Alcohol can cause risky surgical complications for patients who drink in the days leading up to a procedure, but signs of dangerous alcohol use aren’t always obvious on a patient’s chart. Artificial intelligence could help bring such problems to light, a new analysis suggests.
Patients’ electronic medical records contain diagnostic codes, but they can also include information such as notes, test results or billing data that may hint at risky alcohol use.
To pick up on contextual clues, the researchers programmed a natural language processing model to identify both diagnostic codes and other indicators of risky alcohol use, such as drinks per week exceeding recommended thresholds or a history of medical issues associated with alcohol misuse.
Misusing alcohol around a surgery is associated with higher infection rates, longer hospital stays and other surgical complications. Among the patients studied, 4.8 percent had charts that included a diagnosis code related to alcohol use. With the help of contextual clues, the model classified three times as many as being at risk, for a total of 14.5 percent.
The model did about as well as a panel of human alcohol-use experts, matching their classifications for a subset of records 87 percent of the time.
The findings point to AI as a potential partner for clinicians looking to identify patients who need intervention or postoperative supports, the researchers concluded.
The analysis could “lay the groundwork for efforts to identify other risks in primary care and beyond, with appropriate validation,” V.G. Vinod Vydiswaran, an associate professor of learning health sciences at the University of Michigan Medical School and the paper’s lead author, said in a news release. “Essentially, this is a way of highlighting for a provider what is already contained in the notes made by other providers, without them having to read the entire record.”
The researchers say they plan to eventually make the model public, but note that it will have to be trained on medical records from individual facilities.
Automated-detection of risky alcohol use prior to surgery using natural language processing
Alcohol: Clinical & Experimental Research