P219 The relevance of ileitis as diagnosed by capsule endoscopy: A comparison with double balloon enteroscopy
B. Hall*, G. Holleran, D. Brennan, D. McNamara
Adelaide and Meath Hospital, Gastroenterology, Dublin, Ireland
Capsule endoscopy (SBCE) is non-invasive and allows a complete view of the small bowel in the majority of cases. A drawback to SBCE remains the inability to obtain biopsies and the relevance of small bowel inflammation as detected by SBCE has been questioned. Current guidelines would suggest enteroscopy (DBE) with histological assessment should be performed when ileitis is detected on SBCE. This study aimed to determine the clinical accuracy of SBCE in diagnosing small bowel CD as compared to DBE.
Patients with evidence of ileitis on SBCE between June 2010 and July 2013 were identified from our database. Importantly, record was made of whether the capsule reader considered the ileitis to be consistent with CD in each case. All patients with evidence of ileitis were cross-referenced using our endoscopy database to identify patients who had undergone DBE at our institution. A retrospective chart review was undertaken to ascertain final diagnosis. Exclusion criteria included any documentation of NSAID use in the 3 months prior to SBCE and any patient with less than 6 months follow up post procedure.
A total of 820 SBCE procedures were performed during the three-year time-period. Approximately 140 (17%) had documentation of ileitis on SBCE. Of these, data was available on 22 (16%) patients had undergone DBE; mean age 45 years (range 18-73), 12 (55%) male with a mean follow up of 7 months (range 6-18). Table 1 demonstrates the findings of both SBCE and DBE. There was a statistically significant, albeit weak, correlation between SBCE and DBE findings (R=0.516, p<0.013). In terms of final diagnosis, 8 (36%) patients were subsequently diagnosed with definitive small bowel CD. Final diagnosis was based on numerous factors including the SBCE, DBE and histological findings coupled with radiological investigations and clinical symptoms. Of the remaining 14 patients, 8 (36%) were diagnosed with functional bowel disease and 6 (27%) with NSAID enteritis. All patients who were eventually diagnosed with CD had abnormalities consistent with small bowel CD on SBCE. SBCE had a moderate degree of correlation with final diagnosis (R=0.638, p<0.001). The positive and negative predictive value for SBCE was 70% and 92%, respectively based on final diagnosis.
SBCE is effective at detecting ileitis. It has a high positive and negative predicitve value although there will always be a tendancy towards a poorer positive predictive value until capsules are developed which are capable of interacting with their environment. A high index of suspicion is required for alternated diagnoses and DBE should be performed if there are doubts about the likely diagnosis.