Potential drug-drug interactions (PDDIs) represent a significant causality for adverse drug events. Unfortunately, current resources for providers are incomplete and inaccurate. We propose a new PDDI knowledge representation paradigm that we hypothesize will yield more clinically relevant evidence than is currently possible. Starting from our extensive body of preliminary work, we will build a framework that implements the new paradigm using statins and psychotropics (antidepressants and antipsychotics). We expect that the framework will be generalizable to PDDIs involving other drugs, including those predicted using methods from pharmacology and bioinformatics. We will advance three research aims while building the exemplar framework. The first research aim is to derive a new meta-data standard for representing PDDI knowledge that satisfies the information needs of pharmacist working in different care settings. An information needs inquiry will result in clinical scenarios that will then inform, and later validate, the new standard. We will design the standard so that it reflects the best thinking of Semantic Web community and will have a high likelihood of widespread adoption. We will then combine the new standard with semantic annotation and best practices for publishing Linked Data to create a Semantic Web knowledge base of statin and psychotropic PDDIs. The second research aim is to compare PDDI evidence on the Semantic Web with existing PDDI knowledge resources for completeness, accuracy and currency. We will validate a mechanism for linking statin and psychotropic PDDI assertions to relevant evidence on the Semantic Web. Because pharmacogenomics can impact many PDDIs, we will also link to an interoperable representation of this evidence. Two pharmacists will then compare the coverage and quality of the PDDI evidence on the Semantic Web with three existing resources using a new PDDI evidence scoring tool. The third research aim is to investigate a process for filling in gaps in clinically useful PDDI knowledge that cannot be filled with available evidence. We will utilize a consensus-based approach to select high priority PDDIs and evaluate their clinical relevance by retrospective cohort studies. We will extend the Linked Data PDDI knowledge base with the results of these studies, and make the knowledge base publicly available via a pilot web portal. The proposed work will contribute to public health by making more effective use of PDDI evidence, filling in important gaps in drug safety knowledge, and spurring innovations in drug information retrieval.
We propose a new knowledge representation paradigm for potential drug-drug interactions that we hypothesize will yield more clinically relevant evidence than is currently possible. The proposed work will contribute to public health by making more effective use of drug-drug interaction evidence, filling in important gaps in drug safety knowledge, and spurring innovations in drug information retrieval.
|Knowledge Base workgroup of the Observational Health Data Sciences and Informatics (OHDSI) collaborative (2017) Large-scale adverse effects related to treatment evidence standardization (LAERTES): an open scalable system for linking pharmacovigilance evidence sources with clinical data. J Biomed Semantics 8:11|
|Ie, Kenya; Chou, Eric; Boyce, Richard D et al. (2017) Potentially Harmful Medication Use and Decline in Health-Related Quality of Life among Community-Dwelling Older Adults. Drugs Real World Outcomes 4:257-264|
|Romagnoli, Katrina M; Nelson, Scott D; Hines, Lisa et al. (2017) Information needs for making clinical recommendations about potential drug-drug interactions: a synthesis of literature review and interviews. BMC Med Inform Decis Mak 17:21|
|Freimuth, R R; Formea, C M; Hoffman, J M et al. (2017) Implementing Genomic Clinical Decision Support for Drug-Based Precision Medicine. CPT Pharmacometrics Syst Pharmacol 6:153-155|
|Romagnoli, Katrina M; Boyce, Richard D; Empey, Philip E et al. (2017) Design and evaluation of a pharmacogenomics information resource for pharmacists. J Am Med Inform Assoc 24:822-831|
|Utecht, Joseph; Brochhausen, Mathias; Judkins, John et al. (2017) Formalizing Evidence Type Definitions for Drug-Drug Interaction Studies to Improve Evidence Base Curation. Stud Health Technol Inform 245:960-964|
|Voss, E A; Boyce, R D; Ryan, P B et al. (2017) Accuracy of an automated knowledge base for identifying drug adverse reactions. J Biomed Inform 66:72-81|
|Huser, Vojtech; DeFalco, Frank J; Schuemie, Martijn et al. (2016) Multisite Evaluation of a Data Quality Tool for Patient-Level Clinical Data Sets. EGEMS (Wash DC) 4:1239|
|Boyce, Richard D; Handler, Steven M; Karp, Jordan F et al. (2016) Preparing Nursing Home Data from Multiple Sites for Clinical Research - A Case Study Using Observational Health Data Sciences and Informatics. EGEMS (Wash DC) 4:1252|
|Hochheiser, Harry; Ning, Yifan; Hernandez, Andres et al. (2016) Using Nonexperts for Annotating Pharmacokinetic Drug-Drug Interaction Mentions in Product Labeling: A Feasibility Study. JMIR Res Protoc 5:e40|
Showing the most recent 10 out of 19 publications