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 Influence of Systemic Glucocorticoid Therapy upon the Risk of Non-Serious Infection in Patients with Rheumatoid Arthritis.

Dixon3,  William G., Kezouh2,  Abbas, Bernatsky1,  Sasha R., Suissa2,  Samy

McGill UHC/RVH, Montreal, QC, Canada
McGill University
The University of Manchester, Manchester, United Kingdom


Glucocorticoid (GC) therapy is a strong risk factor for serious infections in patients with RA. However, the association between GCs and non-serious infections (NSI) is not well studied. Although NSIs are not life-threatening, they are common: respiratory infections alone account for up to 400 general practice consultations annually per 1000 patients. Even a modest increase in relative risk due to GCs may represent a large increase in the attributable risk and a significant health burden.


Using data from Quebec health administrative databases from 1985–2003, we assembled a cohort of 16,207 patients with RA aged >65. GC and DMARD therapy were identified from drug dispensing records. NSI cases were defined as the first occurrence of either a community physician billing code for infection or community-dispensed anti-infectives. The incidence of NSI was estimated in the whole cohort. The incidence of NSI in GC-exposed and GC-non-exposed person time was calculated in a cohort analysis to enable estimation of the attributable risk. A nested case-control analysis was performed, matching each case to 5 controls. Matching was done on entry date and time in cohort, using risk set sampling. Analysis considered drugs dispensed within 45 days prior to the index date (date of infection for each case-control set), adjusting for age, sex, markers of disease severity, DMARDs and co-morbidity. Oral GC therapy was considered together, then categorised into 5 dose bands. Conditional logistic regression was used to calculate the odds ratio, interpretable as a relative risk.


13,634 first-episode NSI occurred during 28,695 person years (pyrs), generating an incidence rate of 475 events/ 1000 pyrs. The rate of NSI in GC exposed and unexposed person time was 524 and 388/1000 pyrs, respectively. The attributable risk was therefore 135 events/ 1000 pyrs. In the case-control analysis, GC therapy was associated with an adjusted relative risk (aRR) of 1.20 (95% CI 1.15, 1.25). A positive dose response was seen between risk and GC dose. Methotrexate (MTX) was the most commonly used DMARD, and was associated with a no increased risk of NSI (aRR 1.00 (0.95, 1.04)). Patients currently prescribed sulfasalazine or anti-malarials had a lower rate of infection than patients not prescribed those drugs. Cyclophosphamide was associated with a higher risk of infection (aRR 2.14 (1.51, 3.03)). The results for anti-TNF therapy were inconclusive given the small number of patients exposed to them. All GC risk estimates (including 0–5mg/day) were higher than that seen for MTX.

 Cases n = 13,634Controls n = 68,170Adjusted RR* (95% CI)
Oral GC exposure within last 45 days37.932.51.20 (1.15, 1.25)
Average daily dose of oral GC therapy   
  • <5 mg PEQ3.23.11.10 (0.99, 1.22)
  • 5–9.9 mg PEQ17.817.11.10 (1.04, 1.16)
  • 10–14.9 mg PEQ9.57.91.25 (1.17, 1.34)
  • 15–19.9 mg PEQ2.72.11.26 (1.12, 1.42)
  • >=20 mg PEQ4.72.31.85 (1.68, 2.05)
Any GC: oral or injection39.233.71.19 (1.14, 1.24)
Current DMARD use   
  • Methotrexate33.333.01.00 (0.95, 1.04)
  • Sulfasalazine2.43.00.79 (0.70, 0.89)
  • Leflunomide0.30.31.00 (0.71, 1.39)
  • CQ/HCQ29.430.70.93 (0.89, 0.98)
  • Azathioprine1.91.51.05 (0.91, 1.22)
  • Cyclophosphamide1.70.72.14 (1.51, 3.03)
  • Gold7.16.51.08 (0.99, 1.18)
  • Anti-TNF therapy0.10.11.48 (0.87, 2.52)
  • Others** (0.92, 1.25)
GC = glucocorticoid, PEQ = prednisolone equivalent, RR = relative risk*Adjusted for age, sex, markers of disease severity, and co-morbidity**Includes ciclosporin, mycophenolate mofetil, and D-penicillamine


GC therapy is associated with an increased risk of NSI. The magnitude of the risk increases with dose, and is higher than that seen with MTX (although residual confounding may exist). Whilst the RR is low at 1.20, the absolute risk is high, with one additional infection seen for every 8 patients treated for 1 year.

To cite this abstract, please use the following information:
Dixon, William G., Kezouh, Abbas, Bernatsky, Sasha R., Suissa, Samy; The Influence of Systemic Glucocorticoid Therapy upon the Risk of Non-Serious Infection in Patients with Rheumatoid Arthritis. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :342
DOI: 10.1002/art.28111

Abstract Supplement

Meeting Menu