Mining reported adverse events induced by potential opioid-drug interactions

Area of Science:

• Pharmacovigilance
• Drug Safety
• Computational Medicine

Background:

• Opioid analgesics are common for end-of-life pain but carry risks of addiction and overdose.
• Adverse events (AEs) from opioid drug-drug interactions (ODIs) are not well-characterized.
• This study examines ODIs and severe AEs using real-world data.

Purpose of the Study:

• To investigate potential severe adverse events (AEs) arising from drug-drug interactions between opioid and nonopioid medications (ODIs).
• To explore the association between ODIs and severe AEs using a large database of adverse event reports.
• To evaluate the feasibility of predicting AEs associated with polypharmacy.

Main Methods:

• Utilized millions of adverse event reports from the FDA Adverse Event Reporting System (FAERS).
• Employed odds ratio (OR)-based analysis and visualization for single drugs and pairwise ODIs.
• Applied multilabel (multi-AE) learning models for polypharmacy AE prediction.

Main Results:

• Identified the 12 most prescribed opioids in FAERS.
• OR analysis revealed diverse AEs linked to individual opioids.
• Many ODIs were found to dramatically increase the risk of severe AEs.
• Multilabel models for oxycodone ODIs achieved area under the curve values of 0.81–0.88 for 5 severe AEs.

Conclusions:

• Data analysis and visualization effectively identified novel polypharmacy-associated AEs from FAERS data.
• The approach successfully confirmed known drug interactions.
• New opioid-specific AEs were identified, potentially influencing prescribing practices.