Background & aims: Systemic corticosteroid therapy increases risk of postoperative sepsis in Crohn's disease. This study investigates its effect on risk for sepsis in non-operated patients.
Methods: A retrospective case-control study was performed in 432 patients with Crohn's disease (the 94% of our database for whom adequate documentation could be retrieved). Two analyses were performed. The first tested the hypothesis that patients with perforating Crohn's disease (n = 86) were more likely to develop intra-abdominal or pelvic abscess (n = 29) if they had received systemic corticosteroids during the previous 3 months. The second analysis, confined to interventions since 1998, tested the hypothesis that corticosteroid therapy was more common during the 3 months before presentation with intra-abdominal or pelvic abscess (n = 12) than during the 3 months after presentation with a relapse of nonperforating disease (n = 24 consecutive patients). In both analyses adjustment was made for any other significant variable.
Results: Systemic corticosteroid therapy was associated with an adjusted odds ratio (OR) for intra-abdominal or pelvic abscess of 9.03 (95% confidence interval [CI], 2.40-33.98) in patients with perforating Crohn's disease. Patients receiving prednisolone > or = 20 mg per day had an OR of 2.81 (95% CI, 0.99-7.99) for abscess compared with those receiving a lower dose. In patients with relapsed active disease, corticosteroid therapy was associated with an unadjusted OR of 9.31 (95% CI, 1.03-83.91) for intra-abdominal or pelvic abscess. Neither smoking nor azathioprine usage was associated with increased risk for abscess.
Conclusions: Systemic corticosteroid therapy for Crohn's disease is associated with increased risk for intra-abdominal or pelvic abscess.