Published online Nov 7, 2026. doi: 10.3748/wjg.121527
Revised: May 13, 2026
Accepted: June 17, 2026
Published online: November 7, 2026
Processing time: 176 Days and 0.3 Hours
Standard withdrawal time (SWT) is widely used as a quality indicator for colo
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.
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.
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 con
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.
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.
- Citation: 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, Pei ZC. Disentangling true inspection time from therapeutic intervals: Validation of effective withdrawal time as a colonoscopy quality metric. World J Gastroenterol 2026; 32(41): 121527
- URL: https://www.wjgnet.com/1007-9327/full/v32/i41/121527.htm
- DOI: https://dx.doi.org/10.3748/wjg.121527
Colonoscopy is regarded as the gold standard method for the screening and diagnosis of colorectal polyps, and the quality of the procedure directly determines the adenoma detection rate (ADR) and the risk of post-procedural colorectal cancer (CRC)[1,2]. Key quality indicators include withdrawal time and ADR[3]; international guidelines mandate a minimum 6-minute withdrawal to ensure adequate inspection[4].
Despite widespread adherence to this guideline, the optimal withdrawal time remains debated. Several studies have suggested that longer durations (8-9 minutes or more) may improve mucosal inspection[2,5-7]. Moreover, the standard withdrawal time (SWT), which primarily reflects the duration of withdrawal, fails to quantify the “effectiveness” of mucosal observation. This limitation arises because the SWT often includes periods of polypectomy or obscured visualization[8]. Consequently, the SWT may not be a fully reliable indicator of examination quality. We previously developed an artificial intelligence (AI) method to calculate the effective withdrawal time (EWT)[9], but its associations with key quality indicators such as the ADR and polyp detection rate (PDR) have not been fully evaluated. Subsequent studies suggested that the EWT may outperform the SWT for lesion detection, but such findings were limited by small sample sizes and insufficient subgroup analyses[10].
Currently, no standardized method exists for quantifying effective mucosal observation time, and traditional stopwatch timing is labor-intensive and error-prone[11,12]. Furthermore, to align with international definitions of negative colonoscopy, existing studies frequently restrict analysis to colonoscopies without therapeutic interventions, which may limit generalizability[10,13]. In the present study, we used an AI algorithm to calculate EWT and validated it through manual verification. We then evaluated the association of EWT with ADR and compared its performance with SWT.
This prospective study (from December 2024 to September 2025) conducted at Tianjin Union Medical Center included 1044 patients. From an initial cohort of 1059 procedures, 15 (1.4%) were excluded due to incomplete examinations (failure to reach the caecum) or patient intolerance. The procedures were performed by six endoscopists, including four senior physicians (> 8 years of experience; approximately 1000 annual cases) and two junior physicians (approximately 7 years of experience; approximately 600 annual cases). Ethical approval for this study was obtained from the Medical Ethics Committee of Tianjin Union Medical Center (approval No. 2022-C10).
All examinations were performed using Olympus CV-290 colonoscopes (Olympus Optical Co., Tokyo, Japan)[14]. Bowel preparation was completed using the Boston Bowel Preparation Scale (BBPS). The colonoscope was advanced to the cecum, and the mucosa was inspected during withdrawal. Detected lesions were documented, and polyps were resected using appropriate techniques such as cold biopsy, cold snare, or endoscopic mucosal resection. Resected specimens were fixed in 10% neutral buffered formalin and submitted for histopathological examination, which was performed by experienced gastrointestinal pathologists blinded to endoscopic findings, in accordance with World Health Organization criteria[15]. Each working day was divided into a morning session and an afternoon session, with procedures commencing at 08:30 and concluding at 16:30. Physicians could take a midday break from 12:00 to 13:30. However, sometimes they may have had to forgo their break due to a high volume of patients. Four female and two male endoscopists participated in this study, and they were aged between 30 years and 40 years, Endoscopists’ age and sex were not included in the multivariable analysis due to the limited sample size.
