Using Natural Language Processing to Identify Adverse Drug Events of Angiotensin Converting Enzyme Inhibitors
An OMNY Health™ Poster
Natural language processing (NLP) is a potential mechanism for identifying adverse drug events (ADEs) with better efficiency and accuracy.
On June 9-11, 2021, this poster was presented at the International Forum on Quality and Safety in Healthcare EUROPE.
Discussion and Conclusions:
- This specific NLP system successfully identified several classes of ADEs among patients exposed and not exposed to ACE-I therapy.
- Analyses in progress include (1) assessing additional ICD-10 codes positively correlated with ACE-I use; (2) incorporating structured data to reduce confounder effects; and (3) using a larger sample.
- NLP systems for identifying ADEs from clinical text have great potential; to gain widespread adoption, they must first be properly validated against established clinical knowledge.