全球智库信息集成服务系统

The rising prevalence of opioid use nationwide coupled with the high share of SSDI applicants with conditions associated with opioid use, such as musculoskeletal conditions, suggests that opioid use may be common and increasing among SSDI applicants. Although applicants cannot qualify for SSDI solely on the basis of drug addiction, in some cases opioid use may exacerbate the effects of other conditions that meet the SSDI qualifications. Understanding the pattern of opioid use among SSDI applicants has important implications for projecting resources needed to adjudicate SSDI applications and for the size and composition of the SSDI caseload.Yet, little is known about the rates of opioid use among SSDI applicants because of data limitations. Although SSDI applicants are required to report medications, medications are recorded as a combination of coded and open-ended text fields making them difficult to use for research. This study addresses this challenge by testing the use of machine learning to classify free-form text of medication information in the Social Security Administration’s administrative data. The study uses an innovative, supervised machine-learning algorithm to identify opioids recorded in free-form text and combines that information with opioids identified in populated medication codes. With this information, we produce statistics on the prevalence of opioid use among SSDI applicants, fiding that 35 percent of applicants reported use of one or more opioids when they applied to SSDI in 2013. The most frequently reported opioids were Tramadol, Hydrocodone with APAP, Oxycodone, Percocet, and Vicodin. Nearly half of applicants mentioned pain as the reason for the use of medicine at least once; among those using opioids, 89 percent reported using the medicine for pain management.