To calculate EWT, we developed an AI system that segments colonoscopy videos into three phases: (1) Therapeutic (instrument visibility); (2) Polyp observation; and (3) Obscured (poor visibility). The algorithm integrates object detection and image classification models to classify video frames. Two YOLOv8n-based object detection models were trained to identify instruments and polyps using > 36000 annotated images from 1917 patients across three hospitals, ensuring a multicenter dataset. The training data covered a wide spectrum of colorectal lesions, including diminutive, small, and large polyps with various morphologies, enhancing the diversity and representativeness of the dataset. All images were annotated by experienced endoscopists with more than 8 years of clinical experience, and ambiguous cases were reviewed by senior experts to ensure annotation quality. In addition, a Vision Transformer classification model was used to detect obscured frames. The model was trained on 9913 colonoscopic images (Figure 1). Frames were categorized as belonging to the therapeutic, polyp observation, or obscured phase. The SWT was defined as the duration from cecal intubation to scope removal. The EWT was then calculated by excluding the durations corresponding to the therapeutic and obscured phases. Model training and inference were performed on a workstation equipped with an NVIDIA GeForce RTX 3070 GPU.
To validate the AI algorithm, a subset of cases was manually reviewed[9]. This validation set comprised all therapeutic colonoscopies (n = 143) and a computer-generated random selection of 50% of diagnostic cases (n = 451). The manually verified EWT values were then compared with the algorithm-generated results to evaluate their agreement and consistency. Manual review was conducted by trained endoscopists. Reviewers performed assessments independently and were blinded to the AI results. Therapeutic procedure time commenced upon the entry of endotherapeutic devices (e.g., biopsy forceps or snares) into the endoscopic view and concluded once these devices were fully retracted and the respective interventions were terminated. During manual verification of obscured phase, frames were classified as obscured when more than 50% of the image area appeared darkened or reddened. Frames with severe darkness, redness, or glare were also classified as obscured. If the discrepancy between the manual EWT and the AI-calculated EWT exceeded 1 minute, the corresponding video was re-examined by a senior investigator to reach a final consensus. These measures minimized interobserver variability and provided a reliable reference for evaluating AI performance.
Continuous variables were expressed as median with interquartile range. Comparisons were performed using the Mann-Whitney U test, Kruskal-Wallis test, or χ2 test, as appropriate. Equivalence between AI-calculated and manual EWT was assessed using Two One-Sided Tests[16,17]. This method was selected to rigorously demonstrate that the AI mea
A total of 1044 patients were included (486 males, 46.6%), of which 901 (86.3%) underwent diagnostic colonoscopies and 143 (13.7%) underwent therapeutic colonoscopies. The therapeutic group was significantly older (median 64.0 vs 58.0 years, P < 0.001) with a higher proportion of males (58.0% vs 44.7%, P < 0.001) compared with the diagnostic group. The median BBPS was 6 (range: 1-9). The biopsy submission rate, PDR, and ADR were higher in the therapeutic group (all P < 0.001). SWT and EWT were both longer in the therapeutic vs diagnostic group (860 seconds vs 278 seconds; 478 seconds vs 179 seconds; both P < 0.001; Table 1).
| Item | Total (n = 1044) | Therapeutic (n = 143) | Diagnostic (n = 901) | P value |
| Age, years | 59 (46, 67) | 64.0 (53, 69) | 58.0 (45, 66) | < 0.001 |
| Sex | 0.004 | |||
| Female | 558 (53.4) | 60 (42.0) | 498 (55.3) | |
| Male | 486 (46.6) | 83 (58.0) | 403 (44.7) | |
| Bowel preparation score | 6 (5, 7) | 6 (6, 7) | 6 (5, 7) | 0.243 |
| Bowel preparation score group | 0.073 | |||
| 1-3 | 73 (7.0) | 9 (6.3) | 64 (7.1) | |
| 4-6 | 648 (62.1) | 78 (54.5) | 570 (63.3) | |
| 7-9 | 323 (30.9) | 56 (39.2) | 267 (29.6) | |
| Physician experience | 0.001 | |||
| Junior | 325 (31.1) | 25 (17.5) | 300 (33.3) | |
| Senior | 719 (68.9) | 118 (82.5) | 601 (66.7) | |
| Number of polyps | < 0.001 | |||
| ≤ 3 | 522 (50.0) | 30 (21.0) | 492 (54.6) | |
| > 3 | 522 (50.0) | 113 (79.0) | 409 (45.4) | |
| Biopsy rate1 | 467 (44.7) | 114 (79.7) | 353 (39.2) | < 0.001 |
| PDR2 | 749 (71.7) | 143 (100) | 606 (67.3) | < 0.001 |
| ADR3 | 170 (36.4) | 79 (69.2) | 91 (25.8) | < 0.001 |
| CRC detection rate4 | 12 (2.6) | 1 (0.9) | 11(3.1) | 0.004 |
| SWT (seconds) | 305 (227, 444) | 860 (573, 1230) | 278 (219, 371) | < 0.001 |
