Time as a marker of quality in diagnostic upper endoscopy: a retrospective study of its association with clinically significant findings

Micheal Tadros; Dharma Ayer; Caesar Ferrari; Kumael Jafri; Sonia Samuel; Dana Gornick; Paul J. Feustel

doi:10.5946/ce.2025.342

Articles

Page Path: HOME > Clin Endosc > Ahead-of print articles > Article

Original Article Time as a marker of quality in diagnostic upper endoscopy: a retrospective study of its association with clinically significant findings: Micheal Tadros¹, Dharma Ayer², Caesar Ferrari², Kumael Jafri², Sonia Samuel³, Dana Gornick⁴, Paul J. Feustel²; DOI: https://doi.org/10.5946/ce.2025.342
Published online: February 24, 2026

¹Department of Gastroenterology, Albany Medical Center, Albany, NY, USA

²Albany Medical College, Albany, NY, USA

³Department of Internal Medicine, Albany Medical Center, Albany, NY, USA

⁴Department of Internal Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA

Correspondence: Micheal Tadros Department of Gastroenterology, Albany Medical Center, 43 New Scotland Avenue, Albany, NY 12208, USA E-mail: tadrosm1@amc.edu

• Received: September 10, 2025 • Revised: October 24, 2025 • Accepted: October 31, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

684 Views
81 Download

Full Article

Download PDF

Abstract
Graphical abstract
INTRODUCTION
METHODS
RESULTS
DISCUSSION
NOTES
REFERENCES

See letter "Seeing more by looking longer: the value of time in diagnostic upper endoscopy".

Abstract

Background/Aims
Procedural duration has been proposed as a marker of esophagogastroduodenoscopy (EGD) quality; however, formal benchmarks remain undefined. This study examined the association between EGD duration and clinically significant findings (CSFs), as well as procedural behaviors contributing to diagnostic yield.
Methods
We conducted a retrospective analysis of 120 first-time diagnostic EGDs performed at a single academic center and categorized procedures as lasting <3, 3 to 6, or >6 minutes. Logistic regression models were used to assess associations among procedure time, quality metrics (photographs, landmarks, biopsies), clinical indications, and CSF detection.
Results
Longer procedures were independently associated with a high rate of CSF detection. Compared with procedures lasting <3 minutes, those lasting >6 minutes had 17.5-fold greater odds of detecting a CSF (p=0.02). Procedures performed for bleeding or “other” indications were independently associated with a high diagnostic yield (p=0.02 and 0.03, respectively). Although increased photo count and landmark documentation correlated with time, neither was independently predictive of CSFs after adjustment. Procedural duration also showed a stepwise increase in the number of biopsies obtained, photographs taken, and landmarks identified.
Conclusions
Longer EGD duration was independently associated with increased CSF detection in a stepwise manner. Procedural time may serve as a practical, unifying measure correlated with multiple performance behaviors reflective of comprehensiveness in diagnostic EGD.
Keywords: Endoscopy; Quality indicators; Time factors

Graphical abstract

INTRODUCTION

Esophagogastroduodenoscopy (EGD) is an important diagnostic tool for evaluating upper gastrointestinal (GI) disorders. It is one of the most frequently performed procedures by gastroenterologists and is used to visualize the esophagus, stomach, and duodenum.¹ Primary indications for EGD include dysphagia, gastroesophageal reflux disease (GERD), upper abdominal pain, GI bleeding, and screening for preneoplastic conditions.² The ability of EGDs to detect clinically significant findings (CSFs) may depend on several factors, including patient indication, procedural technique, and endoscopist experience. High-quality examinations are essential to ensure accurate diagnosis and appropriate patient management, especially in cases where early pathological detection can influence outcomes.³

