Drug-drug interaction and acute kidney injury development: A correlation-based network analysis

Area of Science:

• Nephrology
• Pharmacology
• Critical Care Medicine

Background:

• Drug-induced nephrotoxicity is a common, preventable cause of acute kidney injury (AKI).
• Drug-drug interactions can lead to additive or synergistic toxicity, complicating prediction and detection.
• The critical care setting involves complex medication use, including new and off-label drugs, increasing nephrotoxicity risks.

Purpose of the Study:

• To evaluate medication utilization patterns and their combinations associated with AKI in critically ill patients.
• To identify specific drug combinations contributing to nephrotoxicity.
• To explore the role of estimated glomerular filtration rate (eGFR) in assessing drug-induced nephrotoxicity risk.

Main Methods:

• Correlation-based network analysis (CNA) was employed on a large cohort of ICU admissions (2007-2018).
• Pairwise medication-AKI correlation analysis assessed synergistic or additive drug effects.
• Isolated medication analysis investigated inherent nephrotoxicity of individual drugs.

Main Results:

• 244 associations were found among 1,555 unique medication types across 147,289 ICU admissions.
• Pairwise analysis revealed 233 significant correlations, while isolated analysis identified ten significant AKI associations.
• Substantial differences in AKI associations were observed based on eGFR levels, indicating its importance in risk assessment.

Conclusions:

• An AI model was developed to identify patient-agnostic relationships between medications/pairs and AKI incidence.
• This model can serve as a continuous quality assurance tool for monitoring drug-associated nephrotoxicity risks.
• The study underscores the need to consider eGFR in evaluating drug interactions and nephrotoxicity.