Identification of Potential Drug-Drug Interactions Using EMR Text-Mining on Atherosclerotic Heart Disease Patients

Atherosclerotic heart disease patients often exhibit comorbidities, leading to polypharmacy and a heightened risk of Drug-Drug Interaction (DDI). This study investigated the harmful potential of Drug-Drug Interactions (DDIs) in patients diagnosed with atherosclerotic heart disease using Electronic Medical Records (EMR) from a private hospital in the East Java region, Indonesia. The research employed a comprehensive methodology encompassing preprocessing, drug name extraction, mapping, and assessing concomitant medications by leveraging text-mining technology on 24,672 records. DDI identification is carried out based on the DDInter database. Based on 24,660 data that have been cleaned, 149 generic drugs were obtained, of which 117 are available in the DDInter database. The findings revealed 33 DDIs at the major risk level among 117 drugs. Among those data, the number of records containing DDI at major, moderate, and minor risk levels or their combination in EMR data was 15,514 (62.91%), and many records were found to have more than one risk level. The DDIs found were 10.45% (2,577 records) at the major risk level, 46.80% (11,542 records) at the moderate risk level, and 44.78% (11,043 records) at the minor risk level. The prominent presence of major-risk and moderate-risk interactions underscores the significance of addressing DDIs in clinical practice. Healthcare services can be reinforced through education, adopting computerized prescribing systems, and enhancing drug information dissemination to mitigate these risks. The results contribute valuable insights into the prevalence of DDIs in atherosclerotic heart disease patients, guiding efforts to improve patient safety and optimize pharmaceutical interventions.