Fig. 1.

Study design. The nest population consists of patients who had a prescription for second-line T2DM therapy. Cases were those with two or more separate diagnoses of RA and a prescription event for RA drugs. Matched controls were selected from the nest who did not have a diagnosis of RA when the case developed. Matching on age (± 5 years), sex, index date (± 90 days), and event date (lag window is equal between cases and controls). RA, rheumatoid arthritis; index date, date of first prescription for second-line T2DM therapy; event date, first RA diagnosis post index date; lag window, index date–event date