Background/Purpose: Glucocorticoids (GC) reduce RA-related disability and joint damage; RA guidelines endorse short term use during DMARD initiation and flares. Long-term high-dose GC exposure (>3 months, ≥10mg/day prednisone equivalent) can be toxic, but risk/benefit balance varies across patients. GC exposure patterns and user characteristics have not been described in a national commercially insured cohort. Provider-level factors may affect patterns of GC use in RA, and may interact with patient factors to alter risk/benefit balance. We hypothesize that GC are commonly used for incident RA and are continued inappropriately once DMARD treatment is established, and that wide provider-level variation in GC use exists.

Methods: Using OptumInsight^TM commercial claims data, we identified 9,221 adults with incident RA diagnosed 2010-2014 and ≥12 months of preceding medical and pharmacy benefits. We assessed GC exposure for 3 months before and 12 months following diagnosis (“study period”), cumulatively and stratified by 3 month quarter and GC prescriber specialty (rheumatologist, primary care provider, other). We examined variation among 117 rheumatologists by dividing per-patient distribution of GC dose and duration for each quarter into quartiles.

Results:   6,717 RA patients (73%) received GC during the study period. There were no clinically important differences in demographics or baseline health status between GC users and non-users, or by cumulative GC exposure level. GC use rose with frequency of rheumatologist visits and number of DMARD prescriptions. 76% of patients filled a DMARD prescription in quarter 1, and 17% received a biologic DMARD by quarter 4 (Table 1).

During quarter 1, 53% of patients received GC with per-patient mean daily dose 15mg prednisone equivalent/day and duration 57 days. During quarter 4, 29% of patients received GC with per-patient mean daily dose 14mg/day and duration 48 days. Rheumatologists prescribed >60% of all dispensed GC, with mean daily dose 11-18mg and duration 30-60 days per quarter (Table 1). Per-patient dose and duration of GC prescribed vary widely at the provider level over the study period (Fig. 1).

Conclusion: In this commercially insured incident RA cohort, rheumatologists commonly prescribe long term high dose GC up to 1 year after RA diagnosis, despite appropriate DMARD and biologic use. Rheumatologist practices regarding GC use for RA vary widely. Further work is needed to evaluate the relationship between 1) specific patient and provider-level factors and GC exposure, and 2) GC exposure and DMARD initiation and persistence in this population.