Enhancing Adverse Drug Event Detection in Electronic Health Records Using Molecular Structure Similarity: Application to Pancreatitis

Abstract

Background Adverse drug events (ADEs) detection and assessment is at the center of pharmacovigilance. Data mining of systems, such as FDA’s Adverse Event Reporting System (AERS) and more recently, Electronic Health Records (EHRs), can aid in the automatic detection and analysis of ADEs. Although different data mining approaches have been shown to be valuable, it is still crucial to improve the quality of the generated signals. Objective To leverage structural similarity by developing molecular fingerprint-based models (MFBMs) to strengthen ADE signals generated from EHR data. Methods A reference standard of drugs known to be causally associated with the adverse event pancreatitis was used to create a MFBM. Electronic Health Records (EHRs) from the New York Presbyterian Hospital were mined to generate structured data. Disproportionality Analysis (DPA) was applied to the data, and 278 possible signals related to the ADE pancreatitis were detected. Candidate drugs associated with these signals were then assessed using the MFBM to find the most promising candidates based on structural similarity. Results The use of MFBM as a means to strengthen or prioritize signals generated from the EHR significantly improved the detection accuracy of ADEs related to pancreatitis . MFBM also highlights the etiology of the ADE by identifying structurally similar drugs, which could follow a similar mechanism of action. Conclusion The method proposed in this paper provides evidence of being a promising adjunct to existing automated ADE detection and analysis approaches.

Integrative Approaches for Predicting In Vivo Effects of Chemicals from their Structural Descriptors and the Results of Short-Term Biological Assays

“…al. showed increased precision when a chemical similarity model was pooled with a model based on clinical notes [66,67]. In the second case, an ensemble chemical-biological model may compensate for the invalid predictions by the QSAR model outside its chemical coverage area.…”

Similarity-based modeling in large-scale prediction of drug-drug interactions

“…This approach can generate sets of potential DDI candidates for both pharmacokinetic and pharmacodynamic interactions. The set of new potential DDIs could be used to filter out candidates extracted from pharmacovigilance databases, such as Electronic Health Records, and to strengthen the signals obtained through data mining 6,7 .…”

Similarity‐Based Modeling Applied to Signal Detection in Pharmacovigilance

“…Similarity methods were applied to medication‐wide association studies performed in an administrative claims database with 46 million patients. However, similarity‐based modeling can be applicable to improve signal detection steps using other data mining algorithms or other type of pharmacovigilance data, such as the FDA Adverse Event Reporting System or electronic health records 21 , 22 . The method also allows rationalizing the relevance of the signals to optimize the decision‐making process.…”

“…Our previous system, 20 although capable of generating an enriched subset of ADE candidates, showed some limitations in its ability to exclude some false positives from the final signal selection. Other studies using other data algorithms and data sources showed the potential to provide an enriched set of drug candidates that can cause the ADE 21 , 22 . However, improvement in the precision in this enriched set of candidates was achieved through the application of 2D structural similarity 21 , 22 …”