High Definition Colonoscopy (HDC) vs. Dye Spraying Chromo-colonoscopy (DSC) in Screening Patients With Long-standing Inflammatory Bowel Disease (IBD)

Research Question:

Is High Definition White Light Colonoscopy (HDWLC) not inferior to Dye Spraying Chromo-colonoscopy (DSC) in detecting all neoplastic lesions during screening patients with colonic IBD for more than eight years?

Study setting: The study will be conducted at Beth Israel Deaconess Medical Center. Patients will be randomized to HDWLC with biopsies every 10 cm versus DSC.

Study population: Adult patients colonic IBD for more than eight years.

Recruitment The study team will conduct a day by day chart review for all patients with a history of inflammatory bowel disease who were scheduled for screening colonoscopy and find candidates who meet the eligibility criteria. A study team member will meet with the potential candidates during their visit to the gastroenterology procedure room before having their colonoscopy and offer them to participate in the study.

Procedures: Participants will undergo either HDWLC with biopsies every 10 cm or DSC based on the randomization. Both procedures are standard of care and are being done at BIDMC before starting this study. All the endoscopists are IBD specialists with expertise in both procedures. All physicians were given a review article and video of chromoendoscopy to further standardize practice. The clinicians who are conducting the HDWLC are allowed to do a targeted DSC if required as per hospital protocol. Both arms of the study are using high definition Olympus colonoscopes

Study design: A randomized controlled, noninferiority trial.

Sample size:

Existing literatures suggest that the detection rate of the DSC arm is 20%. The investigators set 10% as the non-inferiority margin. With a total of 400 subjects (200 in each arm), the study will have 80% power to detect non-inferiority with the type I error rate controlled at 0.05. To account for a potential 25% data attribution, the study team will enroll 500 subjects.

Statistical Analysis:

The baseline characteristics of the two trial arms will first be tabulated for potential imbalance in variables. Continuous variables will be summarized by typical parameters such as mean, standard deviation and range and compared using two-sample T-test (if the normality assumption holds) or Wilcoxon rank-sum test (if the normality assumption does not hold). Normality of distribution will be determined using the Kolmogorov-Smirnov goodness-of-fit test. Categorical data will be summarized by frequency and percentage and analyzed using the Chi-square or Fishers exact test, as appropriate.

The primary outcome will be compared using Z-test based on normal approximation of the sample proportions.

Outcome analysis:

Primary outcome analysis: Intention to treat Secondary outcome analysis: Per protocol analysis.