AB0265 OPIOIDS AND ANALGESIC USE IN EARLY RHEUMATOID ARTHRITIS: A LONGITUDINAL ANALYSIS OF LINKED REAL-WORLD PRESCRIPTION DATA

Background: Large numbers of patients with rheumatoid arthritis (RA) receive regular opioids despite significant toxicity and a lack of evidence supporting their use in non-cancer pain. In order to address this situation, we need to understand when opioids are started in early RA where this has not been studied. Objectives: To examine the temporal trend of opioid prescriptions before and after RA symptom onset and to compare this with DMARD and NSAID prescriptions. Methods: RA participants (cases) were recruited as part of the Scottish Early Rheumatoid Arthritis (SERA) inception cohort1. Controls without RA (five per case), matched for sex, age and post code over the same time period, were obtained through routine data linkage. Prescription data between Jan 2009 to Nov 2019 of cases and matched controls were compared using date of RA symptom onset as reference point. The Prescriptions Per Participant (PPP) for each three-month block was estimated by dividing the number of prescribed drugs in the selected drug classes (assigned using the British National Formulary) in that time block by the number of participants in each group. The differences between mean PPP of the RA cases and controls in each time block were tested by t-test for independent groups and subsequent adjustment for multiple testing. Results: 1,720,335 prescriptions were available for analysis with 421,961 items for 950 RA cases and 1,299,374 items for 4,558 matched controls. As expected, DMARD prescriptions in the SERA cases increased after the symptom onset period and were then sustained (Figure 1: top left panel). NSAID prescriptions in RA cases peaked during the 3 months after symptom onset and then reduced progressively (top right panel). Opioid analgesic prescriptions for the RA cases increased two-fold during the reference period and then reduced 6-9 months post-symptom onset. However, unlike NSAIDs, after this there was no further significant reduction in opioid prescriptions in the RA cases, which remained stable and significantly higher than in the controls for the remaining study period. The non-opioid analgesic mean PPP increased sharply at the time of symptom onset, with a steady gradual upward trend over time (lower right panel).