A pipeline to extract drug-adverse event pairs from multiple data sources

Abstract

BackgroundPharmacovigilance aims to uncover and understand harmful side-effects of drugs, termed adverse events (AEs). Although the current process of pharmacovigilance is very systematic, the increasing amount of information available in specialized health-related websites as well as the exponential growth in medical literature presents a unique opportunity to supplement traditional adverse event gathering mechanisms with new-age ones.MethodWe present a semi-automated pipeline to extract associations between drugs and side effects from traditional structured adverse event databases, enhanced by potential drug-adverse event pairs mined from user-comments from health-related websites and MEDLINE abstracts. The pipeline was tested using a set of 12 drugs representative of two previous studies of adverse event extraction from health-related websites and MEDLINE abstracts.ResultsTesting the pipeline shows that mining non-traditional sources helps substantiate the adverse event databases. The non-traditional sources not only contain the known AEs, but also suggest some unreported AEs for drugs which can then be analyzed further.ConclusionA semi-automated pipeline to extract the AE pairs from adverse event databases as well as potential AE pairs from non-traditional sources such as text from MEDLINE abstracts and user-comments from health-related websites is presented.

Social Media Mining for Public Health Monitoring and Surveillance

“…Early techniques primarily focused on lexicon-based approaches, where the natural language mentions of the elements of interest are encoded in lexicons and these are utilized to detect their mentions in text. 2,45,46,55 These techniques have led to the development of health-related lexical resources from social media sources (e.g., the Consumer Health Vocabulary 100 ). The use of colloquial language, however, limits the performance of such approaches.…”

Extraction of ObjectProperty-UsageMethod Relation from Web Documents

“…However, unlike our research, the semantic relations extracted by previous researches [4,5,[7][8][9] mostly occurred within one sentence containing either the relation between two NPs (NP1 and NP2 of a sentence expression(S) as S NP1 VP; VP verb NP2 |…) or the relation between NP1 and VP.…”

“…respectively. In 2014, [7] applied the semi-automatic pipeline detection and the extraction of drugadverse event (drug-AE) pairs from unstructured data, i.e. user-comment blogs and MEDLINE abstracts, and the structure database (Food and Drug Administration Adverse Event Reporting System).…”

“…Several strategies [4][5][6][7][8][9] have been proposed to determine the semantic relation from the textual data.…”

“…There are several techniques [4][5][6][7][8][9] that have been used to extract the semantic relations involved with the object properties and the procedural knowledge based on explanations from documents (see Section 2) whilst the other previous researches [10] on the semantic relation determination from texts relies on the relations, i.e. is-a, part-of, and cause-effect, between two entities of noun phrases without boundary consideration.…”