Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

The Comparative Incidence of Serious Infections among Rheumatoid Arthritis Patients Who Switch Biologic Agents.

Curtis3,  Jeffrey R., Xie2,  Fenglong, Chen2,  Lang, Baddley2,  John, Beukelman4,  Timothy, Saag5,  Kenneth G., Winthrop1,  Kevin

Univ of Alabama at Birmingham, Birmingham
University of Alabama– Birmingham, Birmingham, AL
University of Alabama-Birmingham, Birmingham, AL
University of Alabama-Birmingham, Birmingham, AL


The safety profile of the anti-TNF agents has been well characterized. However, it is unclear whether the comparative safety of newer biologics with different mechanisms of action is similar to the anti-TNF agents. Our objective was to evaluate the incidence of hospitalized infections among rheumatoid arthritis (RA) patients switching to a new biologic.


Using the administrative databases of a large U.S. healthcare organization from 1/2005-8/2009, we identified enrollees with RA based on having at least 1 physician visit for RA (ICD9 714.X). To control for confounding related to characteristics that cause patients to switch from one biologic to another, eligible patients were those switching to a new biologic who had been treated with a different biologic in the preceding 6 months, defined as the baseline period. Observation time began on the switch date. To further control for confounding, we created pair-wise propensity scores (PS) for each drug contrast using covariates measured at baseline. PS were categorized into quintiles; non-overlapping regions of the PS distributions were trimmed. Exposure was defined as current use of the biologic, allowing for an extension of 90 days. Outcomes of interest were hospitalized infections identified using previously-validated algorithms. Cox proportional hazards models evaluated the incidence rate of hospitalized infections, comparing each biologic to infliximab, selected as a convenient reference standard for uniform comparison. Observation time was censored at the first hospitalized infection, end of study, or lack of current exposure.


Among a total of 101,906 unique RA patients enrolled in the health plan, 1923 switched to a new biologic and contributed 1915 person-years (py) of observation. Characteristics of these eligible patients were mean ± SD age 49.3 ± 11.2 years, 78% women. A total of 82 first hospitalized infections were identified; the rate of hospitalized infections ranged from a low of 1.8 (switch to etanercept) to a high of 6.2 (switch to rituximab) per 100py. There was insufficient use of golimumab or certolizumab to evaluate these agents. Referent to infliximab, there were no significant differences between the rates of infection for any biologic after adjusting for age, sex, and PS quintile (Table). There were no cases of active tuberculosis or of hospitalized progressive multifocal leukoencephalocathy (PML).


The rate of all hospitalized infections among RA patients switching to a new biologic varies across a relatively small range and was not significantly different between agents.

Hazard Ratio Comparing the Rate of Serious Infections for Persons Switching to Various Biologies

Drug Switching ToIncidence Rate (per 100 pt-years)Crude Hazard RatioAdjusted for age, sex, propensity score
Abatacept5.81.08 (0.57, 2.04)0.87 (0.45, 1.70)
Adalimumab3.80.73 (0.40, 1.34)0.68 (0.36, 1.29)
Etanercept1.80.44 (0.19, 1.05)0.46 (0.19, 1.12)
Infliximab4.81.0 (ref)1.0 (ref)
Rituximab6.21.07 (0.56, 2.05)0.89 (0.43, 1.82)

To cite this abstract, please use the following information:
Curtis, Jeffrey R., Xie, Fenglong, Chen, Lang, Baddley, John, Beukelman, Timothy, Saag, Kenneth G., et al; The Comparative Incidence of Serious Infections among Rheumatoid Arthritis Patients Who Switch Biologic Agents. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :721
DOI: 10.1002/art.28489

Abstract Supplement

Meeting Menu