P275 Relevance of ultrasonographic parameters in predicting inflammatory bowel disease in a pediatric population.
A. Dell'Era1, D. Dilillo2, E. Galli2, F. Meneghin3, F. Furfaro1, C. Bezzio1, F. Penagini2, C. Mantegazza2, F. Scrignoli2, G.V. Zuccotti2, G. Maconi*1
1Ospedale L. Sacco. Gastroenterology Unit, Università degli Studi di Milano, Department of Biomedical and Clinical Sciences, Milan, Italy, 2Ospedale dei Bambini V. Buzzi, Università degli Studi di Milano, Department of Paediatrics, Milan, Italy, 3Ospedale L. Sacco. Università degli Studi di Milano, Department of Paediatrics, Milan, Italy
Bowel ultrasound (B-US) has been widely recognized as a useful examination in patients with suspected IBD, particularly in children, owing to its lack of invasiveness. However, its accuracy relies essentially on one criteria: the detection of increased bowel wall thickening (BWT). The relevance of BWT and the additional value of other US parameters, such as lymph node enlargement and mesenteric hypertrophy (MH), in the diagnosis of IBD have not been investigated so far. This study aims at investigating the diagnostic accuracy of several US parameters in detecting IBDs in a pediatric population.
All patients aged 2-18 years referred to the Pediatric Gastroenterology Clinic of our Hospital from 2007 to 2013 for initial assessment for recurrent abdominal pain and/or altered bowel habits were retrospectively considered.Patients presenting with known organic diseases or already investigated with digestive endoscopy were excluded. Patients were considered eligible if they had a complete B-US report including: altered US bowel pattern (US-BP), BWT, MH, pathologic lymph nodes, free abdominal fluid, presence of stenosis, abscesses or fistulae. Ileocolonoscopy, performed in patients with a high index of suspicion of IBD, on the basis of paediatrician's assessment and biochemical test Results (e.g. calprotectin, CRP)has been used as reference standard. Moreover, children who were not selected for endoscopy initially, were followed for at least one year for the appearance of possible additional symptoms.
Variables Se, % (95% CI) Sp, % (95% CI) PPV, % (95% CI) NPV, % (95% CI) Altered US-BP 78.3 (69.3–85.2) 93.3 (86.6–96.9) 75.0 (65.8–82.5) 94.4 (87.9–97.6) Mesenteric hypertrophy 65.2 (55.6–73.8) 92.2 (85.2–96.2) 68.2 (58.7–76.4) 91.2 (84.0–95.5) BWT > 3 mm 69.6 (60.1–77.7) 96.7 (90.9–99.0) 84.2 (75.9–90.1) 92.6 (85.6–96.4) BWT>3 mm + US-BP + MH 56.5 (46.9–65.7) 100 (95.9–100) 100 (95.9–100) 90.0 (82.6–94.6) BWT>3 mm or US-BP or MH 82.6 (74.1–88.9) 86.7 (78.7–92.1) 61.3 (51.6–70.2) 95.1 (88.9–98.1)
113 patients (mean age 10.8 years [range 2.1-17.7], 65 male) were enrolled. 23 IBD (20.4%; 8 ulcerative colitis, 12 Crohn's disease and 3 indeterminate colitis) were diagnosed. Among the bowel US variables considered, only US-BP, MH and BWT>3 mm were found useful to identify IBD on univariate binary logistic analysis. On multivariate analysis, these factors were independent predictors of IBD, even after adjustment for age and sex: US-BP (OR 9.8;95%CI 1.6-59.0); MH (OR 5.2;95%CI 1.1-25.1) and BWT>3 mm (OR 5.4;95%CI 0.7-40.1). Diagnostic accuracy of single US parameters and their combination, in distinguish between IBD and non IBD patients, is reported in table.
Among several US parameters suggestive of IBD, only the increased BWT, MH and altered echopattern are independent predictors of IBD and useful in distinguishing IBD from non-IBD patients. Owing to their high specificity and NPV, these parameters can be useful in identifying patients who did not need diagnostic invasive procedures in the short time.