Selection of symptomatic patients with Crohn's disease for abdominopelvic computed tomography: Role of serum C-reactive protein

Alan N. Desmond, Kevin O'Regan, Neera Malik, Sebastian McWilliams, Siobhan O'Neill, Eamonn M. Quigley, Fergus Shanahan, Michael M. Maher

Research output: Contribution to journalArticlepeer-review

6 Scopus citations


Background: Results of previous studies have shown that repeated abdominopelvic computed tomography (CT) examinations can lead to substantial cumulative diagnostic radiation exposure in patients with Crohn's disease (CD). Improved selection of patients referred for CT will reduce unnecessary radiation exposure. This study examines if serum C-reactive protein (CRP) concentration predicts which symptomatic patients with CD are likely to have significant disease activity or disease complications (such as abscess) detected on abdominopelvic CT. Methods: All abdominopelvic CTs performed on patients with CD at a tertiary referral centre during the period June 2003 to June 2008 were identified. CT findings were coded by a pair of independent blinded senior radiologists for (i) small bowel luminal disease, (ii) large bowel luminal disease, (iii) mesenteric inflammatory changes, (iv) penetrating disease (fistulas, abscess, or phlegmon), (v) acute disease complications (obstruction or perforation), and (vi) acute non-CD findings. Imaging findings were correlated with serum CRP checked within 14 days before imaging. The reference range for CRP was defined as 0-5 mg/L. Results: A total of 147 patients with symptomatic CD had a CRP assay performed within 14 days before undergoing abdominopelvic CT. The median time from CRP assay to imaging was 2 days (interquartile range, 0-6 days). Median CRP before imaging was 24 mg/L (interquartile range, 6-88 mg/L). CT was normal in 34 of 147 case (23.1%). Patients with normal CRP (n = 36) were significantly less likely to have penetrating disease (odds ratio [OR], 0.04 [95% confidence interval {CI}, 0.01-0.7]; P <.001) or large bowel luminal disease (OR, 0.3 [95% CI, 0.1-0.8]; P <.05). Normal CRP excluded penetrating disease with a sensitivity of 1.0 (95% CI, 0.87-1.0). CRP levels did not correlate with the presence of small bowel luminal disease (n = 82), mesenteric inflammatory changes (n = 68), or acute disease complications (n = 10). Conclusion: Symptomatic patients with CD and normal serum CRP are unlikely to have evidence of abscess, fistulating disease, or large bowel luminal disease detected on abdominopelvic CT. However, abdominopelvic CT may demonstrate evidence of clinically significant non-penetrating CD or complications, including perforation and acute obstruction, regardless of serum CRP concentration.

Original languageEnglish (US)
Pages (from-to)267-274
Number of pages8
JournalCanadian Association of Radiologists Journal
Issue number4
StatePublished - Nov 2012


  • C-reactive protein
  • Computed tomography
  • Crohn's disease

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Selection of symptomatic patients with Crohn's disease for abdominopelvic computed tomography: Role of serum C-reactive protein'. Together they form a unique fingerprint.

Cite this