Pos1337 long-term opioid use among patients with rheumatic and musculoskeletal diseases: impact of varying definitions

Background People living with rheumatic and musculoskeletal diseases (RMD) are frequently prescribed opioids for non-cancer pain. A proportion of RMD patients newly prescribed an opioid will transition to long-term opioid use, and represent a high-risk subgroup for opioid dependence, abuse and harms. However, definitions of long-term opioid use in the literature vary considerably, making it difficult to characterise the scale of the issue and design future interventions to address it [1,2] . Objectives (1) To evaluate the proportion of patients transitioning to long-term opioid use in new users across 6 RMD conditions using varying definitions used in the literature (2) To assess the proportion of long-term opioid users who transition to opioid dependence. Methods Patients aged ≥18 years with a diagnosis of rheumatoid arthritis (RA), psoriatic arthritis (PsA), axial spondyloarthritis (AxSpA), systemic lupus erythematosus (SLE), osteoarthritis (OA) and fibromyalgia and without prior cancer, with a new episode of opioid use between 01/01/2006 and 31/10/2021 and at least a 1-year follow-up in the Clinical Practice Research Datalink (CPRD) were included. CPRD is a database of anonymised UK primary care electronic health records representative of the national population. Long-term opioid users were defined using 3 different definitions: 1) Standard (most commonly used): ≥3 opioid prescriptions issued within a 90-day period, or ≥90 days opioid supply, in the first year of follow-up (excluding the first 30 days). 2) Stringent : ≥10 opioid prescriptions filled over >90 days, or ≥120-day opioid supply in the first-year follow-up. 3) Broad : ≥3 monthly prescriptions (no need to be consecutive) in the first 12 months. Opioid dependence was defined as RMD patients who had relevant Read Codes within 5 years after a new episode of opioid use. The proportions of long-term opioid use and opioid dependence for RMDs were calculated. Results This study included 841,047 patients of whom 12,260 had a code for RA, 5,195 PsA, 3,046 AxSpA, 3,081 SLE, 796,276 OA, and 21,189 fibromyalgia. The highest proportion of long-term opioid users among the 6 RMDs was patients with fibromyalgia (27.4% for Standard , 20.9% for Stringent , and 33.7% for Broad ), followed by RA (25.7%, 18.5%, and 32.3% respectively) and AxSpA (23.8%, 17.3%, and 29.6% respectively) (Figure 1-stacked bar chart). On average, using Broad definition showed 10-13% higher than Stringent definition for all RMDs. As the Venn diagram in Figure 1, 241,727 patients met any of the definitions of long-term use, of which half fulfilled all 3 definitions. The Broad definition was able to identify additional half of long-term users, with 24.0% overlapping with the Standard definition. In total, 685 (0.06%) RMD patients were diagnosed with opioid dependence within 5 years after starting opioids. Similar proportions of opioid dependence were observed in long-term opioid users across all definitions: 332 (0.18%) for Standard , 281 (0.23%) for Stringent , and 355 (0.15%) for Broad definitions. Moreover, 323 out of 685 RMD patients (47.2%) who had a diagnosis of opioid dependence were not classified as long-term opioid users by the 3 definitions. Conclusion Around 1 in 3 fibromyalgia patients and 1 in 4 RA/ AxSpA patients fulfilled definitions for long-term opioid use within 12 months after starting an opioid. The low prevalence of opioid dependence across all RMDs, defined using Read Codes alone is likely to be considerably underrepresented in clinical practice. This reflects both coding practises in primary care and under recognition of the issue in those on long-term opioids. References [1]Karmali, RN et al. Long-term opioid therapy definitions and predictors: A systematic review. Pharmacoepidemiol Drug Saf. 2020; 29: 252– 269. [2]Abdel Shaheed C et al. Rethinking “long term” opioid therapy BMJ 2019; 367:l6691. Figure 1 Long-term opioid users by definitions and overlap between them Acknowledgements This work was funded by a FOREUM grant (grant ID: 125059), MJ is funded through an NIHR Advanced Fellowship (NIHR301413). The authors would like to thank Ruth Costello and Ramiro Bravo for data management. Disclosure of Interests Joyce (Yun-Ting) Huang: None declared, David Jenkins: None declared, Belay Birlie Yimer: None declared, Carlos Ramirez Medina: None declared, Niels Peek: None declared, Mark Lunt: None declared, William Dixon Consultant of: WGD has received consultancy fees from Google unrelated to this work., Meghna Jani: None declared.