Effective communication of cross-sectional imaging findings in Crohn's disease: comparing conventional EMR reporting...

The purpose of the article is to compare information regarding small bowel lesions in Crohn's disease (CD) patients communicated by a published scoring system and radiology reports from electronic medical record (EMR) of cross-sectional abdominal imaging.

Two gastrointestinal radiologists (reference readers) blinded to EMR reports scored cross-sectional imaging exams using a published scoring system. Investigators compared EMR and radiologist scores based on the mentioned findings and severity documentation of each variable. Statistical analysis involved means and difference in proportions and logistic regression modeling.

Seventy-three CD patients, with average age 40.6 years (± SD 14.4), having 80 small bowel lesions on imaging were included. EMR reports reliably mentioned within the consensus score included thickness (79%, p = 0.000), enhancement (70%, p = 0.000), active inflammation (86%, p = 0.000), perienteric fluid (82%, p = 0.000), and presence of stricture (62%, p = 0.002). Minimal lumen diameter (19%, p = 0.000), comb sign (19%, p = 0.000), lesion length (57%, p = 0.06), and fistula (50%, p = 1.0) were reported less often. There was a strong association between the EMR and scoring scale in noting severity of active inflammation (88%, p = 0.000), perienteric fluid (76%, p = 0.000), and internal fistula (71%, p = 0.000). The proportion matching severity values of comb sign and minimal lumen were 24% and 21%, respectively (p = 0.000). Severity matches for stricture were less likely among the non-GI radiologists (odds ratio = 0.33, SE = 0.168, p = 0.029). The odds of reporting stricture and fistula severity were 3.6 and 5.7, respectively, on MRE.

Findings and severity of inflammation were communicated consistently. Stricture severity including minimal luminal diameter, was less reliably reported, though its prognostic significance impacts management.