Utility of Candidate Genes From an Algorithm Designed to Predict Genetic Risk for Opioid Use Disorder - JAMA

This study examines whether 15 genetic variants used in an opioid use disorder risk algorithm can reliably identify people at increased risk. Researchers analyzed information from 452,664 participants in the VA Million Veteran Program who had been exposed to opioids. Together, the variants explained only 0.40% of the variation in opioid use disorder risk. A research model using these variants correctly classified people with and without the disorder approximately 53% of the time in an independent sample. The findings raise concerns about inaccurate results, false reassurance, stigma, and potential bias in care. The study evaluated the genetic variants rather than directly reproducing the manufacturer’s proprietary algorithm, an important distinction when interpreting its conclusions.