Nationally Representative Health Care Expenditures of Community-Based Older Adults with Pain in the United States Prescribed Opioids vs Those Not Prescribed Opioids

Abstract

Abstract Objective To compare health care expenditures between older US adults (≥50 years) with pain who were prescribed opioid medications and those who were not. Design Cross-sectional. Setting Community-based adults in the 2015 Medical Expenditure Panel Survey (MEPS). Subjects Nationally representative sample of US adults alive for the calendar year, aged 50 years or older, who reported having pain in the past four weeks. Methods Older US adults (≥50 years) with pain in the 2015 MEPS data were identified. The key independent variable was opioid prescription status (prescribed opioid vs not prescribed opioid). Hierarchical linear regression models assessed health care expenditures (inpatient, outpatient, office-based, emergency room, prescription medications, other, and total) in US dollars for opioid prescription status from a community-dwelling US population perspective, adjusting for covariates. Results The 2015 study cohort provided a national estimate of 50,898,592 noninstitutionalized US adults aged ≥50 years with pain in the past four weeks (prescribed opioid N = 16,757,516 [32.9%], not prescribed opioid N = 34,141,076 [67.1%]). After adjusting for covariates, individuals prescribed an opioid had 61% greater outpatient (β = 0.477, P < 0.0001), 69% greater office-based (β = 0.524, P < 0.0001), 14% greater emergency room (β = 0.131, P = 0.0045), 63% greater prescription medication (β = 0.486, P < 0.0001), 29% greater other (β = 0.251, P = 0.0002), and 105% greater total (β = 0.718, P < 0.0001) health care expenditures. There was no difference in opioid prescription status for inpatient expenditures (P > 0.05). Conclusions This study raises awareness of the economic impact associated with opioid use among US older adults with pain. Future research should investigate these variables in greater depth, over longer time periods, and in additional populations.