We searched endoscopy reports to identify all patients referred to the gastroenterology department at the Hillel-Yaffe medical center, a university affiliated hospital in Israel, for ileocolonoscopy in the years 2010-2016 as part of a diagnostic work-up for a suspected IBD. These patients suffered from different clinical presentations such as chronic abdominal pain, diarrhea and anal complaints deemed by the referring gastroenterologist as suspicious for IBD. Patients were included if they had normal ileocolonoscopy provided that TI biopsies were performed. Normal ileocolonoscopy denotes to findings of normal colon and TI mucosa. mild nodularity characteristic of lymphoid hyperplasia was also considered normal. Findings of erythema, friability, granularity, erosions, ulcers or strictures were considered abnormal and were not included. Moreover, only patients with a full data set, including demographic details (age, sex), indication for exam, endoscopic findings, and available biopsy results were included in the final analysis. Patients with prior colonoscopy or IBD diagnosis, as well as patients who had undergone prior colon resection were excluded.
Our staff included 6 senior endoscopists (more than 15 years experience, with more than 500 annual colonoscopies) who performed these ileo-colonoscopy procedures or directly supervised their performance by trainees to ensure visualization and assessment of terminal ileum mucosa. More than 95% of patients included in the study had adequate bowel preparation (some patients were included after repeated exams due to poor preparation).
According to the biopsy results, patients were divided into normal and MI groups (figure 1). Normal group included patients with normal biopsy or reactive and nonspecific changes, while MI group included findings of active ileitis (normal villous architecture and polymorphonuclear or mild to moderate eosinophilic infiltrate), chronic active ileitis (distorted villous architecture; mixed inflammatory infiltrates of any severity) and granulomatous inflammation (non-caseating granuloma formation, no evidence of fungi or parasites). In order to determine the long term clinical outcome in both groups we reviewed electronic files of clinic, imaging and endoscopic reports available in our department. For patients who continued follow-up elsewhere, we used national electronic system that enables access to patients' reports and clinical data, though with usually restricted access to all data resources. Some of these patients were invited for clinical assessment in our department. The follow-up was usually discontinued once a new diagnosis of Crohn's disease was confirmed or with the last report available by the time this study has started.
Crohn's diagnosis was counted when expert gastroenterologist clearly confirm that the patient underwent subsequent follow up examination including recurrent ileo-colonoscopy, capsule endoscopy and/or bowel imaging procedures with new findings that meet the diagnostic criteria for Crohn's disease. We performed also multivariate analysis to identify independent association of age, sex and MI with crohn's disease development. In the MI group we classified patients according to inflammation severity in order to evaluate histological correlation with final outcome.
The local institutional Helsinki ethics board in Hillel Yaffe medical center approved the study (reference number 0112-17) and granted exemption from informed consent as patients were receiving standard care without relation to the study. Data collection did not influence medical practice.
Statistical analysis:
Continues parameters were presented by means ±standard deviations, and categorical parameters were expressed by using frequencies and percentages. Differences between the MI and normal groups were compared by t-test in quantitative parameters, and Fisher's exact test in the categorical parameters. Multivariate analysis by logistic regression were used for prediction of subsequent Crohn's disease diagnosis by age, gender and group. P<0.05 was considered as significant. SPSS version 25 was used for the all statistical analysis.