ScotCap – A large observational cohort study

Abstract Aim The aim of this work was to evaluate the performance of colon capsule endoscopy (CCE) in a lower gastrointestinal diagnostic care pathway. Method This large multicentre prospective clinical evaluation recruited symptomatic patients (patients requiring investigation of symptoms suggestive of colorectal pathology) and surveillance patients (patients due to undergo surveillance colonoscopy). Patients aged 18 years or over were invited to participate and undergo CCE by a secondary‐care clinician if they met the referral criteria for a colonoscopy. The primary outcome was the test completion rate (visualization of the whole colon and rectum). We also measured the need for further tests after CCE. Results A total of 733 patients were invited to take part in this evaluation, with 509 patients undergoing CCE. Of these, 316 were symptomatic patients and 193 were surveillance patients. Two hundred and twenty‐eight of the 316 symptomatic patients (72%) and 137 of the 193 surveillance patients (71%) had a complete test. It was found that 118/316 (37%) of symptomatic patients required no further test following CCE, while 103/316 (33%) and 81/316 (26%) required a colonoscopy and flexible sigmoidoscopy, respectively. Fifty‐three of the 193 surveillance patients (28%) required no further test following CCE, while 104/193 (54%) and 30/193 (16%) required a colonoscopy and flexible sigmoidoscopy, respectively. No patient in this evaluation was diagnosed with colorectal cancer. Two patients experienced serious adverse events – one capsule retention with obstruction and one hospital admission with dehydration due to the bowel preparation. Conclusion CCE is a safe, well‐tolerated diagnostic test which can reduce the proportion of patients requiring colonoscopy, but the test completion rate needs to be improved to match that of lower gastrointestinal endoscopy.


Study design and participants
This large, multicentre, prospective clinical evaluation took place in three NHS Scotland health boards (Highland, Grampian and Western Isles). Patient recruitment was divided into two cohorts: symptomatic (patients requiring investigation of symptoms suggestive of colorectal pathology) and surveillance (patients being monitored because of an increased risk of development of colorectal cancer).
Inclusion and exclusion criteria are described in Table 1. All potential patients were invited to participate by introductory letter from the regional research and the endoscopy administration departments. Patients who declined CCE were treated according to standard care and offered colonoscopy. Demographic and baseline clinical data were collected by a data manager and recorded in the CASTOR Electronic Data Capture system, (a research database platform) for all patients undergoing CCE. The evaluation protocol is available online (https://www.dhi-scotl and.com/media/ og5b0 g3n/ scotc ap-clini cal-evalu ation -proto col-v2-0-final -27-08-19-1.pdf).

Procedures
The capsule used in this evaluation was the PillCam™ COLON 2; the technical details and method use are detailed elsewhere [16]. Patients underwent CCE procedures in seven geographically convenient locations (four community healthcare centres and three district general hospitals) to minimize patient travel. The pretest dietary restriction, bowel preparation and booster regimen were consistent throughout the evaluation (Appendix 1). Procedures were supervised by trained nursing staff following a standardized protocol (Appendix 2). CCE reports were generated by consultant gastroenterologists based in Scotland (CCE reader) and returned electronically to the referring hospital clinician. An audit of 22 random reports was performed by an independent, experienced CCE reader, with low interobserver error (Appendix 3, Table C1).
The CCE reports detailed whether the examination was complete (if incomplete, the extent of colon visualized was reported), and included the bowel cleanliness rating according to the Boston bowel preparation scale by colonic segment (right colon, transverse colon, left colon including rectum) and the location, size and morphology of any colonic pathology [17]. CCE examinations were defined as complete if the capsule was excreted during its battery life