Unlike colonoscopy, which has well-established quality indicators, such as adenoma detection rate and withdrawal time, there are fewer universally accepted quality indicators for EGD.⁴ Several GI societies updated their published guidelines on EGD quality indicators in 2024 and 2025. The most recent versions emphasize the importance of measures such as appropriate indications, comprehensive photo documentation, adequate mucosal visualization, and targeted biopsy sampling.³^,⁵ However, these guidelines currently lack formal recommendations regarding procedure duration despite growing interest in its potential impact on examination quality.³

Recent studies have increasingly highlighted the potential role of intraprocedural duration as a marker of EGD quality. A 2022 European multicenter study, with a median examination time of 4.2 minutes, reported that longer EGD duration was associated with higher detection of clinically significant lesions, including an increased cancer detection rate, as well as a high biopsy rate and improved neoplasm identification.⁶ Similarly, a prospective trial demonstrated that implementing a minimum EGD inspection time of 6 minutes enhanced detection of upper GI lesions, supporting time as a potential factor in quality improvement initiatives.⁷ An earlier Singaporean study also found that endoscopists with a mean procedure time exceeding 7 minutes identified significantly more high-risk gastric lesions compared with those with shorter examination times, in a predominantly Chinese patient population with smaller proportions of Malay, Indian, and other ethnic groups.⁸ These findings collectively suggest that time spent during EGD may influence diagnostic yield.

Despite these findings, no American gastroenterology society has established a formal grade or strength of recommendation regarding procedure time as a quality indicator for upper endoscopy.³^,⁵ Additionally, existing research has yet to fully examine how procedural time varies across common indications or correlates with specific clinical findings. This study aimed to address these gaps by investigating the relation between CSF detection and EGD duration while also exploring how procedure time varies according to indication and other procedural metrics.

METHODS

Study population and data collection

A retrospective chart review was conducted at a single academic center for patients who underwent EGD between December 15, 2023, and December 31, 2023. Only procedures performed for first-time diagnosis were included. The exclusion criteria were therapeutic EGDs and procedures performed for known or previously diagnosed conditions, including Barrett’s esophagus, varices, screening or surveillance of GI neoplasia, and percutaneous endoscopic gastrostomy placement.

Patient-level and procedural data were recorded. Procedural duration was collected both as a continuous measure and as ordinal categories (<3, 3–6, and >6 minutes). The duration was measured continuously from intubation to withdrawal, including biopsy time. These intervals were selected based on previous studies. Prior research has identified approximately 3 minutes as a threshold distinguishing faster from slower examinations, whereas the average inspection times for diagnostic EGDs typically range between 5 and 10 minutes, with 6 minutes used as a threshold for increased lesion detection. Accordingly, 3 minutes was chosen as the lower cutoff and 6 minutes as the threshold for the highest category to align with these ranges and facilitate comparisons.¹^,⁷^,⁹

Additional variables included the presence of CSFs, pathology results, number of biopsies, number of photographs, number of clearly documented anatomical landmarks, visualization clarity, and clinical indications for EGD. Anatomical landmarks were defined based on published quality guidelines and included, but were not limited to, the lower esophagus/cardia, gastroesophageal junction/fundus, gastric body and antrum, duodenal bulb, and second portion of the duodenum.³ Indications were categorized as anemia; bleeding (such as hematemesis, hematochezia, or melena); dysphagia; nausea; abdominal or epigastric pain; reflux or GERD; or other (such as weight loss, suspected esophagitis, chronic diarrhea, or suspected celiac disease). Definitions of CSFs and indication categories are presented in Table 1. While specific CSF definitions are listed in Table 1, findings were grouped into broader diagnostic categories (such as inflammatory/erosive, ulcerative, preneoplastic/neoplastic, structural/vascular, and polypoid) at the time of data collection for analysis and reporting.

Visualization clarity was assessed using the POLPREP scale, with each photograph rated on a scale of zero to three for mucosal visibility.¹⁰ The mean clarity score was calculated for each patient based on all captured images. Missing data were handled using pairwise deletion, whereby patients were included in analyses for variables with available data. If a specific variable (such as photograph count) was missing for a patient, it was excluded from the analysis.

