Disentangling true inspection time from therapeutic intervals: Validation of effective withdrawal time as a colonoscopy quality metric

doi:10.3748/wjg.121527

Advanced Search

BPG is committed to discovery and dissemination of knowledge

Home / Archive / Volume 32, Issue 41

This Article

(0)

Peer-Review Report of This Article

CrossCheck and Google Search of This Article

Academic Rules and Norms of This Article

Supplementary Materials of This Article

Citation of this article

Corresponding Author of This Article

Research Domain of This Article

Article-Type of This Article

Open-Access Policy of This Article

Times Cited Counts in Google of This Article

Journal Information of This Article

Publication Name

World Journal of Gastroenterology

ISSN

1007-9327

Publisher of This Article

Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA

- Full Article (XML)

Prospective Study

Copyright: ©Author(s) 2026. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial (CC BY-NC 4.0) license. No commercial re-use. See permissions. Published by Baishideng Publishing Group Inc.

World J Gastroenterol. Nov 7, 2026; 32(41): 121527
Published online Nov 7, 2026. doi: 10.3748/wjg.121527

Disentangling true inspection time from therapeutic intervals: Validation of effective withdrawal time as a colonoscopy quality metric

Zi-Ye Peng, Jia-Xin Li, Yue Huang, Xiang-Yu Wang, Qi An, Zhao-Yu Yin, Yan-Ru Li, Jia-Yi Sun, Yu-Wei Wang, Xiao Yang, Qi Zhang, Xi-Mo Wang, Shu-Yi Zhang, Zheng-Cun Pei

Zi-Ye Peng, Yue Huang, Xiang-Yu Wang, Qi An, Zhao-Yu Yin, Xi-Mo Wang, Zheng-Cun Pei, Medical School, Tianjin University, Tianjin 300072, China

Jia-Xin Li, Xi-Mo Wang, Department of Surgery, Tianjin Third Central Hospital, Tianjin 300110, China

Jia-Xin Li, Central Hospital, Tianjin University, Tianjin 300072, China

Yan-Ru Li, Jia-Yi Sun, Yu-Wei Wang, Xiao Yang, Qi Zhang, Shu-Yi Zhang, Department of Endoscopy, Tianjin Union Medical Center, Tianjin 150300, China

Co-first authors: Zi-Ye Peng and Jia-Xin Li.

Co-corresponding authors: Shu-Yi Zhang and Zheng-Cun Pei.

Author contributions: Peng ZY and Li JX contributed equally to this work as co-first authors; Peng ZY, Li JX and Huang Y analyzed the data and wrote the manuscript; Peng ZY, Li JX, Huang Y, Wang XY, An Q, Yin ZY and Wang XM performed the research; Li YR, Sun JY, Wang YW, Yang X and Zhang Q contributed new reagents and analytic tools; Zhang SY and Pei ZC supervised the study and revised the manuscript as co-corresponding authors; Peng ZY, Li JX, Huang Y, Wang XY, An Q, Yin ZY, Li YR, Sun JY, Wang YW, Yang X, Zhang Q, Wang XM, Zhang SY and Pei ZC designed the research study; all authors have read and approved the final manuscript.

AI contribution statement: During the preparation of this work, the authors used Gemini (Google) to improve the language, readability, and grammatical accuracy of the manuscript and answering-reviewers document. After using this tool, the authors reviewed and edited the content as needed and took full responsibility for the content of the published article. AI tools did not participate in study design, scientific interpretation, data analysis, or decision-making. No AI-generated images were used in this manuscript.

Supported by the Tianjin Health Science and Technology Project, No. TJWJ2025MS017; and the Natural Science Foundation of Tianjin, No. 24JCQNJC01990.

Institutional review board statement: The study was conducted in accordance with the Declaration of Helsinki, and approved by the Medical Ethics Committee of Tianjin Union Medical Center (approval No. 2022-C10).

Clinical trial registration statement: Trial registration was not required for this study as it was an observational study and did not involve any clinical interventions.

Informed consent statement: All participants provided informed consent.

Conflict-of-interest statement: All authors declare no conflict of interest in publishing the manuscript.

CONSORT 2010 statement: The authors have read the CONSORT 2010 Statement, and the manuscript was prepared and revised according to the CONSORT 2010 Statement.

Data sharing statement: The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Corresponding author: Zheng-Cun Pei, PhD, Medical School, Tianjin University, No. 92 Weijin Road, Nankai District, Tianjin 300072, China. peizhengcun@tju.edu.cn

Received: March 27, 2026
Revised: May 13, 2026
Accepted: June 17, 2026
Published online: November 7, 2026
Processing time: 172 Days and 11.5 Hours

Abstract

BACKGROUND

Standard withdrawal time (SWT) is widely used as a quality indicator for colonoscopy, but it does not distinguish true mucosal inspection from time spent on therapeutic or non-diagnostic activities. This limitation may lead to inaccurate assessment of inspection quality and its association with lesion detection.

AIM

To assess the association between artificial intelligence (AI)-derived effective withdrawal time (EWT) and lesion detection, and identify the optimal mucosal inspection window vs SWT.

METHODS

In this prospective study, colonoscopy videos from December 2024 to September 2025 were analyzed using an AI-based system that automatically identified and excluded therapeutic and visually obscured phases. The algorithm was validated against manual assessment. Adenoma detection rate and polyp detection rate were used as measures of lesion detection. The associations between withdrawal times (EWT and SWT) and these outcomes were analyzed using generalized additive models to characterize non-linear associations and identify potential plateau effects.

RESULTS

AI-derived EWT demonstrated near-perfect concordance with manual verification (r = 0.998). Longer EWT was strongly associated with higher lesion detection. In quintile analyses, adjusted odds ratios for adenoma detection increased progressively across EWT levels. Although SWT showed a similar trend, the associations were consistently weaker. EWT also demonstrated superior predictive performance, with area under the curve values of 0.843 for polyp detection rate and 0.945 for adenoma detection rate. Nonlinear modeling revealed diminishing returns, with rapid increases in detection when EWT was < 6 minutes, slower gains between 6 minutes and 10 minutes, and a plateau thereafter.

CONCLUSION

AI-derived EWT better reflects quality than SWT by excluding therapeutic time; an optimal mucosal inspection window of approximately 6-10 minutes, offering a more precise and clinically actionable metric.

Keywords: Colonoscopy; Artificial intelligence; Adenoma detection rate; Withdrawal time; Quality control

Core Tip: Effective withdrawal time (EWT), derived from an artificial intelligence-based system, provides a refined measure of true mucosal inspection by excluding therapeutic and visually obscured phases during colonoscopy. In this study, EWT showed stronger associations with adenoma detection rate and polyp detection rate than standard withdrawal time, along with superior predictive performance. Nonlinear analysis revealed a plateau effect, suggesting that an inspection time of approximately 6-10 minutes may be optimal. EWT may serve as a more precise and clinically actionable quality metric.