Kaid Health's natural language processing model was successful in identifying prolonged opioid usage three months after orthopedic surgery, according to a study published July 6 in Regional Anesthesia & Pain Medicine.
Researchers from the University of California San Diego compared the AI model's capabilities in detecting prolonged postoperative opioid use with the traditional method, in which clinician reviewers track usage from electronic medical records.
"The study revealed an impressive 90 percent concordance between the AI model and the clinicians, highlighting the utility of the Kaid model in extracting valuable clinical insights from medical charts to enhance patient care," Rodney Gabriel, MD, one of the authors of the study, said in a July 10 press release from Kaid Health.
The model is designed to draw conclusions based on relevant patient information from structured and unstructured EMR data, such as medications, conditions, issues and test results. Kaid's AI model can also support administrative tasks.