Development of a multivariable prediction model to assess potential drug-drug interactions in chronic kidney disease patients

Abstract

Potential drug-drug interactions (pDDIs) are highly prevalent in chronic kidney disease (CKD) patients, owing to the existence of various comorbidities and the large number of drugs used to treat them. This study aimed to evaluate the number and types of pDDIs observed in the study population and to develop a prediction model based on various risk factors. It was conducted retrospectively at a tertiary care teaching hospital and included 392 CKD patients. Relevant patient demographics and clinical details were collected and documented in case record forms. Using the Micromedex® Drug-Reax® System, the acquired data were screened to identify and classify pDDIs, and Poisson regression was used to identify independent risk factors associated with the number of pDDIs. Data entry and analysis were done using IBM Statistical Package for the Social Sciences software v20.0. A total of 2,054 interacting drug pairs were found and male gender, comorbid conditions like ischemic heart disease, hypertension, diabetes mellitus, and congestive heart failure, a higher number of therapeutic subgroups, and drugs per prescription were identified as independent risk factors associated with an increase in the number of pDDIs. The presence of liver disease was the only factor that reduced the number of pDDIs. Our study identified the significant risk factors for pDDIs in CKD patients and developed a prediction model. This can play a significant role in the early detection of pDDIs using prior information about the patient characteristics and attributes of various administered drugs.