| Quintile 1 (Q1) | 176 (88, 212) | 450 (258, 539) | 169 (88, 205) | |
| Quintile 2 (Q2) | 240 (212, 268) | 597 (558, 730) | 229 (206, 250) | |
| Quintile 3 (Q3) | 305 (268, 345) | 872 (733, 971) | 278 (250, 315) | |
| Quintile 4 (Q4) | 400 (346, 512) | 1159 (1012, 1421) | 348 (315, 398) | |
| Quintile 5 (Q5) | 836 (513, 3857) | 1700 (1422, 3857) | 508 (399, 2156) | |
| EWT (seconds) | 197 (143, 284) | 478 (356, 764) | 179 (138, 240) | < 0.001 |
| Quintile 1 (Q1) | 105 (25, 132) | 246 (127, 325) | 101 (25, 125) | |
| Quintile 2 (Q2) | 154 (133, 171) | 378 (333, 424) | 147 (125, 162) | |
| Quintile 3 (Q3) | 198 (171, 225) | 480 (424, 566) | 179 (162, 204) | |
| Quintile 4 (Q4) | 253 (225, 333) | 660 (570, 909) | 228 (204, 255) | |
| Quintile 5 (Q5) | 515 (334, 1702) | 1076 (921, 1702) | 338 (256, 1277) |
Instrument detection achieved near-perfect accuracy (mAP50 of 0.995, precision of 0.998, recall of 0.997). Polyp detection was robust (mAP50 of 0.949, precision of 0.895, recall of 0.937). Similarly, the image classification model showed reliable and balanced performance, with a mAP50, precision, and recall all approximating 0.938. Manual review confirmed only minimal discrepancies between the AI-calculated and manually verified EWT in both the therapeutic and screening cohorts, with mean differences of -0.4 seconds and 1.7 seconds, respectively. The AI-calculated EWT correlated strongly with manual verification across both groups (r = 0.998, P < 0.001). Despite significant subcomponent differences in the diagnostic group (P < 0.001), Two One-Sided Tests confirmed equivalence within ± 5 seconds (P < 0.001; Supplementary Table 1). Bland-Altman analysis further supported the high agreement between the AI-calculated and manually verified EWT. The bias was minimal, and the majority of observations fell within the 95% limits of agreement (-10.47 to 7.09 seconds for the diagnostic group and -11.52 to 12.39 seconds for the therapeutic group, Supplementary Figure 1), confirming strong consistency between the AI-based and manual measurements.
Based on the distribution of SWT and EWT, the first two quintiles had similar withdrawal time ranges and low detection rates; therefore, quintile 1 and quintile 2 were combined as the reference group to provide more stable estimates and better reflect the dose–response relationship. After this adjustment, clear dose–response relationships between longer withdrawal time and higher lesion detection rates were observed for both SWT and EWT (Figure 2). EWT showed a strong dose–response association with adenoma detection, with aORs increasing progressively across quintiles and reaching 360.46 (95%CI: 107.11-1577.04) in Q5. Sensitivity analyses excluding the upper 3% of EWT values (n = 31) yielded consistent results, and the overall dose-response relationships for both ADR and PDR were preserved (Supplementary Table 2 and Supplementary Figure 2). Although SWT also showed increasing aORs, they were consistently lower than those for EWT. A similar trend was observed for polyp detection. The aORs for EWT increased from 2.50 in Q3 to 6.53 in Q4 and 86.68 in Q5 (Figure 2). Again, SWT demonstrated a weaker gradient in comparison with EWT. For CRC detection, both SWT and EWT showed an overall upward trend across withdrawal time categories, although estimates were unstable due to the limited number of cases. Overall, EWT demonstrated a steeper and more consistent dose-response association with adenoma and polyp detection than SWT, reinforcing its potential superiority as a quality indicator for colonoscopy.
The predictive ability of EWT for polyp and adenoma detection reached 0.843 and 0.945, respectively, in the continuous-variable models. Moreover, whether modeled as a continuous variable or as quintiles, EWT consistently showed higher AUCs than SWT for predicting both polyp and adenoma detection (P < 0.01; Figure 3), indicating superior discriminative performance. Results from the quintile and continuous models were highly concordant, with increasing EWT associated with progressively higher probabilities of polyp and adenoma detection. In contrast, the effect of SWT increased more gradually, and its continuous influence diminished in the higher range of values (Figures 2 and 3), further supporting the notion that EWT more accurately captures the duration of effective mucosal inspection. For CRC detection, no significant differences in predictive performance were observed between SWT and EWT due to the limited number of cancer cases. Nevertheless, in the quintile model, EWT achieved an AUC of 0.832 for predicting CRC detection, although this finding should be interpreted cautiously given the small sample size.
