3 Committee discussion

The diagnostics advisory committee considered evidence on PillCam COLON 2 for investigating the colon from several sources. This included evidence submitted by Medtronic, a review of clinical and cost evidence by the external assessment group (EAG), and responses from stakeholders. Full details are available in the project documents for this guidance.

The condition

3.1

Colorectal polyps are small growths on the inner lining of the colon. Polyps are very common, affecting about 15% to 20% of the UK population aged 50 or over, and most people are unaware they have them. Polyps are not usually cancerous, but some may develop into adenomas (advanced polyps) which can be precancerous. People with adenomas have a higher risk of developing colorectal cancer if the adenomas are not removed. Colorectal cancer is the fourth most common cancer in the UK (see Cancer Research UK's webpage on cancer incidence for common cancers). There are approximately 44,000 new cases and 16,800 deaths per year, with over half of people diagnosed with colorectal cancer in England surviving at least 10 years after diagnosis (see Cancer Research UK's webpage on bowel cancer statistics). Early diagnosis improves survival.

Current practice

3.2

For people with symptoms that may suggest colorectal cancer, NICE's diagnostics guidance on quantitative faecal immunochemical testing (FIT) to guide colorectal cancer pathway referral in primary care recommends that:

adults with a FIT result of 10 microgram/g or more should be referred using a suspected cancer pathway referral
adults with a FIT result below 10 microgram/g should be referred to an appropriate secondary care pathway if there is a strong clinical concern of cancer because of ongoing unexplained symptoms. The referral would be for the most appropriate test or an appointment with a specialist.

3.3

People who have had high-risk findings on a previous colonoscopy may have follow-up surveillance colonoscopy or colorectal imaging after 3 years.

3.4

Colonoscopy is often used for diagnosing colorectal cancer. It can visualise the entire colon, and biopsies can be taken and examined histologically to confirm a diagnosis. Depending on the clinical situation, polyps may be removed when they are identified using cauterisation or a snare. Colonoscopy needs preparation of the colon using diet modification and laxatives and is most frequently done as an outpatient procedure. It is associated with very rare but serious complications, such as perforation of the bowel and heavy bleeding which may need a transfusion.

3.5

Colonoscopy may not be suitable for elderly people and those with comorbidities such as kidney disease. For these people, CT colonography may be used. Additionally, people who are unwilling to have a colonoscopy may have CT colonography. People having a CT colonography consume a contrast-based material, need air insufflation and are exposed to ionising radiation. If pathology is found then a follow-up colonoscopy, or flexible sigmoidoscopy if pathology is limited to the distal colon, may be needed.

3.6

After bowel polyps are removed, they are sent for laboratory testing. In rare cases, surgery may be needed to treat polyps by removing part of the bowel. If colorectal cancer is confirmed, NICE's guideline on colorectal cancer recommends further imaging tests, such as CT or MRI, to stage the cancer and determine which treatment is needed. Colonoscopy may also find other bowel conditions such as Crohn's disease, ulcerative colitis and diverticulosis, which may need further treatment and follow up.

Unmet need

3.7

NHS endoscopy services are under considerable strain and in some areas there are long waiting lists for colonoscopy. Healthcare professionals have observed that many people on the suspected colorectal cancer referral pathway have no abnormal pathology found at colonoscopy. PillCam COLON 2 is indicated for use as an alternative to colonoscopy to help rule out polyps, colon cancer and other pathology, and to direct people who need a biopsy or polypectomy to therapeutic colonoscopy. This may help reduce waiting lists and aid early cancer detection by prioritising people who need further tests and treatment.

Clinical effectiveness

Diagnostic accuracy

3.8

The committee considered 1 study that included a population directly applicable to the scope. This study included people with symptoms who were referred for colonic investigation, based on NICE criteria before 2021. Sensitivity and specificity for detecting clinically significant polyps of 100% and 98% were reported in 66 people. This evidence was limited by there being a:

small number of people
low prevalence of polyps
high risk of bias because of concerns arising from the unblinded interpretation of the diagnostic tests.