What does this paper add to the literature?
This clinical evaluation demonstrated that colon capsule endoscopy (CCE) can be safely introduced into a diagnostic care pathway as an alternative to colonoscopy, and reduces the need for colonoscopy. The completion rate for CCE is acceptable but still needs to be improved to reduce the number of flexible sigmoidoscopy procedures that are required following CCE.
or the anal cushions were visualized. The bowel preparation was deemed adequate if the rating was at least fair for all segments and the CCE reader judged the overall quality to be acceptable. Any relevant extracolonic (e.g. small bowel) findings were also reported.
The CCE report was reviewed by the referring clinician who decided if further management to investigate CCE findings, or incomplete tests, was needed. In general, polyps were examined by endoscopy as per European guidelines [18]. All patients for whom areas of the colon were inadequately visualized by the capsule due to an incomplete test or inadequate bowel preparation were referred for further investigation. Any further investigation (colonoscopy, flexible sigmoidoscopy or CTC) was carried out in line with local standard care. The CCE reports were made available to those doing further endoscopic procedures. The findings of the follow-up investigations, and any associated pathology results, were collated by a data manager and recorded in CASTOR. The urgency rating (urgent suspected cancer, urgent, routine) of the follow-up investigation request was also recorded. Adverse events related to the CCE procedures were recorded following a review of unplanned admissions in the patient's electronic health record by a data manager and clinical researcher (CM). Adverse events associated with follow-up investigations were not recorded.

Outcomes
The primary outcome was the CCE test completion rate (excretion of the capsule within its battery life or visualization of the anal cushions). The secondary outcomes were rate of uptake of CCE, successful bowel preparation rate, findings from CCE (polyp, inflammation, colorectal cancer), the need for a further diagnostic bowel test (colonoscopy, flexible sigmoidoscopy, CTC) and findings and/or pathology found at further investigation (polyp, inflammation, colorectal cancer). In addition, we recorded the urgency and indication of follow-up investigations, and the detection rate of colonic polyps and colorectal cancer in follow-up tests.

Statistical analysis
The primary outcome was determined as the proportion of complete CCE tests out of the total number of CCE procedures. The uptake of CCE was calculated as the proportion of patients who successfully swallowed the capsule out of those invited to participate in the evaluation. The number of patients requiring each follow-up test was calculated as a sum of those who had undergone that test during the evaluation period plus those who were scheduled to undergo the test following the end of the evaluation period.
The requirement for follow-up investigation was recorded from patients' CCE reports by a clinical researcher (CM). Follow-up tests were classified as 'due to CCE findings' if there were any findings reported by CCE necessitating endoscopy, regardless of whether the CCE examination was adequate. Investigations carried out solely due to an incomplete CCE and or inadequate bowel preparation were classified as 'inadequate procedure'.
Polyp matching analysis was done for patients who underwent follow-up endoscopic examination. Only colonic segments adequately visualized by CCE were considered for polyp matching purposes. For a polyp detected at CCE to be considered a true positive it had to match a polyp found at endoscopy located within the same or adjacent colonic segment (right, transverse and left colon including the rectum), and the size measured at CCE, plus or minus 50%, had to overlap with the size measured at endoscopy, plus or minus F I G U R E 1 PillCam™ COLON 2 (copyright Medtronic)  We also calculated the number of colonoscopy appointments, classified by urgency rating, made available by the use of CCE. If a patient's original referral was prioritized as urgent and they underwent CCE with no follow-up required, one urgent colonoscopy appointment was 'saved'. Thus, the demand for tests with different urgency could be calculated following CCE, taking into account the total number of follow-up tests requested in different urgency categories.  (Table D1). The clinical evaluation, and collection of follow-up data, were terminated on 15 January 2021.

RE SULTS
Baseline patient characteristics for each cohort that swallowed the CCE capsule are shown in Table 2. In total, 316 patients were symptomatic patients and 193 were surveillance patients. The majority of the symptomatic patients were female (57%) while the majority of surveillance patients were male (59%). A change in bowel habit was the most common referral symptom in the symptomatic group (66%). The presence of previous polyps was the most common reason for surveillance (52%). The mean full blood count haemoglobin for symptomatic patients was 141 g/l. The FIT data for symptomatic patients are available in Table 2 Table 5  This was managed with intravenous fluids. Two patients experienced nonserious adverse events. One patient developed a Mallory-Weiss tear due to vomiting after bowel preparation. One patient reported pain on excretion of the capsule. In addition, one patient experienced technical failure of the recorder, leading to them undergoing a colonoscopy, and one patient's data recorder was temporarily misplaced due to a logistical error, resulting in them undergoing a CTC.

