SKOUT system

Performance Metrics

Clinical Evidence

The SKOUT evidence base is unusually self-contained: a pilot, a registration randomised trial and a post-hoc histology analysis, all led by the same principal investigator and all concerning the same locked detection task. The pilot (Shaukat A, et al., Endosc Int Open 2021;9:E263-E270, PMID 33553591) enrolled 83 screening and surveillance patients at a single outpatient endoscopy centre and compared them with 283 historical controls. The adenoma detection rate was 54.2 percent with the device against 40.6 percent in the historical cohort, without a significant increase in hyperplastic-polyp resection — a hypothesis-generating design, not a controlled comparison. The registration trial (Shaukat A, Lichtenstein DR, Somers SC, et al., Gastroenterology 2022;163:732-741, PMID 35643173; ClinicalTrials.gov NCT04754347) randomised patients across five US academic and community centres with 22 board-certified gastroenterologists, with 1,359 participants in the primary analysis. It had two co-primary endpoints. The first was adenomas per colonoscopy — a deliberate choice, since APC credits every additional adenoma found rather than only the first per patient — tested for superiority: APC rose from 0.83 to 1.05, a 27 percent relative increase (P=.002). The second was the true histology rate, the share of polyp resections with clinically significant histology, tested for non-inferiority: 67.4 percent with CADe against 71.7 percent standard (P for non-inferiority <.001). That non-inferiority framing is the paper's basis for saying the extra detections did not disproportionately add non-neoplastic resections; the raw count of non-neoplastic lesions resected was nevertheless higher in the CADe arm, 375 against 284. The adenoma detection rate, as a secondary endpoint, was 47.8 percent against 43.9 percent — a roughly four-point difference that did not reach statistical significance (P=.065) and is the fairer comparator to the headline figures of other CADe trials. An independent Gastroenterology editorial discussing the trial (PMID 35780871) framed it among the higher-quality studies in the field. The trial group returned to the true histology rate in a post-hoc subgroup analysis (Shaukat A, et al., Gastrointest Endosc 2024, PMID 38964478; 1,423 participants), which compared the rate — the share of resected polyps confirmed as clinically relevant histopathology — between arms and found it similar overall for adenomas, sessile serrated lesions and large hyperplastic polyps, with higher CADe-arm rates confined to subgroups (endoscopists with 11 to 20 years of experience, afternoon procedures). Read together with the class-wide meta-regression literature, which documents wide heterogeneity in the CADe effect across trials and operators, the editorial reading is that SKOUT reliably finds more adenomas per procedure, that the additions are histologically ordinary rather than disproportionately meaningful, and that — as for the class as a whole — the link from more detections to fewer interval cancers remains inferential.

Inside the Auris+ Listing

Regulatory Approvals

Safety Record

No SKOUT recalls, FDA safety communications or published adverse-event reports were identified in publicly available sources as of August 2026. The MAUDE adverse-event database and the openFDA recall endpoint could not be queried directly from this environment, so this should be read as "none found in public reporting" rather than an exhaustive vigilance audit; the registration trial itself reported no device-related adverse events. As a detection-only overlay, the device's harm surface is behavioural rather than mechanical: a false-negative frame does not clear a segment and must not substitute for careful mucosal inspection, extra detections can drive additional polypectomies (the trial's true-histology co-primary endpoint was non-inferior with CADe, although the raw count of non-neoplastic resections was higher, 375 against 284, and the post-hoc true-histology analysis found the flagged polyps histologically similar to conventionally found ones), and the indication is limited to white-light colonoscopy — performance outside that modality is not cleared. Detection pauses automatically when instruments are in the field so the overlay cannot obstruct a resection.

Intended Use & Indications

SKOUT is indicated as a computer-aided detection tool that provides colorectal-polyp location information to assist qualified and trained gastroenterologists in identifying potential colorectal polyps during colonoscopy examinations of adult patients undergoing colorectal-cancer screening or surveillance. The indication is detection and localisation (CADe), not characterisation (CADx): the device marks where to look and makes no statement about what a lesion is. It is limited to standard white-light colonoscopy, is not intended to replace a full patient evaluation or serve as a primary interpretation of the procedure, and in every case the decision to inspect, biopsy or resect — and the histological verdict — remains with the endoscopist and the pathology laboratory. A frame without a bounding box is not an assurance that a segment is lesion-free.