The association of pain interference and opioid use with healthcare utilization and costs, and wage loss among adults with osteoarthritis in the United States

Aim: To examine associations of opioid use and pain interference with activities (PIA), healthcare resource utilization (HRU) and costs, and wage loss in noninstitutionalized adults with osteoarthritis in the United States (US). Methods: Adults with osteoarthritis identified from the Medical Expenditure Panel Survey for 2011/2013/2015 were stratified by no-opioid use with no/mild PIA, no-opioid use with moderate/severe PIA, opioid use with no/mild PIA, and opioid use with moderate/severe PIA. Outcomes included annualized total HRU, direct healthcare costs, and wage loss. Multivariable regression analyses were used for comparisons versus no-opioid use with no/mild PIA (referent). The counterfactual recycled prediction method estimated incremental costs. Results reflect weighted nationally representative data. Results: Of 4,921 participants (weighted n = 20,785,007), 46.5% had no-opioid use with no/mild PIA; 23.2% had no-opioid use with moderate/severe PIA; 9.6% had opioid use with no/mild PIA; and 20.7% had opioid use with moderate/severe PIA. Moderate/severe PIA and/or opioid use were associated with significantly higher HRU and associated costs, and wage loss. Relative to adults with no/mild PIA, opioid users with moderate/severe PIA were more likely to have hospitalizations, specialist visits, and emergency room visits (all p <.001). Relative to the referent, opioid use with no/mild PIA had higher per-patient incremental annual total healthcare costs ($11,672, 95% confidence interval [CI] = $11,435–$11,909) and wage loss ($1,395, 95% CI = $1,376–$1,414) as did opioid use with moderate/severe PIA ($13,595, 95% CI = $13,319–$13,871; and $2,331, 95% CI = $2,298–$2,363) (all p <.001). Compared with the referent, estimated excess national total healthcare costs/lost wages were $23.3 billion/$1.3 billion for opioid use with no/mild PIA, and $58.5 billion/$2.2 billion for opioid use with moderate/severe PIA. Limitations: Unobservable/unmeasured factors that could not be accounted for. Conclusions: Opioid use with moderate/severe PIA had significantly higher HRU, costs, and wage loss; opioid use was more relevant than PIA to the economic burden. These results suggest unmet needs for alternative pain management strategies.