Also, the study only included a subset of the symptomatic population of interest to this evaluation, providing no data for surveillance populations. The committee acknowledged the results and limitations of this evidence, accepting that further diagnostic accuracy evidence is needed to inform decision making.

3.9

Because of the scarcity of evidence identified in directly applicable populations, 5 studies including populations indirectly applicable to the scope were considered by the committee. The proportion of people within scope in these studies ranged from 64% to potentially as few as 11%. Pooled sensitivity and specificity estimates were lower than those seen in the directly applicable evidence. The EAG commented that the difference in diagnostic accuracy between mixed and within-scope populations was likely caused by potential differences in size and distribution of polyps between the groups, introducing uncertainty into the evidence. But it was unable to predict what impact this would have on diagnostic accuracy estimates. There was limited evidence available on the diagnostic accuracy of PillCam COLON 2 for adenomas and colorectal cancer. The committee acknowledged the limitations of the evidence but agreed that including this indirect evidence supports decision making. It concluded that further research should be done to reduce uncertainties in the diagnostic accuracy estimates.

Patient preference

3.10

The committee considered evidence on patient preference from 5 studies and testimony from a lay committee member with lived experience. Preference and satisfaction data for PillCam COLON 2 compared with colonoscopy and CT colonography was mixed. The committee noted that patient preference was likely to be influenced by the clinical situation. The committee acknowledged some points of dissatisfaction with PillCam COLON 2, including:

the provision of information
problems with the bowel preparation regimen
pain and discomfort after swallowing the capsule
the wait for results and a subsequent colonoscopy when needed
the impact on daily life of both the bowel preparation regimen and wearing the sensor and recorder.

3.11

The committee considered that for some people there are potential barriers to having a colonoscopy. This includes people:

who have previously had a negative experience of colonoscopy
who have been sexually abused
on anticoagulants that would need to be stopped before colonoscopy
with certain religious or cultural beliefs.

The committee agreed that these groups would benefit from the option of a less-invasive test.

Ongoing studies

3.12

The committee discussed the ongoing ColoCap study being done in the UK, due to be published in 2027. The committee agreed that this study is likely to provide diagnostic accuracy and patient-preference evidence that is directly applicable to each of the 3 populations described in the decision problem.

Cost effectiveness

Model structure and inputs

3.13

The committee considered a health economic model that was developed to assess the cost effectiveness of PillCam COLON 2 compared with both colonoscopy and CT colonography. The model contained 6 separate populations: symptomatic FIT-positive patients, symptomatic FIT-negative patients, and post-polypectomy surveillance patients, each further split by whether they were willing to undergo colonoscopy or not. The committee was aware that the polyp size categories identified in the literature for diagnostic accuracy did not directly align with those used in the long-term model, which introduces uncertainty.

Model results

3.14

Base-case results showed that PillCam COLON 2 was more costly and less effective than colonoscopy in all populations, resulting in it being dominated. Compared with CT colonography, PillCam COLON 2 was more costly and, depending on the population, either less effective or more effective. This led to it being dominated or having incremental cost effectiveness ratios (ICERs) ranging from £434,488 to £7,208,331 per quality adjusted life year gained, which are well above the range that NICE considers a cost-effective use of NHS resources. The committee noted that differences in quality-adjusted life years gained between tests was small, but that the difference in costs between them heavily contributed to the results. The committee noted that in all scenarios tested, PillCam COLON 2 was either dominated, or had a high ICER. This included an optimistic scenario that assumed a PillCam COLON 2 completion rate of 85% and equal diagnostic accuracy of PillCam COLON 2 and colonoscopy. The committee concluded that PillCam COLON 2 does not appear to be a cost-effective test when compared with either colonoscopy or CT colonography in any of the 6 modelled populations.

