Estimating risk of reported versus theoretical drug-drug interactions in headaches medicine: an exhaustive comparison between DrugBank and FAERS database for abortive and preventive combinations.

backgroundPolypharmacy is common in headache medicine. This project uses DrugBank and Food and Drug Administration's (FDA) Adverse Event Reporting System (FAERS) to determine the likelihood of a theoretical and empirically reported interaction between a given number of headache abortive and preventive medications.

methodsAll headache medications referenced in "The American Headache Society Position Statement on Integrating New Migraine Treatments into Clinical Practice" and Szperka's "Migraine care in the era of COVID-19: clinical pearls and plea to insurers" were included. All possible combinations of up to three abortives and preventatives were screened for drug-drug interactions through searches in DrugBank and FAERS. If at least one drug-drug interaction was listed, then it was included in our analysis. The percentage of combinations containing an interaction was then compared across the two databases.

resultsOf the 38 abortives and 23 preventives included, once more than 3 drugs are used in any combination, a drug-drug interaction was >99% likely, per DrugBank. However, per FAERS, the reported interaction was 39% to 60% likely once 3 or 4 drugs are used in any combination. In FAERS, the likelihood of interaction rised most dramatically once 3 or more drugs are used.

conclusionsTheoretical interactions, based on DrugBank, maybe overstated when compared to actual observed interactions in FAERS. Future direction is needed delving into the types of interactions recorded in each database.