Identification of Opioid Involved Health Outcomes Using Linked Hospital Care and Mortality Data (306607)
Keywords: Opioids, Natural Language Processing, Hospital Data, Mortality Data
As a federal statistical agency the National Center for Health Statistics (NCHS) provides information to guide actions and policies to improve the health of the American people, including, public health surveillance on opioid misuse and overdose. Specifically, NCHS houses three data sources that when combined, enhance researchers’ ability to analyze data on hospital care and deaths related to opioid-involved drug overdose. This presentation discusses use of natural language processing to improve identification of opioid-specific hospital encounters and deaths using three data sources, the National Hospital Care Survey (NHCS) linked to the National Death Index (NDI) and the National Vital Statistics restricted-use mortality file to create an enhanced file on care and outcomes. The data files will allow researchers to examine factors from hospital encounters and death records, including patient characteristics, risk factors, emergency department and hospital care practices, and causes of death associated with opioid-related drug overdoses presenting in hospital settings. This presentation will also highlight some examples using these enhanced data files.