Leveraging Semantic Technologies to Bring Active Pharmacovigilance at the Point of Care
Any type of content formally published in an academic journal, usually following a peer-review process.
Pharmacovigilance is recognized worldwide as an important public health issue. Nevertheless, there is a lack of comprehensive tools to support healthcare professionals in drug safety risk assessments in the clinical setting. In the scope of the PVClinical project, we employ semantic technologies for developing a point-of-care platform for the early identification and assessment of possible Adverse Drug Reactions (ADRs). Our approach relies on the following technical pillars: (a) leveraging the OMOP Common Data Model as a reference data model to encode relevant Electronic Health Record (EHR) data at the local site; (b) the Linked Data paradigm for integrating data from various and heterogeneous data sources beyond local EHRs (e.g. bibliographic databases, Spontaneous Reporting Systems, as well as social media platforms) in a common knowledge graph (KG), and (c) advanced Analytics for exploiting the KG, including text mining and statistical inference for risk assessment. The ultimate goal is to support clinicians in obtaining comprehensive information on potential drug safety risks that shall be assessed in their daily practice.