Outcomes and predictors

The primary outcome was the presence of CSFs on EGD. The primary predictor variable was procedural duration evaluated ordinally (<3, 3–6, and >6 minutes). Additional predictors included the indications for EGD, number of photographs taken, number of biopsies obtained, number of clearly identified landmarks, and operator type (fellow vs attending). Indication was coded as a categorical variable with reflux/GERD as the reference group because of its relatively low diagnostic yield in this cohort and its role as a clinically low-risk baseline.¹¹ Secondary analyses evaluated associations between procedural duration (both continuous and ordinal) and the number of photographs, biopsies, landmarks clearly identified, and indications.

Statistical analysis

Initial descriptive statistics were calculated for all variables, including means, standard deviations, medians, and ranges, where appropriate. To evaluate the association between time categories and CSF detection, we performed a chi-square test and a Cochran–Armitage trend test for ordinal association. Univariate binary logistic regression models were then constructed with CSF as the dependent variable, using each predictor independently. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. A multivariate binary logistic regression model was constructed to assess the independent contributions of each variable after adjusting for potential confounders. All candidate variables from the univariate analyses were included in the multivariable logistic regression model to adjust for potential confounding and evaluate the independent contribution of each factor.

To assess the relation between procedural duration and other continuous quality indicators (including number of photographs, landmarks, and biopsies), we used both one-way analysis of variance (ANOVA) and univariable linear regression models. For time versus indication, we performed Pearson’s chi-square tests and ordinal logistic regression using IBM SPSS ver. 30.0 (IBM Corp.) to determine whether the presence of specific indications influenced the likelihood of belonging to a particular time category. All other analyses, including chi-square, Cochran–Armitage, and logistic regression models for CSFs, were conducted using R ver. 4.4.1 (R Foundation for Statistical Computing). Descriptive statistics, one-way ANOVA, and linear regressions for time were performed using Minitab Statistical Software ver. 19 (Minitab LLC).

Ethics approval

This study was reviewed by the Institutional Review Board of Albany Medical Center and was determined to be exempt from full Institutional Review Board review and approval under 45 CFR 46.104(d)(4)(iii), as it involved secondary research using private health information regulated under the HIPAA Privacy Rule. The exemption determination was issued under InfoEd Protocol #6944 on April 11, 2024.

RESULTS

A total of 120 EGD procedures were included in the final analysis. The mean age of the patients was 50.6±20.5 years. The mean visualization clarity score was 2.59±0.37, indicating generally high mucosal visibility across procedures. On average, 4.93±1.99 anatomical landmarks were clearly identified, 1.58±1.72 biopsies were taken, and 8.6±3.2 photographs were captured per procedure. The median procedural duration was 6.82 minutes, and the mean was 9.00±7.48 minutes. Overall, 15.2% (n=18) of the procedures lasted <3 minutes, 31.3% (n=37) lasted 3–6 minutes, and 53.8% (n=64) exceeded 6 minutes. All procedures were performed under monitored anesthesia care unless contraindicated, and no procedure was complicated by an adverse event that affected its completion or interpretation. The baseline demographic and procedural characteristics are summarized in Table 2.

CSFs and procedural time

Procedural time was categorized into three intervals (<3, 3–6, and >6 minutes) to evaluate whether duration influenced CSF detection. There was a significant overall association between time interval and presence of CSFs (χ²=16.97, df=2, p<0.001). Subsequently, the Cochran–Armitage test for trends was conducted to examine whether this relation followed a stepwise pattern. The test further supported a stepwise increase in CSF detection with longer procedural durations (Z=−3.05, p<0.01). Figure 1 shows the proportional increase in CSF detection with longer procedure durations.

