Gastrointestinal ultrasound (GIUS) is a non-invasive imaging modality capable of detecting intestinal inflammation & associated complications. It has comparable sensitivity & specificity to magnetic resonance enterography (MRE) in detecting ileocolonic disease, however it is less expensive (£24 vs £180) & can be performed at point of care.
We aimed to establish the proportion of MREs that could have been performed as GIUS at a tertiary inflammatory bowel disease (IBD) unit, the potential cost savings, & the predicted pathology miss-rates.
All MREs performed in January 2018 were retrospectively reviewed. Demographics, scan indication, IBD characteristics, surgical history, & gastrointestinal & non-gastrointestinal findings were collected. Indications deemed suitable for GIUS included: assessment of disease activity of known small bowel (SB) Crohn’s disease; first assessment for presence of SB disease in IBD; & investigation for SB disease in patients without a known diagnosis of IBD. Obesity, complicated surgical history (>1 resection or strictureplasty involving different segments, or stoma), & known proximal SB disease were deemed unsuitable.
105 MREs were performed in January 2018. 59 (56%) were deemed suitable for GIUS instead of MRE. Most common reasons for unsuitability included complex surgical history (n=17, 37%), obesity (n=14, 30%), non-appropriate indication (n=12, 26%) & known upper gastrointestinal disease (n=10, 22%).
Of suitable cases, 32/59 (54%) had active inflammation detected including 17 (53%) isolated ileal, 8 (25%) ileocolonic, & 6 (19%) isolated colonic. In one case performed as first assessment for SB disease, both ileal & jejunal disease were found, the latter likely to be missed with GIUS. No cases of isolated upper gastrointestinal inflammation were found. Regarding non-gastrointestinal findings in potential GIUS patients, there were two cases of pancreatic cysts necessitating further investigation with serial MRIs & endoscopic ultrasound, yielding a side branch intraductal papillary mucinous neoplasm & a benign serous cyst adenoma. One case of multiple high T2 skeletal lesions was deemed clinically insignificant following further investigations. No other significant extra-intestinal findings not expected to be seen on GIUS were identified.
Over 50% of MREs could have been performed as GIUS instead, with a potential annual cost saving of over £110,000. No instances of inflammation would have been missed based on distribution, although in one case the full extent of disease may not have been identified on GIUS. Incidental non-gastrointestinal findings resulted in multiple investigations but were of limited clinical significance.