Mines D, Gu Y, Liu Q, Horne LN, Harrison MJ. Risk of active tuberculosis (TB) among biologic-naïve RA patients: retrospective cohort study in a US claims database. Presented at the 2008 Annual European Congress of Rheumatology; June 2008. Paris, France. [abstract] Ann Rheum Dis. 2008; 67(Supplement II):91.


BACKGROUND: Several epidemiologic studies suggest that the risk of active TB is elevated in patients with rheumatoid arthritis (RA) even in the absence of biologic therapy. However, the one study in the United States (US) that evaluated this issue found no association.1

OBJECTIVES: To compare the risk of TB among US patients with RA who have not received biologic therapy to that among patients without RA.

METHODS: We conducted a retrospective cohort study from 1998-2005 using the Pharmetrics Patient-Centric Database, which includes comprehensive health insurance claims for about 55 million people across the US. We assembled a cohort of adult RA patients based on two diagnosis codes for RA and dispensing of at least one conventional DMARD. For each RA patient we selected up to 20 patients without RA for a comparator cohort, matching on gender, age, health plan, calendar time and duration of enrollment. Patients who received treatment with a biologic agent (etanercept, infliximab, adalimumab, anakinra, abatacept, or rituximab) prior to cohort entry were excluded from the study, and patients who later received such treatment were censored. We identified cases of incident TB based on relevant diagnosis codes with concurrent claims for at least two specific anti-TB antibiotics, one of which needed to be isoniazid. We calculated incidence rates for each cohort, and using Poisson regression, calculated incidence rate ratios (RR).

RESULTS: The RA cohort consisted of 59,996 patients (75% female, median age 53 years), who were followed a mean 1.7 years. The non-RA cohort included 945,258 patients (76% female, median age 52 years) who were followed a mean 1.9 years. Nearly all study subjects had commercial insurance, and only 2.3% were Medicaid recipients. Main results appear in the Table. Comparing the RA cohort with the non-RA cohort, the RR of TB was 7.79 (95% CI: 2.40-25.3). Adjustment for age, gender, and calendar time led to no material change in the RR.

CONCLUSION: In this large population of US patients, nearly all of whom had commercial health insurance, the risk of TB among RA patients was about 7-fold higher than among patients without RA, even in the absence of biologic therapy.

Share on: