Per-Patient Polyp Detection Rate of Colon Capsule Endoscopy: A Systematic Review and Meta-Analysis with Exploratory Stratification by Referral Pathway

Abstract

Background/Objectives: Colon capsule endoscopy (CCE) is a minimally invasive alternative to optical colonoscopy, but per-patient polyp detection rate (PDR), the proportion of patients with at least one visually reported polyp, varies across referral populations. We estimated any-polyp and size-stratified CCE yield and explored between-study variation. Methods: MEDLINE, Embase, PubMed, and Cochrane CENTRAL were searched from 1 January 2004 to 29 July 2026. Proportions were pooled using random-intercept binomial-normal generalised linear mixed models. PROSPERO: CRD420261464937 (registration not prospective). Results: Of 44 retained studies (12,754 participants), 31 independent studies (11,310 participants) formed the primary pool. Pooled PDR was 53.2% (95% CI 46.1–60.1; prediction interval 19–85%), a descriptive average, not a performance target. Exploratory PDR was 81.2% after positive FIT (k = 4), 49.4% after incomplete colonoscopy (k = 4), and 39.9% in screening without prior FIT enrichment (k = 3); 14 studies had mixed pathways. Pooled detection was 32.7% for polyps ≥ 6 mm (prediction interval 8–74%) and 17.7% for polyps ≥ 10 mm (4–51%). Conclusions: CCE PDR is a visual yield measure, not adenoma detection or diagnostic accuracy. A global CCE PDR is not a portable quality threshold; referral-pathway estimates are descriptive and hypothesis-generating. Future quality metrics require standardised pathways, denominators, and linkage to downstream clinical outcomes.