Richmond Leader

Artificial intelligence and genetics could help doctors treat alcohol addiction, VCU-led research shows

Imagine a patient has been diagnosed with alcohol use disorder, and their healthcare provider is reviewing medication options to help them curb their drinking. The provider asks the patient some basic questions, like alcohol cravings and stress levels, and collects a blood sample for genetic testing. A computer model uses this information to determine which medication would most likely support the patient with managing their alcohol use. With the help of the model, the provider gives a medication recommendation that is the best fit for their patient.

That is a possibility that Albert Arias, M.D., and his research team at Virginia Commonwealth University are trying to help make a reality.

For patients with alcohol-related health problems, abstinence is one of the most important interventions for reversing these issues or preventing them from becoming worse. A number of medications can help reduce a patient’s desire to drink, though success varies between patients.

“How a patient responds to a particular drug is the result of many factors, including their genes, behaviors and environment. Some patients have a powerful response to a medication, and some patients have no response at all,” said Arias, who is an associate professor in the VCU School of Medicine’s Department of Psychiatry. “And while we think about risks, side effects, and other factors when considering medications for patients, we don’t have a good way yet to specifically match a treatment to a person.”

To address this issue, Arias and his colleagues are using artificial intelligence to better understand which individual factors can be used to predict how patients with alcohol use disorder respond to certain medications. He presented the results last fall at the inaugural academic symposium of the Stravitz-Sanyal Institute for Liver Disease and Metabolic Health at VCU, as well as the Research Society on Alcoholism’s Annual Scientific Meeting.

Original source can be found here