Harnessing Natural Language Processing and Machine Learning to Enhance Identification of Opioid-involved Health Outcomes in the National Hospital Care Survey

Articles and reports: 11-522-X202100100016
Description: To build data capacity and address the U.S. opioid public health emergency, the National Center for Health Statistics received funding for two projects. The projects involve development of algorithms that use all available structured and unstructured data submitted for the 2016 National Hospital Care Survey (NHCS) to enhance identification of opioid-involvement and the presence of co-occurring disorders (coexistence of a substance use disorder and a mental health issue). A description of the algorithm development process is provided, and lessons learned from integrating data science methods like natural language processing to produce official statistics are presented. Efforts to make the algorithms and analytic datafiles accessible to researchers are also discussed.

Key Words: Opioids; Co-Occurring Disorders; Data Science; Natural Language Processing; Hospital Care

Issue Number: 2021001
Author(s): Brown, Amy M; Adams, Nikki
Main Product: Statistics Canada International Symposium Series: Proceedings
Format Release date More information
PDF October 22, 2021

Related information

Subjects and keywords

Subjects

Keywords

Date modified: