 moderate, 90 major), followed by OncoAssist (156: 16 minor, 100 moderate, 40 major). UpToDate detected 67 DDIs (7 minor, 58 moderate, 2 major), and Micromedex detected 28 (5 moderate, 23 major). Regarding clinical consequences, DrugBank reported 217 DDIs associated with increased toxicity and 44 with reduced effectiveness; OncoAssist reported 123 and 33, UpToDate 41 and 31, and Micromedex 20 and 8, respectively. The Fleiss’ kappa coefficient for DDI detection was 0.145 (95% CI: 0.126–0.165; p<0.001), indicating slight inter-database agreement.Conclusion and Relevance There is considerable variability among DDI databases regarding the detection and classification of potential interactions in oncohematological prescriptions. To ensure patient safety, a multidisciplinary, case-by-case evaluation is crucial to support safe and individualised therapeutic decisions.Conflict of Interest No conflict of interest