Bowel Sonography and MR Enterography in Children

The objective of this study was to assess the diagnostic agreement between bowel sonography and MR enterography in children with suspected or confirmed inflammatory bowel disease (IBD).

We prospectively studied 33 children undergoing clinically indicated MR enterography. Bowel sonography was performed within 2 hours before MR enterography. The reliability of each modality was assessed by computing the interreader kappa agreement on nine inflammatory markers in eight bowel segments and the interreader intraclass correlation coefficients (ICC) in each bowel segment. The agreement between both modalities in each bowel segment was also calculated using the Kendall coefficient of concordance (KCC) in addition to the sensitivity and specificity of sonography using those of MR enterography as the reference standard.

Sonographic readings are substantially reliable in the colon and terminal and distal ileum (ICC = 0.79-0.88) but are highly unreliable in the mid and proximal small bowel. MR enterography has moderate to almost perfect reliability for all bowel segments (ICC = 0.43-0.96). Sonography was in substantial agreement with MR enterography in most of the large bowel (KCC = 0.67-0.83) and terminal and distal ileum (KCC = 0.75) and in moderate agreement in the remaining bowel segments (KCC = 0.45-0.48). Sonographic findings are specific for disease using MR enterography as the reference standard in all bowel segments (0.83-1.00), but sensitivity markedly varies according to disease location.

Sonography is highly specific and shows substantial agreement with MR enterography in detecting inflammatory and damage changes in the colon and terminal and distal ileum, the segments most likely to be affected by IBD. Therefore, sonography can be a good complement to MR enterography in evaluations of patients with known or suspected IBD, particularly if MR enterography access is limited.