We examined the relationship between ADR, PDR, and withdrawal time during diagnostic colonoscopy. In the multivariable-adjusted fitted curves, ADR increased markedly with longer EWT but demonstrated a clear pattern of diminishing marginal returns. When EWT was < 6 minutes, ADR rose rapidly; between 6 minutes and 10 minutes the increase slowed; and the curve plateaued around 10 minutes (Figure 4). A similar pattern was observed for PDR, which increased steadily up to 6 minutes and then flattened, with minimal improvement and a slight decline beyond 10 minutes. In contrast, when withdrawal time was assessed using SWT, both ADR and PDR increased monotonically without an obvious saturation point. Notably, categorization based on EWT revealed clearer inflection points in detection rates, suggesting that EWT more accurately reflects effective mucosal inspection time and is more sensitive for identifying the saturation point where further inspection yields diminishing returns.
Figure 5 illustrates the fitted curves of ADR as a function of both EWT and SWT across different subgroups defined by polyp burden (≤ 3 polyps vs > 3 polyps), endoscopist experience (junior vs senior), and BBPS score (1-6 vs 7-9). Across all three subgroup stratifications, the pattern of ADR change with increasing EWT was largely consistent with the overall findings, showing a clear pattern of diminishing marginal returns: (1) ADR increased steeply during the early phase of EWT (< 6 minutes); (2) Continued to rise with a reduced slope between 6 minutes and 10 minutes; and (3) Plateaued – or increased only minimally – beyond approximately 10 minutes. A similar pattern was observed for PDR across subgroups (Figure 6).
Subgroup analyses revealed notable differences. Senior endoscopists consistently achieved higher adjusted ADRs across all time points compared with junior endoscopists. In patients with better bowel preparation (BBPS 7-9), a high ADR was achieved even with relatively short EWTs; conversely, the ADR remained substantially lower in patients with a BBPS of 1-6, regardless of withdrawal time (Figure 5). In addition, the ADR was higher among patients with more than three polyps even with a very short EWT, suggesting that baseline lesion burden strongly influences lesion detection. Similar patterns were observed for PDR: Examinations performed by senior endoscopists and in those with good bowel preparation achieved higher PDRs and reached high detection levels even with shorter withdrawal times. In contrast, procedures by junior endoscopists or in patients with suboptimal bowel preparation showed much slower increases in PDR over time.
This prospective study validates EWT as a biologically plausible metric that outperforms SWT by isolating pure mucosal inspection. We identified a “saturation effect” in which diagnostic yield follows a law of diminishing returns, plateauing after 6-10 minutes of effective time.
Current international guidelines advocate for a minimum 6-minute withdrawal time based on SWT[4,18]. However, SWT is inherently confounded by the pathology it seeks to measure. Therapeutic colonoscopies showed longer EWT than diagnostic procedures (478 seconds vs 179 seconds), reflecting additional inspection required for lesion characterization[19]. SWT further prolongs this duration (860 seconds vs 278 seconds) by including non-diagnostic intervals. By isolating EWT, non-diagnostic procedural time is excluded, leaving only effective mucosal inspection time. This likely explains the higher aORs and AUCs observed for EWT, which more directly reflects the endoscopist’s visual engagement.
The translational value of EWT lies in its feasibility for real-time clinical implementation. Our algorithm, based on the lightweight YOLOv8n architecture[20], processes video frames with millisecond-level latency. Although the wide-scale deployment of such AI systems entails initial costs, including hardware upgrades and physician training, by reducing the rate of missed lesions and potentially extending safe surveillance intervals, accurate EWT monitoring offers a cost-effective strategy to optimize resource allocation and improve patient outcomes in CRC screening programs.
