PSY17 - Evaluation of Opioid use Among Patients with Back Disorders and Arthritis

Abstract

This retrospective cohort study evaluated the impact of opioid use in back pain and arthritis on health-related quality of life (HRQoL). Using data from the Medical Expenditure Panel Survey, participants must have back pain or arthritis for three of the five rounds, have no opioid use in Round 1 of the survey, and be cancer-free. Opioid use was evaluated in Rounds 2-4. Outcome measures, assessed in Round 4, were HRQoL [represented by the Mental Component Score (MCS) and the Physical Component Score (PCS)] and subdomains of MCS [Vitality (VT), Social Functioning (SF), Role Emotional (RE), and Overall Mental Health (MH)] and PCS [Bodily Pain (BP), Physical Functioning (PF), Role Functioning (RF), and General Overall Health (GH)]. Estimated propensity for opioid use was used in a 1:1 Greedy match. Linear regressions were used to analyze outcomes after matching. 1,871 eligible participants (690 opioid and 1,181 non-opioid users) were identified. MCS was not different between opioid and non-opioid users. PCS and BP were significantly lower among opioid users versus non-opioid users (PCS: β=-2.26, 95%CI: -4.16 to -0.37, p=0.0194; BP: β=-2.72, 95%CI: -4.72 to -0.72, p=0.0077). MCS and RE subdomain were significantly higher among short-term opioid user (<1 month) compared to non-opioid users (MCS: β=2.71, 95%CI: 0.65 to 4.76, p=0.0099; RE: β=3.44, 95%CI: 1.25 to 5.63, p=0.0021). PCS and its subdomains trended towards improvement with short-term use. Long-term users (≥1 month) had significantly lower SF, PCS, BP, PF, RF, and GH scores compared to non-opioid users. Low-dose opioid users had lower PCS, BP, and RF scores as compared to non-opioid users, although most differences never reached clinical significance. No difference in HRQoL was found between high-dose users and non-opioid users. Opioid use, and particularly long-term opioid use, in those with back pain and arthritis is not associated with improved HRQoL.