3.1 Of the 4 studies that met the inclusion criteria in the scope, 2 were randomised controlled trials (ADENOMA and E‑Cap) and 2 were non-randomised prospective cohort studies (Rameshshanker et al. 2016 and Tsiamoulos et al. 2018).
3.2 ADENOMA was a UK-based, multicentre, single-blind randomised controlled trial (n=1,772) that compared Endocuff Vision-assisted colonoscopy with standard colonoscopy (Ngu et al. 2018). The population comprised adults having a colonoscopy after presenting with clinical symptoms of bowel cancer, for post-polypectomy surveillance, or after a positive faecal occult blood test as part of bowel cancer screening. The results showed that for the whole study population, Endocuff Vision-assisted colonoscopy led to a significant increase in adenoma detection rates compared with standard colonoscopy (p=0.02). Subgroup analyses showed that this was mainly driven by a 10.8% improvement in the screening population, which comprised about 45% of the whole study population; there was no statistically significant improvement in adenoma detection rates in the other patient subgroups.
The 3 single-centre studies are potentially biased by local expertise and so may not be representative of UK practice
3.3 The other randomised controlled trial (E-Cap) was a single-blind, single-centre study (n=531) comparing Endocuff Vision-assisted colonoscopy with standard colonoscopy in adults who had a positive faecal occult blood test as part of bowel cancer screening. Tsiamoulos (2018) was a single-centre pilot service evaluation of Endocuff Vision, set up as a before-after study (n=410) in adults who had a positive faecal occult blood test as part of bowel cancer screening. Rameshshanker (2016) was a single-centre study (n=96) that prospectively compared Endocuff Vision-assisted colonoscopy with standard colonoscopy. These 3 single-centre studies showed various levels of improvement in adenoma detection rates with Endocuff Vision-assisted colonoscopies. The external assessment centre (EAC) noted that the level of improvement was inversely related to the expertise of the colonoscopists involved; that is, more benefits were seen with Endocuff Vision when used by colonoscopists with less expertise. Because of this, results of the single-centre studies are likely to be biased by local expertise and so may not be representative of NHS practice. For full details of the clinical evidence, see section 3 of the assessment report.
The company's model shows that Endocuff Vision is cost saving in people having colonoscopies as part of bowel cancer screening
3.4 The company submitted a de novo cost model which compared Endocuff Vision-assisted colonoscopy with standard colonoscopy in patients having colonoscopies after a positive faecal occult blood test as part of bowel cancer screening. The model comprised a number of interlinked decision trees and Markov models based on an established model of colorectal cancer. The model results showed that over 10 years, using Endocuff Vision in this patient population would save £12 per patient. Sensitivity analyses showed that the cost savings were driven by how much Endocuff Vision improved the adenoma detection rate, because a higher adenoma detection and resection rate is likely to result in a lower incidence of colorectal cancer (see section 4.5).
The EAC's revised model shows that Endocuff Vision is cost saving if it improves the adenoma detection rate by over 3%
3.5 The EAC considered the model structure, use of 1‑year cycles and a 10‑year time horizon to be appropriate. However, it revised the model to include more appropriate parameters based on data from the bowel cancer screening programme, including the stage distribution of screen-detected and symptomatic-detected colorectal cancer (Sagar et al. 2015) and the success rate for standard colonoscopy (Lee et al. 2012). The EAC's revised model showed that over 10 years, using Endocuff Vision in a screening population would save around £53 per patient. Sensitivity analyses showed similar results to the company's analyses, with the cost savings mainly driven by how much Endocuff Vision improved the adenoma detection rate. Specifically, the EAC's revised model showed that Endocuff Vision was cost saving when it improved the adenoma detection rate by over 3%.
3.6 The company did not submit an analysis of Endocuff Vision in people having treatment unrelated to bowel cancer screening. The EAC did additional analyses and concluded that Endocuff Vision was unlikely to be cost saving in this population because of the modest gain in adenoma detection rates seen with Endocuff Vision in the ADENOMA trial. For full details of the cost evidence, see section 4 of the assessment report.