To the Editors:

We read with interest the recently published article in Arthritis Care & Research by Constantinescu and colleagues regarding race and preferences for choosing treatments for rheumatoid arthritis (RA) (1). Several studies have shown an influence of factors such as race, education, and socioeconomic status on disease outcome in patients with RA (2, 3). Minorities may also underutilize treatment resources, further contributing to poorer disease outcomes. The introduction of highly effective therapeutic agents, such as tumor necrosis factor inhibitors (TNF-I), has dramatically altered the treatment approach to RA and improved outcomes for individual patients (4).

Using the Consortium of Rheumatology Researchers of North America (CORRONA) database, we sought to assess the extent to which race/ethnicity, education, insurance type, and disease activity influenced RA patients' likelihood of receiving TNF-I. The CORRONA database is a prospective registry that gathers information on demographics, disease characteristics, and therapies of both US and Canadian patients with various rheumatologic conditions (5). The characteristics of our cohort of RA patients were as follows: 75% were women, and the mean ± SD age was 60 ± 13.4 year; 88% were white, 7% were Hispanic, 4% were African American, and 1% were Asian. The mean ± SD disease duration was 11.6 ± 9.9 years, and maximal education levels were 49% for primary/secondary school and 51% for college. The majority of patients had some form of medical insurance (73% private, 42% Medicare, and 6% Medicaid).

Compared with those with less than a college education, college graduates had lower disease activity (Disease Activity Score in 28 joints [DAS28] 3.3 versus 3.8; swollen joint count 3.9 versus 4.7; visual analog scale [VAS] pain score 29 versus 34; and erythrocyte sedimentation rate 22 versus 26; all P < 0.01). Compared with whites, Hispanics had higher disease activity (DAS28 4.0 versus 3.5; tender joint count 5.2 versus 3.4; VAS pain score 34 versus 31; all P < 0.01), but were less likely to receive TNF-I (36.3% versus 42.3%; odds ratio [OR] 0.78). The use of TNF-I was comparable among whites, African Americans, and Asians. Compared with those with only primary/secondary school education, college graduates were more likely to receive therapy with TNF-I (45.4% versus 38.4%; OR 1.33). Even after adjustment for education, age, sex, disease duration, and disease activity, Hispanics remained less likely to receive treatment with TNF-I (OR 0.71, 95% confidence interval [95% CI] 0.58–0.89). Patients with private insurance were more likely to receive TNF-I (OR 1.22), but Hispanics continued to receive less TNF-I after adjusting for types of insurance (OR 0.76, 95% CI 0.61–0.95). Understanding these factors can help us target certain subgroups with a more optimal therapeutic approach in an attempt to prevent potentially disabling joint disease and its sequelae.

Susan J. Lee MD*, Joel Kremer MD†, Arthur Kavanaugh MD‡, * University of California, San Diego, † Albany Medical College, Albany, New York, ‡ University of California, San Diego.