A key finding of this study is the non-linear relationship between inspection time and detection rates. Unlike SWT, which showed a deceptive monotonic increase[21-23], EWT curves for ADR and PDR rose sharply before 6 minutes, increased more gradually between 6 minutes and 10 minutes, and plateaued thereafter. This suggests a limit to lesion detection during a single inspection. Extending inspection beyond 10 minutes yields minimal diagnostic gain, and further time effectively amounts to “over-inspection” without yield. This plateau effect is likely multifactorial. First, prolonged inspection may be associated with cognitive and visual fatigue, which can impair sustained attention and subtle pattern recognition, thereby counterbalancing the theoretical benefits of extended observation time. Second, given the limited surface area of the colonic wall, mucosal inspection gradually becomes exhaustive once EWT reaches a certain threshold. Consequently, the likelihood of additional diagnostic yield diminishes with further time prolongation. Together, quality improvement should ensure procedures reach the 6-10 minutes effective window, rather than indiscriminately encouraging longer withdrawal times, which may reduce unit efficiency without improving patient outcomes[18].
The universality of this saturation phenomenon was further reinforced by its stability across diverse clinical scenarios. EWT is not intended to replace SWT but to refine it. Current guidelines recommending a minimum SWT of 6 minutes provide a pragmatic and easily measurable benchmark for clinical practice. However, SWT inherently includes non-diagnostic intervals, which may dilute its ability to accurately reflect true mucosal inspection. In this context, EWT can serve as a complementary metric that enhances the interpretability of SWT by isolating effective inspection time. Our findings suggest that achieving an SWT ≥ 6 minutes should ideally be accompanied by sufficient effective inspection, with an EWT in the range of 6-10 minutes representing an optimal balance between diagnostic yield and procedural efficiency. Therefore, rather than functioning as an alternative standard, EWT may be integrated as an adjunctive indicator within existing quality frameworks to better ensure that adequate withdrawal time translates into meaningful mucosal evaluation. SWT remains a valuable, easily measurable proxy for resource management. However, EWT complements this by providing a granular analysis of how that time is spent, specifically filtering out non-diagnostic intervals. By using SWT as a baseline and EWT as a precision indicator, endoscopy units can achieve a more comprehensive quality assurance framework, ensuring that time spent strictly correlates with effective mucosal visualization. The non-linear plateau persisted regardless of bowel preparation quality or endoscopist seniority, suggesting the “10-minute ceiling” reflects limits of visual inspection. Notably, senior endoscopists achieved the plateau phase of detection more rapidly than junior endoscopists, and high-quality bowel preparation (BBPS 7-9) supported peak detection rates earlier than intermediate preparation. This implies that “effective” time is synergistic with the “quality” of the view; superior visualization allows the endoscopist to reach the efficiency frontier faster, whereas suboptimal conditions require more time to achieve the same yield[24,25]. Furthermore, senior physicians achieved higher detection rates at equivalent EWTs compared with junior physicians, highlighting that while time is a prerequisite for quality, the cognitive skill of recognition remains a key variable[26].
The strengths of this study are underpinned by its robust methodological design. Unlike most existing AI-based quality metrics, which are largely derived from retrospective analyses of video repositories[27-29], this study was prospective in nature, ensuring high data fidelity and minimizing selection bias. Furthermore, our analysis was powered by a substantial sample size (n = 1044), providing statistical stability to our findings. The validity of the EWT metric was further solidified by a rigorous validation process, where the AI algorithm demonstrated high concordance when benchmarked against manual assessment.
This study has several limitations. Firstly, this study was conducted at a single tertiary medical center in Tianjin, China. Although the use of standardized endoscopic protocols and consistent patient management within a single center ensured high internal validity for the initial validation of the EWT algorithm, the specific demographic characteristics of the study population, as well as the limited number of physician participants, may potentially limit the generalizability of our findings to other geographic regions or healthcare settings. Secondly, sufficient adjustments for potential confounders were inapplicable in this study. Patients with comorbidities, particularly diabetes, may have poorer bowel preparation due to impaired gastrointestinal motility, and physician fatigue may accumulate with consecutive procedures or extended working hours, potentially impairing sustained attention, visual acuity, and clinical judgment. Future studies should collect this information. Thirdly, all endoscopic examinations in this study were performed using Olympus CV-290 endoscopy systems. As a result, variations in imaging quality, color rendering, and video processing algorithms across manufacturers may affect algorithm performance. Lastly, the number of CRC cases was relatively low (n = 12), resulting in wide confidence intervals for cancer-specific detection estimates. Therefore, these findings should be interpreted with caution and considered exploratory or for reference only. Future studies should include multi-center cohorts with diverse patient populations and incorporate multiple endoscopy systems to comprehensively evaluate the generalizability and clinical utility of EWT as a universal quality metric. Future studies should incorporate fatigue and workload metrics to clarify the independent and interactive effects of fatigue and experience on colonoscopy quality outcomes. In addition, ADR in this study was calculated based on patients who underwent biopsy rather than all examined patients, due to incomplete histological confirmation in cases where biopsy was declined. This approach may limit direct comparability with studies using the standard ADR definition.