CSFs were identified in 49 of the 120 procedures (40.8%). The most frequent category was inflammatory or erosive lesions (n=20, 40.8%), including gastritis, erosive gastropathy, eosinophilic esophagitis, and reflux-related esophagitis. Ulcerative findings (n=14, 28.6%) included esophageal, gastric, and duodenal ulcers, both bleeding and nonbleeding, and findings consistent with peptic ulcer disease. Preneoplastic and neoplastic lesions (n=7, 14.3%) included intestinal metaplasia, dysplasia, and carcinoma. Structural and vascular abnormalities (n=5, 10.2%) included hernias, ectopic mucosa, and varices. Patients with polyps (n=3, 6.1%) presented with multiple benign-appearing gastric polyps, some with overlying infection but without dysplasia.

Univariable associations with CSF detection

To assess the individual associations with CSF detection, binary logistic regression models were used for different procedural and clinical variables. Compared with procedures lasting <3 minutes, those lasting 3–6 minutes were associated with a 9.2-fold increase in the odds of CSF detection (OR, 9.20; 95% CI, 1.10–77.23; p=0.04), whereas procedures lasting >6 minutes had a 23.3-fold increase (OR, 23.30; 95% CI, 2.92–185.90; p=0.003). Each additional photograph captured was associated with a 15% increase in the odds of CSF detection (OR, 1.15; 95% CI, 1.03–1.28; p=0.01). Similarly, the number of anatomical landmarks clearly identified was significantly associated with CSF detection (OR, 1.30; 95% CI, 1.07–1.59; p=0.01). The number of biopsies obtained was not significantly associated with CSF detection (OR, 1.16; 95% CI, 0.94–1.43; p=0.18).

The effect of indication on CSF detection was also assessed using reflux/GERD as the reference group. Procedures performed for bleeding were associated with a 4.79-fold increase (OR, 4.79; 95% CI, 1.10–20.79; p=0.04). Other indications, including anemia (OR, 2.93), dysphagia (OR, 2.44), nausea (OR, 1.22), abdominal or epigastric pain (OR, 2.39), and “other” (OR, 3.21), showed elevated ORs but did not reach statistical significance. As shown in Figure 2, the distribution of CSFs varied by indication, with the highest proportions observed in procedures performed for bleeding, followed by “other,” anemia, and dysphagia. Lastly, operator type (whether the procedure was performed by a fellow or attending) was not significantly associated with CSF detection (OR, 1.33; 95% CI, 0.64–2.37; p=0.45).

Multivariable logistic regression model for CSFs

A multivariable binary logistic regression model was used to determine the independent effects of each variable while adjusting for potential confounders. As shown in Table 3, procedure times greater than 6 minutes remained significantly associated with increased odds of CSF detection (OR, 17.52; 95% CI, 1.70–180.71; p=0.02), whereas the 3–6 minutes group did not reach statistical significance (OR, 8.49; 95% CI, 0.89–80.88; p=0.06). Among clinical indications, bleeding (OR, 10.00; 95% CI, 1.40–71.43; p=0.02) and the “other” category (OR, 8.42; 95% CI, 1.20–59.19; p=0.03) were significant independent predictors. The number of photographs (OR, 1.04; 95% CI, 0.91–1.19), biopsies (OR, 1.23; 95% CI, 0.87–1.74), and landmarks clearly identified (OR, 1.13; 95% CI, 0.90–1.42) were not independently associated with CSF detection after adjustment. Operator type (fellow vs attending) also remained non-significant (OR, 0.54; 95% CI, 0.17–1.69).

Time associations with procedural characteristics

Given that procedural time emerged as a strong independent predictor of CSFs, additional analyses were performed to explore the relations between procedural and clinical variables. Specifically, we examined how time was related to photo documentation, the number of landmarks clearly identified, the number of biopsies, and indications.

