Quantified motility assessment in paediatric inflammatory bowel disease: a multi-reader study

Quantitative bowel motility in inflammatory bowel disease using magnetic resonance enterography (MRE) shows promise as a biomarker of disease activity against histopathological and biochemical markers of inflammation when used by experienced readers in a research setting.

The aim of this study was to test multi-reader agreement in quantified bowel motility by non-experienced readers.

A single-centre, retrospective search included all patients <18 years with paediatric inflammatory bowel disease and more than one MRE between 2012-2022. Cine images from the first two MREs were processed using quantified motility software. Two experienced readers provided a reference standard for response using a consensus sMaRIA (decrease of ≥2 points) and measured the change (Δ) in motility index. Two senior and three junior readers visually assessed response for each patient and recorded Δ motility index. Inter-reader agreement was assessed for each pair of readers as the Bland-Altman bias and limits of agreement for Δ motility index (arbitrary units, a.u.) and %agreement in the Δdirection.

Percentage agreement in the Δdirection was fair to excellent. Junior reader agreement with experienced readers was 73-84%, and senior reader agreement with experienced readers was 67-85%. Bland-Altman plots of Δ motility index demonstrated no systematic bias and the junior-experienced readers' limits of agreement were acceptable.

Quantified bowel motility in paediatric inflammatory bowel disease measured by readers of varied experience yields good agreement.