Overall Objective: To optimize timing of surveillance colonoscopy.

Principal research question and specific aims: To assess the impact of access to a hand-held application on the timing of surveillance colonoscopy. To assess whether access to the tool improves adherence to recommended guidelines for follow-up colonoscopy intervals.

Colonoscopy is commonly used for surveillance of patients with high risk of developing colorectal cancer, including those with family history of colorectal cancer and those with colorectal polyps. The recommended timing of surveillance colonoscopy varies by the estimated risk for development of colorectal cancer. The estimated risk varies by family history of colorectal cancer (number of affected individuals, age of the persons affected with CRC) and characteristics of the colorectal polyps (size, number, and histology of colorectal polyps (tubular or villous; high grade or low-grade dysplasia; sessile serrated polyp, sessile serrated polyp with dysplasia, hyperplastic polyp or traditional serrated adenomas). Guidelines take all of these factors into account in the recommendations for follow-up colonoscopy and hence are difficult to recall for the busy clinicians. Colonoscopy surveillance is frequently performed at shorter or longer than the recommended time intervals. The investigators have developed a smart phone application in which the characteristics of the patients can be inputted and the tool provides the recommended time interval for surveillance colonoscopy, based on North American guidelines. The investigators are proposing a pilot randomized trial to determine sample size estimates for a larger trial to assess the utility of this application in clinical practice.

Study design and intervention: This will be a cluster randomized controlled trial in which individual endoscopy physicians (gastroenterologists and surgeons) will be randomized (stratified by physician specialty) to one of two groups: A) provided access to the application (intervention group) or B) not provided access to the application (control group). In the intervention group, the application can be downloaded to smart phones for portability which will allow access in an endoscopy suite or in clinic or used as a reference at other times. It can also be accessed online (all endoscopy rooms in Winnipeg have computers with internet access for the endoscopy physicians' use). The clusters of patients will be defined by the endoscopy physician providing the colonoscopy.

Intervention: Access to the application which will be available both as a web-based as well as a smart phone application.

Setting: Colonoscopies included in the study will be those performed in the Winnipeg endoscopy units.

A research assistant will extract data from medical records maintained by the hospitals and physicians privately. Individual cases of colonoscopies will be identified through the city-wide endoscopy program database. Study data will be collected and managed using the REDCap (Research Electronic Data Capture) electronic data capture tools hosted at the University of Manitoba.

Concealment of Allocation and Randomization: Randomization will be performed after the endoscopy physician consent and enrollment into the study. A computer generated randomization sequence will used in blocks of eight. Stratification for surgical and gastroenterology endoscopy physician will be performed as there might be differential adherence to recommendations by speciality of training.

Blinding: Randomization will be performed and the application distributed by study personnel different than those involved in data extraction from medical records so that the person extracting the data remains blinded to the randomized group assignment.

Patient consent: Patients will not be approached by the study investigators and will get routine care decided by their physicians. Information on recommendations will be extracted from medical records retrospectively; hence, patient consent will not be required.

Duration of treatment and duration of follow-up: This is a cross-sectional study of those undergoing colonoscopy with no follow-up after any pathology is reported.

Outcomes: The primary outcome of analysis will be congruence (agreement) of the colonoscopy physician recommendation with the guideline recommendation.

Endoscopy physician recommendations will be determined from review of the medical records (endoscopy reports and letters/notes in the medical charts). Listing of colonoscopies performed by each enrolled colonoscopy physician will be obtained from the centralised common endoscopy reporting system. Absence of documented recommendations will be considered an incongruent recommendation in the primary analysis and excluded in a sensitivity analysis.

Medical records will be reviewed to obtain the history relevant to determining the timing of surveillance (characteristics of polyps, family history, inclusion and exclusion criteria). Absence of documented relevant history will be considered an incongruent recommendation in the primary analysis and excluded in a sensitivity analysis.

Secondary outcomes will be collected by a survey distributed to the group of endoscopy physicians that were provided the app to determine their opinions about the app. The survey will inquire about ease of use, perception of reliability, likelihood of future use, whether the app provided helpful information and how often they used the app and when. Percentage responses will be tabulated for this descriptive outcome. Investigators will also include open-ended questions about any desired changes to the app.

Another secondary outcome will be the proportion of cases which have a documented recommendation.

Similarly, in this pilot study, the investigators will assess the proportion of the medical records which have complete information (family history of colorectal cancer/polyps, different polyp characteristics) required to make surveillance colonoscopy recommendations as per the guidelines.

Inclusion Criteria:

For physician participants, there are no age limits/parameters. All physicians performing colonoscopy on adult patients in Winnipeg will be eligible to participate (other than those involved in pilot testing and study investigators). Participating physicians will be informed that the introduced intervention will be an information tool to help improve care for individuals undergoing colonoscopy, and will be asked to consent to review of their patient records. Participating physicians will be informed that no individual-level information will be disclosed at any time and all analysis will be on anonymized data.

For patient participants:

• Patients must be older than 50 years and up to 75 years old for inclusion in the study data.
• Adequate bowel preparation defined by Boston Bowel Preparation Scale Score of ≥ 2 in each of segments of the colon (Boston bowel prep scale score recording is mandatory in the Winnipeg city-wide endoscopy reporting system);
• Colonoscopy completed to the cecum/ileocolonic anastomosis;
• Colonoscopy performed between 1 to 4 months before randomization of endoscopy physician (to determine baseline adherence) or between 3 to 7 months after randomization (to determine effect of the intervention).

Exclusion Criteria:

Physicians will be excluded if they are away for more than six weeks continuously in the six months after randomization. In this situation, an alternate physician will be recruited and randomized. Thus there is no anticipated loss of follow-up.

Patients will be excluded if any of the following criteria are met:

1. History of inherited CRC predisposition (Lynch Syndrome, Familial adenomatous polyposis, others);
2. Inflammatory bowel disease;
3. Partially excised polyp or endoscopy physician recommending early colonoscopy to document complete excision;
4. Endoscopy physician documenting a rationale for not repeating colonoscopy in future such as co-morbid conditions.

Study findings will be disseminated via presentations at professional forums and published in a peer-reviewed journal.

No specific feedback will be provided to individual participants as we will perform and report aggregate analysis.

Randomization will be performed after the endoscopy physician consent and enrolment into the study. A computer generated randomization sequence will used in blocks of eight. Stratification for surgical and gastroenterology endoscopy physician will be performed as there might be differential adherence to recommendations by specialty of training.Randomization will be performed and the application distributed by study personnel different than those involved in data extraction from medical records so that the person extracting the data remains blinded to the randomized group assignment. Patients of participating physicians will not require blinding as they will not be approached by the study investigators and will get routine care decided by their physicians.