Clinical reliability and concordance of drug-drug interaction tools as clinical decision support systems in patients with epilepsy

PURPOSE: To evaluate the concordance and clinical correlation of drug-drug interaction tools (DDITs) as clinical decision support systems (CDSS) in patients with epilepsy. METHODS: All patients from the Epi2020 study cohort receiving ≥1 anti-seizure medication (ASM) and ≥1 concomitant drug (CD) were included. Individual treatment regimens were analyzed for drug-drug interactions (DDIs) using five DDITs (Drugbank, Drugs.com, mediQ, Medscape and WebMD). To address the clinical significance of possible DDIs, Spearman correlation between the number of detected DDIs and two established adverse event (AE) metrics (LAEP, QOLIE-31) was performed using post-hoc correction by Fisher's z-transformation (FZT) for the total number of drugs taken. Multivariate ordinal regression analysis (MORA) was performed to identify ASM or CD classes associated with severe DDIs. RESULTS: Overall, 140 patients (57.9% female, median age 48 years) taking a median number of 4.0 drugs (2.0 ASMs and 2.0 CDs) were included. DDI were found in 51.4%-84.3% and severe interactions in 2.1%-17.8% of patients, depending on the DDIT. The concordance rate between DDITs was only 6.4% for all and only 63.6% for severe detected DDI, respectively. All concordant cases involved no detected interactions. The number of DDIs significantly correlated with AE metrics in 3/5 DDITs. Following the FZT, none of the DDI/AE correlations were superior to that between the number of DDIs and the number of drugs taken. MORA identified topiramate, valproate, zonisamide, hormones, and antipsychotics as independent predictors for the detection of severe DDIs. CONCLUSION: When using DDITs as CDSS, it is important to consider that the results of different tools may vary greatly from one another and do not necessarily correlate with clinical AEs. STUDY REGISTRATION: The Epi2020 study was registered under the trial registration number: DRKS00022024, U1111-1252-5331.