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.

Factors Associated with Upper and Lower Gastrointestinal Perforation in a Cohort of Patients with Rheumatoid Arthritis.

Curtis5,  Jeffrey R., Lanas4,  Angel, Werther2,  Winifred, John2,  Ani, Johnson1,  David A., Schulman3,  Kathy L.

Eastern Virginia Medical School, Norfolk, VA
Genentech, a Member of the Roche Group, South San Francisco, CA
Outcomes Research Solutions, Inc. Bolton, MA
Universidad de Zaragoza, Zaragoza, Spain
University of Alabama at Birmingham, Birmingham, AL


To evaluate the incidence of upper and lower gastrointestinal (GI) perforation and to identify associated risk factors in a cohort of patients (pts) with rheumatoid arthritis (RA).


In this retrospective, claims-based study (January 2001-June 2009), pts aged >18 were selected from the MarketScan® databases if they had at least 2 diagnoses of RA (ICD 9 CM 714.0, 714.3), between 30 and 365 days apart, and at least 1 year of continuous enrollment (defining "baseline period") prior to the beginning of follow-up time. Pts were excluded if they had a baseline history of hospitalized GI perforation or any GI malignancy. A validated algorithm for identifying hospitalized GI perforation in claims data was used to identify the first upper GI (UGI) and first lower GI (LGI) perforation. Pts were followed up until the date of GI perforation, database disenrollment, or study end. The GI perforation rate was reported in years of observed person time (PY). A Cox proportional hazards model was used to estimate the relative risk of GI perforation in the first 12 months of follow-up as a function of baseline demographics (age, gender, region, population density), comorbidities (Charlson Comorbidity Index), and use of RA and GI medications during the baseline period. These medications included NSAIDs, corticosteroids, biologic agents, methotrexate (MTX), and other DMARDs.


The study population included 143,433 RA patients (mean age [SD], 57.7 [14.1]; 74.8% female) with mean (SD) follow-up of 2.9 (1.9) years. Hospitalization with GI perforation occurred in 0.5% of pts (n=696), with 6.6% of patients dying during admission. At baseline, pts with GI perforation were older (57.6 vs. 62.0 years; p<.0001) than pts without perforation and had lower RA severity scores (p=.01) and higher Charlson Comorbidity scores (p<.01). Overall, the GI perforation rate per 1000 PY was 1.73 (95% CI, 1.61–1.86), with the majority of events concentrated in the LGI: 1.44 (1.32–1.55) vs. UGI 0.30 (0.24–0.35). The unadjusted GI perforation rate per 1000 PY was higher in pts aged >=65 years (2.3), with exposure to corticosteroids without accompanying NSAID (2.2), among pts with a pre-period history GI disturbance (3.4), specifically, diverticulitis (19.3), diverticulosis without diverticulitis (6.7), unspecified noninfectious gastroenteritis/colitis (4.9), esophageal or GI hemorrhage (4.2), ulcer (3.8), Crohn's disease (3.7), and ulcerative colitis (2.1). The adjusted relative risk of GI perforation associated with demographic and RA-related medications during the first year of follow-up is presented in the Table.


GI perforation is a rare but serious condition impacting patients with RA, most frequently in the LGI tract. The short-term risk of GI perforation increases with age and is highest among patients with a history of diverticulitis, diverticulosis, and exposure to corticosteroids and NSAIDs.

Table. Adjusted Relative Risk of GI Perforation by Significant (p<.05) Risk Factor

Baseline ParameterHazard Ratio% Confidence Limits
Age <40 Referent 
Age 40 to 642.601.215.58
Age 65 and older3.551.637.76
Exposure to NSAIDS & corticosteroids
No exposure to either NSAIDS or corticosteroids Referent 
Exposure to NSAID, no corticosteroid1.771.132.79
Exposure to corticosteroid, no NSAID2.231.453.45
Exposure to both NSAID & corticosteroid1.711.112.63
Exposure to MTX vs. no exposure to MTX0.750.570.99
Diverticulosis without diverticulitis4.261.7310.45

To cite this abstract, please use the following information:
Curtis, Jeffrey R., Lanas, Angel, Werther, Winifred, John, Ani, Johnson, David A., Schulman, Kathy L.; Factors Associated with Upper and Lower Gastrointestinal Perforation in a Cohort of Patients with Rheumatoid Arthritis. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1031
DOI: 10.1002/art.28798

Abstract Supplement

Meeting Menu