20 | Opioid Use by Disability Insurance Applicants

Since 2007, the Social Security Administration has collected data on medication use among applicants for Social Security Disability Insurance (SSDI). The administration can then use the medication data to identify opioid use among SSDI applicants. But the data set is so large and the data themselves are unstructured, with the majority of applicants reporting drug names in open-ended text fields, so the agency couldn't use the information to inform policy and programs. On this episode of On the Evidence, we talk with April Yanyuan Wu, a researcher at Mathematica, who used supervised machine learning to uncover new insights based on those data, including an estimate on the prevalence of opioid use among SSDI applicants. This episode is part of a series produced by Mathematica in support of the Association for Public Policy Analysis and Management (APPAM) and its fall research conference. More information about April's research on opioids and SSDI applicants is available here: https://www.mathematica.org/our-publications-and-findings/publications/trends-in-opioid-use-among-social-security-disability-insurance-applicants