The current 6-minute SWT recommendations remain widely adopted. Our findings suggest that EWT offers an important complementary metric rather than a replacement. By eliminating non-diagnostic intervals, EWT provides a more physiologically relevant measure of mucosal inspection and reveals a clear dose–response relationship with lesion detection. The optimal inspection window appears to be between 6 minutes and 10 minutes of effective time: Ensuring at least 6 minutes is essential, whereas extending beyond 10 minutes offers diminishing returns. Real-time EWT monitoring may provide endoscopists with actionable feedback to optimize inspection efficiency without unnecessarily prolonging procedures.
| 1. | Nierengarten MB. Colonoscopy remains the gold standard for screening despite recent tarnish: Although a recent study seemed to indicate that colonoscopies are not as effective as once thought at detecting colorectal cancer, a closer look at the study clears the confusion: Although a recent study seemed to indicate that colonoscopies are not as effective as once thought at detecting colorectal cancer, a closer look at the study clears the confusion. Cancer. 2023;129:330-331. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 23] [Reference Citation Analysis (0)] |
| 2. | Tiankanon K, Aniwan S. What are the priority quality indicators for colonoscopy in real-world clinical practice? Dig Endosc. 2024;36:30-39. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 13] [Reference Citation Analysis (0)] |
| 3. | Desai M, Rex DK, Bohm ME, Davitkov P, DeWitt JM, Fischer M, Faulx G, Heath R, Imler TD, James-Stevenson TN, Kahi CJ, Kessler WR, Kohli DR, McHenry L, Rai T, Rogers NA, Sagi SV, Sathyamurthy A, Vennalaganti P, Sundaram S, Patel H, Higbee A, Kennedy K, Lahr R, Stojadinovikj G, Campbell C, Dasari C, Parasa S, Faulx A, Sharma P. Impact of withdrawal time on adenoma detection rate: results from a prospective multicenter trial. Gastrointest Endosc. 2023;97:537-543.e2. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 27] [Cited by in RCA: 32] [Article Influence: 10.7] [Reference Citation Analysis (2)] |
| 4. | Barclay RL, Vicari JJ, Doughty AS, Johanson JF, Greenlaw RL. Colonoscopic withdrawal times and adenoma detection during screening colonoscopy. N Engl J Med. 2006;355:2533-2541. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1058] [Cited by in RCA: 982] [Article Influence: 49.1] [Reference Citation Analysis (7)] |
| 5. | Yang LS, Thompson AJ, Taylor ACF, Desmond PV, Holt BA. Quality of upper GI endoscopy: a prospective cohort study on impact of endoscopist education. Gastrointest Endosc. 2022;96:467-475.e1. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1] [Cited by in RCA: 10] [Article Influence: 2.5] [Reference Citation Analysis (1)] |
| 6. | Garborg KK. How many minutes should we take for withdrawal - guideline recommendations and upcoming evidence. Best Pract Res Clin Gastroenterol. 2026;80:102019. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 4] [Reference Citation Analysis (0)] |
| 7. | Xu BX, Xu CZ, Zhang HY, Chen XJ, Wei BN, Yang C. Personalizing withdrawal time by insertion time to achieve target adenoma detection rate in colonoscopy. World J Gastroenterol. 2025;31:111364. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 2] [Reference Citation Analysis (0)] |
| 8. | Kadle N, Westerveld DR, Banerjee D, Jacobs C, Gesiotto F, Moon N, Forde JJ, Conti M, Hatamleh D, Taylor R, Brar T, Riverso M, Jawaid S, Perbtani YB, Zhang Y, Draganov PV, Beyth R, Yang D. Discrepancy between self-reported and actual colonoscopy polypectomy practices for the removal of small polyps. Gastrointest Endosc. 2020;91:655-662.e2. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 6] [Cited by in RCA: 11] [Article Influence: 1.8] [Reference Citation Analysis (0)] |
