Detection of Adverse Drug Reactions in hospitalized patients: a network analysis approach

Abstract

Objective: We aimed to investigate whether network analysis can be used to estimate patterns of Adverse Drug Reactions and drugs involved. Methods: Patients admitted from 18 years of age or older, hospitalized for more than 24 hours, and using at least one drug during hospitalization were included. Results: 8060 patients were observed, and 358 cases of Adverse Drug Reactions were identified (4.43%). The network graph shows that the occurrence of hypotension induced by furosemide, spironolactone and enalapril is related to serum changes in potassium and the occurrence of renal failure. Centered around nausea and vomiting node, there is a great variety of drugs from different classes involved with this Adverse Drug Reaction and without other connections. Conclusion: Network analysis is a promising strategy for identifying patterns that correlate adverse reactions to drugs administered during hospitalization.