Pain Medication and Corticosteroid Use in Ankylosing Spondylitis, Psoriatic Arthritis, and Rheumatoid Arthritis in the United States: A Retrospective Observational Study

Objective We compared pain medication use in patients with ankylosing spondylitis (AS), psoriatic arthritis (PsA), and rheumatoid arthritis (RA) versus matched control over 2 years; a subgroup analysis assessed changes in pain medication use in patients who initiated a biologic during 12 months before and after. Methods This was a retrospective observational cohort study using an administrative claims database. Newly diagnosed adult patients with AS, PsA, or RA identified between 1/1/2014 and 7/31/2017 were included. Demographics, baseline characteristics, and pain medication use were described using descriptive statistics. Differences in pain medication use were assessed using McNemar's/Wilcoxon signed-rank test for categorical/continuous variables. Results The study included 2180 AS, 5681 PsA, and 34,047 RA patients to assess overall pain medication use over 2 years; 188 AS, 921 PsA, and 1599 RA patients were included to assess changes in pain medication use 12 months before and after initiation of biologic. Demographics and baseline characteristics were balanced. In the overall cohort, 74.6% AS, 75.0% PsA, and 83.0% RA patients used any pain medication at baseline versus matched control; pain medications use 2 years after diagnosis date was reported in 73.5% AS, 74.1% PsA, and 81.3% RA patients. Among AS, PsA, and RA patients, use of prescribed NSAIDs (AS: 68.1 vs. 51.1%; PsA: 51.1 vs. 42.5%; RA: 61.1 vs. 41.5%; P < 0.05), glucocorticoids (AS: 56.4 vs. 41.5%; PsA: 57.4 vs. 46.9%; RA: 88.2 vs. 75.3%; P < 0.05), and opioids (AS: 42.6 vs. 36.2% [non-significant]; PsA: 38.1 vs. 33.8%; RA: 52.0 vs. 40.4%; P < 0.05) significantly decreased 12 months after biologic initiation versus prior. Conclusions Use of NSAIDs, glucocorticoids, and opioids are common among patients with AS, PsA, or RA, although the reported use of these co-medications after biologic initiation significantly decreases in the first year of treatment.