| 9. | Li J, Peng Z, Wang X, Zhang S, Sun J, Li Y, Zhang Q, Shi L, Li H, Tian Z, Feng Y, Mu J, Tang N, Wang X, Li W, Pei Z. Development and validation of a novel colonoscopy withdrawal time indicator based on YOLOv5. J Gastroenterol Hepatol. 2024;39:1613-1622. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 4] [Reference Citation Analysis (0)] |
| 10. | Lui TKL, Ko MKL, Liu JJ, Xiao X, Leung WK. Artificial intelligence-assisted real-time monitoring of effective withdrawal time during colonoscopy: a novel quality marker of colonoscopy. Gastrointest Endosc. 2024;99:419-427.e6. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 24] [Reference Citation Analysis (0)] |
| 11. | Coghlan E, Laferrere L, Zenon E, Marini JM, Rainero G, San Roman A, Posadas Martinez ML, Nadales A. Timed screening colonoscopy: a randomized trial of two colonoscopic withdrawal techniques. Surg Endosc. 2020;34:1200-1205. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 20] [Cited by in RCA: 20] [Article Influence: 3.3] [Reference Citation Analysis (1)] |
| 12. | Zhao S, Yang X, Wang S, Meng Q, Wang R, Bo L, Chang X, Pan P, Xia T, Yang F, Yao J, Zheng J, Sheng J, Zhao X, Tang S, Wang Y, Wang Y, Gong A, Chen W, Shen J, Zhu X, Wang S, Yan C, Yang Y, Zhu Y, Ma RJ, Wang R, Ma Y, Li Z, Bai Y. Impact of 9-Minute Withdrawal Time on the Adenoma Detection Rate: A Multicenter Randomized Controlled Trial. Clin Gastroenterol Hepatol. 2022;20:e168-e181. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 63] [Cited by in RCA: 62] [Article Influence: 15.5] [Reference Citation Analysis (1)] |
| 13. | Kawamura T, Oda Y, Toyoizumi H, Kato M, Sekiguchi M, Takamaru H, Mizuguchi Y, Horiguchi G, Kobayashi K, Sada M, Yokoyama A, Utsumi T, Tsuji Y, Ohki D, Takeuchi Y, Shichijo S, Ikematsu H, Matsuda K, Teramukai S, Kobayashi N, Matsuda T, Saito Y, Tanaka K. Risk of colorectal cancer among fecal immunochemical test-positive individuals by timing of previous colonoscopy: A multicenter analysis. J Gastroenterol Hepatol. 2025;40:153-158. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 4] [Reference Citation Analysis (0)] |
| 14. | Ka-luen Lui T, Yee K, Wong K, Leung WK. 1062 Use of artificial intelligence image classifer for real-time detection of colonic polyps. Gastrointest Endosc. 2019;89:AB135. [DOI] [Full Text] |
| 15. | Dekker E, Bleijenberg A, Balaguer F; Dutch-Spanish-British Serrated Polyposis Syndrome collaboration. Update on the World Health Organization Criteria for Diagnosis of Serrated Polyposis Syndrome. Gastroenterology. 2020;158:1520-1523. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 62] [Cited by in RCA: 46] [Article Influence: 7.7] [Reference Citation Analysis (2)] |
| 16. | Dixon PM, Saint-Maurice PF, Kim Y, Hibbing P, Bai Y, Welk GJ. A Primer on the Use of Equivalence Testing for Evaluating Measurement Agreement. Med Sci Sports Exerc. 2018;50:837-845. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 107] [Cited by in RCA: 157] [Article Influence: 22.4] [Reference Citation Analysis (0)] |
| 17. | Amrhein V, Greenland S, McShane B. Scientists rise up against statistical significance. Nature. 2019;567:305-307. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1318] [Cited by in RCA: 1642] [Article Influence: 234.6] [Reference Citation Analysis (0)] |
| 18. | Shaukat A, Rector TS, Church TR, Lederle FA, Kim AS, Rank JM, Allen JI. Longer Withdrawal Time Is Associated With a Reduced Incidence of Interval Cancer After Screening Colonoscopy. Gastroenterology. 2015;149:952-957. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 144] [Cited by in RCA: 192] [Article Influence: 17.5] [Reference Citation Analysis (0)] |
| 19. | Nagl S, Ebigbo A, Goelder SK, Roemmele C, Neuhaus L, Weber T, Braun G, Probst A, Schnoy E, Kafel AJ, Muzalyova A, Messmann H. Underwater vs Conventional Endoscopic Mucosal Resection of Large Sessile or Flat Colorectal Polyps: A Prospective Randomized Controlled Trial. Gastroenterology. 2021;161:1460-1474.e1. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 83] [Cited by in RCA: 78] [Article Influence: 15.6] [Reference Citation Analysis (1)] |
