Medication Extraction and Reconciliation Knowledge Instrument (MERKI) Parser

MERKI extracts patient medication information from structured and free-text clinical data sources. The extracted medication information can be used for medication reconciliation, pharmacovigilance and quality assurance. MERKI outputs medication information in the form of drug name, dose, route and frequency. It relies on a set of parsing rules and a lexicon, both of which can be easily customized.

MERKI is open source under the GNU general public license. When using MERKI, please refer to the following paper: Extracting Structured Medication Event Information from Discharge Summaries. 2008. Sigfried Gold, Noémie Elhadad, Xinxin Zhu, James J. Cimino, and George Hripcsak. Proceedings of the American Medical Informatics Associations Annual Symposium (AMIA'08). pp. 237-241.

Download source code

Demo

Paste a chunk of clinical text into the box below (maximum, 2000 characters), choose an output format, and hit submit to see how the parser extracts medication information from text.

Some random text to use as an example of some parser features: Discharge medications: Procardia XL 60 mg p.o. p.r.n. for severe wheezing, ferros sulfate 300 mg p.o. b.i.d., Cipro 250 mg p.o. q12hQ

XML Table