To explore the potential factors contributing to the stepwise relation between longer procedure times and increased CSF detection, we examined the number of photographs taken across time categories. There was a clear increase in photo documentation with longer procedures: the average number of photographs was 5.1 for procedures <3 minutes, 8.3 for those 3–6 minutes, and 9.8 for those >6 minutes. This trend was statistically significant in both the regression analysis (p<0.001, R²=14.3%) and ANOVA (p<0.001). Furthermore, a significant relation was observed between procedural time and the number of anatomical landmarks clearly identified, with both regression (p=0.006, R²=6.3%) and ANOVA (p<0.001) analyses supporting this association. The mean landmark counts also increased in a stepwise manner with longer durations: 3.44 for <3 minutes, 4.70 for 3–6 minutes, and 5.52 for >6 minutes. Although linear regression did not demonstrate a statistically significant association between procedural time and the number of biopsies obtained, ANOVA revealed significant differences across time categories (p<0.05), with a clear stepwise increase in biopsy counts as procedural duration increased: 0.94 for procedures <3 minutes, 1.23 for 3–6 minutes, and 1.96 for those >6 minutes.

Ordinal regression analysis showed no statistically significant association between time intervals and indications (anemia, bleeding, dysphagia, nausea, abdominal or epigastric pain, and reflux/GERD screening; p>0.05). However, there were observable trends in the distribution of some of these indications across the time intervals. Specifically, 60% of pain cases, 67% of bleeding cases, and 89% of dysphagia cases showed a trend toward the >6 minutes category.

DISCUSSION

This study demonstrated a clear association between longer EGD procedural durations and the detection of CSFs. A stepwise increase in diagnostic yield was observed across time categories, with procedures lasting 3–6 minutes showing a notable elevation in the odds of CSF detection compared with those lasting <3 minutes. This trend was more pronounced for procedures >6 minutes, which were significantly more likely to yield CSFs. This finding supports procedural time as a potential surrogate marker of endoscopic quality, even in first-time diagnostic cases. These results are consistent with prior studies emphasizing the importance of inspection time during EGD for detecting Barrett’s esophagus, gastric intestinal metaplasia, and neoplasia.⁵ Our analysis extends the existing literature by clearly demonstrating a stepwise association between CSF detection and the defined time strata, reinforcing its utility as an important quality metric. While the 6 minutes threshold used in our study may work effectively in a Western context, it likely varies regionally. For example, several Asian countries, such as Japan, have high rates of gastric cancer; thus, the optimal inspection time for diagnostic EGD may be greater.⁵^,⁸ Additionally, recent European and American guidelines have suggested thresholds of >7 minutes, although prior studies have also used the 6 minutes mark as a high-time threshold, which informed the interval selection in this analysis.⁵^-⁷^,¹²

In the univariable analysis, several procedural variables were significantly associated with CSF detection. Each additional photograph captured was associated with a 15% increase in the odds of CSF detection, suggesting that more thorough documentation may contribute to improved lesion identification. Similarly, the number of clearly identified anatomical landmarks was significantly associated with CSF detection, reflecting the importance of a comprehensive mucosal inspection. Although biopsy acquisition is an established quality indicator for EGD, the number of biopsies obtained was not significantly associated with CSF detection in this cohort.⁵ This may reflect study-specific limitations, such as the exclusion of patients undergoing surveillance, where biopsy collection is more routine.

In the multivariable logistic regression model, adjusting for key procedural and clinical factors, procedural durations >6 minutes and select indications (notably bleeding and the “other” category) remained independently associated with increased odds of CSF detection. Importantly, the procedural quality indicators that were significant in the univariable analysis (photographs, landmarks, and biopsies) did not remain independently predictive after adjustment. These findings reiterate the value of procedural duration and select clinical indications as independent predictors of CSF detection while also highlighting the importance of adjusting for confounders when interpreting associations with other quality indicators.

Given the strong association between procedural duration and CSF detection, we explored whether this relation could be explained by procedural behaviors reflective of comprehensiveness. The numbers of photographs captured, anatomical landmarks identified, and biopsies obtained increased across the time categories in a stepwise manner. These variables represent potential contributors to or reflections of procedural quality and may help explain the mechanisms underlying the observed association between time and diagnostic yield. For instance, the observed increase in photo documentation with longer procedures supports the interpretation that an extended duration allows for a more comprehensive mucosal assessment and documentation. Similarly, an increase in biopsy frequency and landmark documentation over time reinforces the potential of procedural duration as an enabling factor for adherence to quality practices. Although these individual variables were not independently associated with CSFs in the multivariable model, their correlation with time suggests that procedural duration may serve as a unifying quality measure.

