O28 Has modern UK management of RA led to a reduction in use of steroids and NSAIDs? 20-year data from the clinical practice research datalink

Abstract

Abstract Background Over 20 years, rheumatoid arthritis (RA) management has evolved to immediate initiation and more effective escalation of disease-modifying anti-rheumatic drugs (DMARDs) to suppress inflammation. We aimed to establish whether this has facilitated reduced long-term use of potentially toxic glucocorticosteroids (GCs) and nonsteroidal anti-inflammatory drugs (NSAIDs). Methods We investigated changes in DMARD, oral GC and NSAID prescribing over 1998-2017 in a UK primary care electronic health records database. We defined RA by coded diagnosis, sensitivity analyses requiring (a) two diagnoses ≥6 months apart; (b) ≥1 DMARD prescriptions. We calculated annual RA incidence and prevalence, annual prescribing and prescribing in the first to fifteenth year from diagnosis. Long-term prescribing was defined as ≥ 3 months (≥6 in a sensitivity analyses) and compared with a non-RA cohort (5:1 matched on age, sex and GP practice). Results We included 71,411 RA patients (44,426 with 2 diagnoses, 45,438 with DMARD). Incidence declined from 5.01 (±0.36) per 10,000 person-years in 1998 to 4.77 (±0.23) in 2011, before increasing following changes in Quality Outcomes Framework indicators. Prevalence rose from 0.70% (±0.013) in 1998 to 0.91% (±0.014) in 2017 and was highest among patients aged ≥70 (2.21%, ±0.05). Long-term DMARD prescribing rose from 30.97% in 1998 to 49.28% in 2017. Long-term NSAID prescribing fell (45.94% to 25.08%) particularly amongst incident RA patients, with only modest change in GCs (20.98% to 15.53%) (Table 1). By comparison, long-term GC prescribing in the matched non-RA population rose (0.90% to 2.01%) and NSAID prescribing changed little (6.37% to 8.41%). The proportion with long-term prescribing declined slightly from the first to the third year following diagnosis for GCs (22.18% to 17.87%) and NSAIDs (41.19% to 34.30%). Sensitivity analyses showed similar results. Conclusion Long-term GC prescribing among newly diagnosed RA patients has not appreciably reduced despite modern treatment paradigms and the expected observed increase in DMARD prescribing. Long-term NSAID prescribing did decrease in primary care for RA (especially from mid-2000 and among newly diagnosed patients), but interestingly, not for non-RA patients. Given the scope for further reduction in GC and NSAID prescribing, the reasons for persistent prescribing, including prescribing behaviours, require further investigation. Disclosures S.S.R. Crossfield: Grants/research support; Medical Research Council (MRC) Leeds Medical Bioinformatics Centre [MR/L01629X]. M.H. Buch: None. P. Baxter: None. S.R. Kingsbury: Grants/research support; National Institute for Health Research (NIHR) Leeds Biomedical Research Centre and the Versus Arthritis Experimental Osteoarthritis Treatment Centre [20083]. M. Pujades-Rodriguez: None. P.G. Conaghan: Grants/research support; National Institute for Health Research (NIHR) Leeds Biomedical Research Centre and the Versus Arthritis Experimental Osteoarthritis Treatment Centre [20083].