To study the association between infection risk and low-dose corticosteroids (LD-CT, defined as a daily dose <10mg/day of prednisone) in rheumatoid arthritis (RA).

Data source: a systematic review of the literature up to June 2009 was performed. Data extraction :all type of infections: bacterial, viral and postoperative; infection severity, RA activity, RA severity, comorbid conditions. Data analysis: descriptive, comparing infection risk between LD-CT-treated and LD-CT-not treated RA.

Of the 1310 screened reports, the literature analysis identified 15 assessing infection risk of LD-CT in RA patients. Of the eight reports that studied all types of infection, six articles found no association between risk of infection and LD-CT, one showed an association between severe infections and LD-CT (OR=8 [1–64]) and another showed a dose-dependent association including doses of less than 5mg/day: RR=1.32 [1.06–1.63] and doses between 6 to 10mg/day: RR=1.95 [1.53–2.46]. Of the three trials that studied infection risk secondary to bacteria, one showed an increased risk (HR=1.7 [1.5–2.0]) while two did not (respectively, exposure to <5mg/day: OR=1.34 [0.85–2.13]; 6 to 9mg/day: OR=1.53 [0.95–2.48] and <5mg/day: OR=1.49 [0.82–2.72]; 5 to 10mg/day: OR=1.46 [0.84–2.54]). None of the three trials studying postoperative infection risk found any association between infection risk and LD-CT treatment. Two reports studied herpes zoster risk and found no association with LD-CT.

There was a paucity of data about LD-CT and infection risk in RA and that risk seems poorly increased. These findings need to be confirmed by further studies.