Colonoscopy capacity sparing

3.15

The model estimated that, in populations willing to have colonoscopy, using PillCam COLON 2 would result in large reductions in the number of colonoscopies, especially in symptomatic populations. The model estimated small reductions in the number of colonoscopies in populations which initially declined colonoscopy, noting that the only comparator in this population was CT colonography. Evidence from 2 studies identified in the literature review also suggested that using PillCam COLON 2 would significantly reduce the number of colonoscopies needed. Notably, capacity-sparing estimates were highest in people with a negative FIT test result. The committee was concerned that interpretation of PillCam COLON 2 images may be done by the same endoscopists who perform colonoscopies. So, it was concerned whether benefits from colonoscopy capacity sparing would be realised in practice. The committee heard from clinical experts that interpretation of PillCam COLON 2 images by trained endoscopists does not take away from their colonoscopy capacity. So, PillCam COLON 2 could feasibly increase colon investigation capacity and reduce waiting times. Clinical experts noted that a partially complete PillCam COLON 2 investigation could be followed by flexible sigmoidoscopy to visualise the lower gastrointestinal tract, rather than a colonoscopy. Flexible sigmoidoscopy requires a less-intensive bowel preparation procedure and takes less time than a colonoscopy, potentially further freeing up endoscopist capacity. Additionally, the committee noted that colonoscopy is used for purposes beyond those included in the scope of this evaluation. Reducing the number of colonoscopies being done for colorectal cancer investigations could feasibly reduce waiting list times in these other populations. This may result in cost savings or improved health outcomes in these people. But, this was beyond the scope of this evaluation and could not be ascertained from the economic model.

Follow-up testing

3.16

The committee noted that the proportion of people who need colonoscopy or flexible sigmoidoscopy after a PillCam COLON 2 investigation was the key driver of the model results. This is because of the additional costs associated with further testing. In the model, if PillCam COLON 2 results in incomplete visualisation of the whole colon, people have a colonoscopy or flexible sigmoidoscopy. Evidence from the literature review suggested large variations in the completion rates of PillCam COLON 2, ranging from 40% to 89%. The NHS England colon capsule endoscopy pilot study reported 70% completion rates. Clinical experts noted that some people are likely to have lower completion rates than others, for example, people with chronic constipation, people who are inactive and people with diabetes. Clinical experts noted that a learning curve exists for colon capsule endoscopy services. Improvements in completion rates, bowel cleansing rates, interpretation of findings and subsequent onward referral rates are likely to occur as centres gain experience in using the technology. Also, people with significant pathology found with PillCam COLON 2 need a therapeutic colonoscopy or flexible sigmoidoscopy. This means that a large proportion of people having PillCam COLON 2 need to have further investigation, incurring additional costs. Because of this, the committee concluded that PillCam COLON 2 is unlikely to be cost effective in people who are likely to have an incomplete test or are likely to have underlying pathology detected with PillCam COLON 2. It agreed that further research is needed to identify the populations who have the lowest risk of needing a colonoscopy after PillCam COLON 2.

Populations that could benefit most

3.17

The committee discussed subpopulations in which clinical and economic benefits may be seen. It considered this was likely in people who had an incomplete index test colonoscopy, because PillCam COLON 2 may allow complete visualisation of the colon. In current practice, this would need follow up with a CT colonography or a repeat colonoscopy. Clinical experts noted that CT colonography may not be suitable for everyone and that a repeat colonoscopy may not be possible. In this scenario, PillCam COLON 2 may be useful as an alternative test. Clinical experts noted that in this situation, a patient could take the PillCam COLON 2 capsule immediately after the incomplete colonoscopy. This would reduce waiting time for a follow-up test and remove the need for an additional bowel cleansing regimen.

Equality considerations

3.18

The committee acknowledged that there was no data on whether socioeconomic status or ethnicity may impact patient preference or other outcomes in this assessment. The committee agreed that this was an important area for future research which may help to establish in which populations PillCam COLON 2 may provide most benefit.