Researchers from Maywood, Ill.-based Loyola Medicine and Loyola University Chicago used an artificial intelligence technique to identify alcohol misuse among trauma patients.
The technique was able to differentiate between trauma patients who misused alcohol and those who did not in 78 percent of cases. Researchers published their findings in the Journal of the American Medical Informatics Association.
Four notes:
1. One in three trauma patients misuse alcohol, and many trauma cases are alcohol-related, according to the study.
2. For the study, researchers used natural language processing and machine learning — two types of artificial intelligence — to sift through records of 1,422 adult patients admitted to Loyola's Level 1 trauma center between April 2013 and November 2016. These records included 91,405 EHR clinician notes.
3. The researchers determined 16 concepts that were indicative of alcohol misuse, including references to intoxication, neglect, drinking problems, liver imaging, sexual activity, marijuana and the B1 vitamin thiamine.
4. Using natural language processing and machine learning, researchers were able to sift through the EHR notes to flag potential patients who misuse alcohol.
Study authors concluded that natural language processing has adequate predictive validity for detecting alcohol misuse among trauma patients, and suggested their technique provides a potential approach to overcome staffing challenges.