DISCUSS ION
In this multicentre prospective clinical evaluation, we found that 72% of symptomatic patients and 71% of surveillance patients had a complete CCE test. The CCE test completion rates need to be improved and must aspire to match those of optical colonoscopy (80%-92%). Whilst different 'booster' regimens have been trialled, substantially better completion rates will only be achieved when a capsule is equipped with a battery that enables the whole colon  [20]. Moreover, the chances of diminutive polyps progressing to a more sinister pathology or harbouring a cancer is 0% at 3 years [21]. Follow-up endoscopy for these small polyps is still widely performed but is probably unnecessary.
The polyp detection rate of CCE is well established [12]. The per-  need for follow-up endoscopy is high, patients may value faster reassurance that they do not have bowel cancer.

Patient acceptance will be important to stakeholders when
considering the introduction of new investigations. Qualitative research conducted in parallel to this study reported that patients found the procedure less invasive and painless compared with colonoscopy [22]. In addition, most patients (83%) indicated they would recommend CCE to others. The acceptance rate to participate in this service evaluation was high (69%), potentially reflecting the availability of CCE in geographically convenient locations or a desire to avoid colonoscopy. These findings are supported by other studies demonstrating CCE to be well-tolerated by patients [23][24][25]. We must recognize, however, that colonoscopy will be preferred by some patients who decline CCE (24%) and it should continue to be discussed in any informed consent process.
Currently, there is a paucity of published cost-effectiveness data comparing CCE with colonoscopy. An economic analysis using interim data from this evaluation has shown an additional per-patient cost of GBP54 for a CCE diagnostic pathway, falling to GBP11 per patient in year 5, assuming wide-scale adoption of the test [26].
Further cost benefit is likely to be accrued as the efficiency of CCE reading is improved with the introduction of machine learning algorithms to reduce the reading time [27]. The introduction of new and alternative colon capsules to the market will also reduce cost.
Further cost-effectiveness research is needed as the technology matures and completion rates improve.
One of the strengths of this evaluation is that it pragmatically evaluates the introduction of CCE into a standard clinical pathway.
We captured clinician decision-making based on CCE as a definitive test. Follow-up procedures were prioritized by CCE findings and not by the urgency of the original GP referral. While the need for follow-up endoscopy after CCE will be considered by some to be too high, the test did allow clinical teams to accurately triage the urgency of any follow-up tests into a realistic timeframe. The overall effect of CCE, within a lower gastrointestinal diagnostic pathway, was to free up urgent colonoscopy appointments by reducing need or retriaging patients into a routine appointment.

PATI ENT CO N S ENT S TATEM ENT
Written consent was obtained from all patients participating in the evaluation.

CLI N I C A L TR I A L R EG I S TR ATI O N
The work was a service evaluation and therefore was not registered as a trial. Figure 1 is a picture of the PillCam COLON 2. Permission to use the picture has been given by Medtronic, UK.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.

CO LO N C A P S U LE E N D OS O PY (CCE) PRO CE D U R E PROTO CO L
1. Patients agreeing to participate are screened by a nurse reviewing their electronic healthcare records.
2. The first telephone consultation is carried out by a nurse with the patient to explain the procedure in detail, confirm consent to continue with the procedure and arrange a test date.
3. The bowel preparation with instructions is sent to the patient's home; regimen detailed in Appendix 1.

4.
A second telephone consultation is carried out by a nurse with the patient to provide additional support for bowel preparation consumption.
5. The patient attends the CCE procedure, which is carried out by a trained nurse.
6. Preprocedure checks are carried out to ensure the patient is safe to continue with the procedure and that the bowel preparation has been adequate. If bowel cleanliness is inadequate, then one sachet of Picolax in 1 l of water is administered.
7. The belt and recorder are fitted, then the capsule is swallowed by the patient.
8. The booster medication is provided to the patient with further instructions.
9. The patient returns home to complete the procedure, returning the belt and recorder the following day.
10. The CCE recording is reported using the rapid reader software (Medtronic) by an NHS Scotland gastroenterologist trained in CCE reading.