Can machine learning aid in determining who should receive opioids? Research has shown that the prevalence of opioids in the Medicare population for a given year ranges between 26% and 52% depending on the cohort. Those who receive opioid prescriptions are at higher risk for adverse drug events and opioid-related complications and injury.
Properly understanding risk is important to assessing whether a Medicare member should be prescribed opioids or continue opioid therapy. Leveraging both machine learning and historical claims data can help identify potential opioid use disorder cases more effectively. Milliman’s Joseph Boschert, Jim Hong, and Kyle Ligon offer more perspective in the paper “More accurately assessing opioid risk in Medicare members.”