| 20. | Jeong SW, Ahmad S, Kim JS, Whangbo T. A lightweight YOLOv8-based model for gastric cancer detection. Comput Biol Med. 2025;196:110689. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 2] [Reference Citation Analysis (0)] |
| 21. | Jung Y, Joo YE, Kim HG, Jeon SR, Cha JM, Yang HJ, Kim JW, Lee J, Kim KO, Song HK, Hwangbo Y, Shin JE. Relationship between the endoscopic withdrawal time and adenoma/polyp detection rate in individual colonic segments: a KASID multicenter study. Gastrointest Endosc. 2019;89:523-530. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 30] [Cited by in RCA: 54] [Article Influence: 7.7] [Reference Citation Analysis (0)] |
| 22. | Bhurwal A, Rattan P, Sarkar A, Patel A, Haroon S, Gjeorgjievski M, Bansal V, Mutneja H. A comparison of 9-min colonoscopy withdrawal time and 6-min colonoscopy withdrawal time: A systematic review and meta-analysis. J Gastroenterol Hepatol. 2021;36:3260-3267. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 25] [Cited by in RCA: 24] [Article Influence: 4.8] [Reference Citation Analysis (0)] |
| 23. | Arora A, McDonald C, Guizzetti L, Iansavichene A, Brahmania M, Khanna N, Wilson A, Jairath V, Sey M. Endoscopy Unit Level Interventions to Improve Adenoma Detection Rate: A Systematic Review and Meta-Analysis. Clin Gastroenterol Hepatol. 2023;21:3238-3257. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 5] [Cited by in RCA: 8] [Article Influence: 2.7] [Reference Citation Analysis (1)] |
| 24. | Johnson DA, Barkun AN, Cohen LB, Dominitz JA, Kaltenbach T, Martel M, Robertson DJ, Boland RC, Giardello FM, Lieberman DA, Levin TR, Rex DK. Optimizing Adequacy of Bowel Cleansing for Colonoscopy: Recommendations From the US Multi-Society Task Force on Colorectal Cancer. Am J Gastroenterol. 2014;109:1528-1545. [RCA] [DOI] [Full Text] [Cited by in Crossref: 116] [Cited by in RCA: 115] [Article Influence: 9.6] [Reference Citation Analysis (0)] |
| 25. | Shahini E, Sinagra E, Vitello A, Ranaldo R, Contaldo A, Facciorusso A, Maida M. Factors affecting the quality of bowel preparation for colonoscopy in hard-to-prepare patients: Evidence from the literature. World J Gastroenterol. 2023;29:1685-1707. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in CrossRef: 91] [Cited by in RCA: 79] [Article Influence: 26.3] [Reference Citation Analysis (0)] |
| 26. | Jover R, Zapater P, Bujanda L, Hernández V, Cubiella J, Pellisé M, Ponce M, Ono A, Lanas A, Seoane A, Marín-Gabriel JC, Chaparro M, Cacho G, Herreros-de-Tejada A, Fernández-Díez S, Peris A, Nicolás-Pérez D, Murcia O, Castells A, Quintero E; COLONPREV Study Investigators. Endoscopist characteristics that influence the quality of colonoscopy. Endoscopy. 2016;48:241-247. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 44] [Cited by in RCA: 49] [Article Influence: 4.9] [Reference Citation Analysis (0)] |
| 27. | Naeem MS, Farooq A, Sadiq Z, Saleem I, Siddique MU, Shirazi A, Farooq S, Sarwar MZ, Ali AA. Evaluating the Safety and Quality of Diagnostic Colonoscopies Performed by General Surgeons: A Retrospective Study. Cureus. 2023;15:e38955. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 1] [Reference Citation Analysis (1)] |
| 28. | Yue G, Zhang L, Du J, Zhou T, Zhou W, Lin W. Subjective and Objective Quality Assessment of Colonoscopy Videos. IEEE Trans Med Imaging. 2025;44:841-854. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 4] [Cited by in RCA: 3] [Article Influence: 3.0] [Reference Citation Analysis (0)] |
| 29. | Zessner-Spitzenberg J, Jiricka L, Waldmann E, Rockenbauer LM, Cook J, Hinterberger A, Majcher B, Szymanska A, Asaturi A, Trauner M, Ferlitsch M. Polyp characteristics at screening colonoscopy and post-colonoscopy colorectal cancer mortality: a retrospective cohort study. Gastrointest Endosc. 2023;97:1109-1118.e2. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 7] [Reference Citation Analysis (0)] |