This interpretation aligns with current recommendations, which endorse both minimum observation times for certain conditions and standardized photo documentation protocols, including anatomical landmark capture, as benchmarks for high-quality EGD.⁵ Studies from Japan have demonstrated that inspection times exceeding 5–7 minutes are associated with significantly high odds of detecting gastric neoplasia.¹³ However, these results may reflect the unique epidemiological context of Japan, where gastric cancer prevalence and screening frequency are higher than in Western countries.¹⁴ The association between an increased photo count and time observed in our study provides further support for this guideline-based approach. International societies have already established photo documentation benchmarks. The European Society of Gastrointestinal Endoscopy recommends at least 10 landmark images in a normal EGD, whereas the Korean Society of Gastrointestinal Endoscopy recommends a minimum of eight. In countries, such as South Korea and Japan, where the incidence of upper GI cancers is high and routine upper endoscopy screening is more prevalent, endoscopists frequently exceed these thresholds, capturing 20–40 images per procedure.¹⁴^,¹⁵ The World Endoscopy Organization also supports extensive photo documentation, recommending up to 28 images to optimize the detection of early gastric cancer and other significant findings.¹⁶ Institutional interventions, such as implementing minimum observation-time policies even in first-time diagnostic EGDs and providing education on EGD quality metrics, may further enhance practice-based improvement strategies informed by procedural duration.

Thus, our key finding is that while quality indicators, such as photographic documentation, anatomical landmark identification, and biopsy frequency, did not retain statistical significance in the adjusted model, they all trended upward with longer procedural durations. Procedural duration remained an independent predictor of CSF detection in both univariable and multivariable analyses. This suggests that adequate procedural time may be the most practical and impactful target for improving diagnostic yield, as it encapsulates multiple interrelated behaviors that contribute to overall endoscopic quality.

A sensitivity analysis restricted to preneoplastic and neoplastic findings (n=7) was performed to evaluate whether the observed association persisted in lesions with malignant potential. Owing to the small number of such cases, the model did not converge and yielded non-estimable coefficients; however, a descriptive review showed that most cases (five of seven, 71%) occurred during procedures >6 minutes, consistent with the direction of the primary analysis. These findings suggest that with a larger sample size, a statistically significant association between time and neoplastic lesions alone may be observed.

This study had some limitations that should be considered when interpreting the findings. It was conducted at a single center with a modest sample size, which affected the CIs and may limit the generalizability of the findings. Additionally, because biopsy time was included within the total procedural duration, this may have introduced some variability, although this approach has been used in other studies assessing endoscopic inspection time.⁶ Patients who underwent therapeutic or surveillance procedures, or those with known diagnosis-related EGDs, were excluded, which may have influenced the strength of the association between biopsy acquisition and CSFs. It is also important to note that factors, such as sedation depth, mucosal cleanliness, and the severity of the patient’s clinical status (such as active bleeding), may have influenced both procedure duration and CSF detection, representing potential confounders. Additionally, the CSF classification was derived from a retrospective chart review, which may have introduced misclassification bias. Operator experience was dichotomized as fellow versus attending, which may not fully capture the variability in procedural expertise; however, all cases were performed under the supervision of fellows at similar levels of training. Finally, we acknowledge the potential for reverse causality, as procedures may last longer after lesion detection, although adjustments were made for biopsy number and indication in the multivariable logistic regression analysis to partially mitigate this effect. Future studies should include larger multicenter cohorts, employ standardized prospective definitions of CSFs, and explore real-time assessments of procedural practices, including those at the endoscopist level, to validate and expand upon these findings.

While a longer inspection time may enhance diagnostic yield, it is also important to acknowledge the potential procedural and sedation-related complications that can accompany prolonged endoscopy. Longer endoscopic procedures are associated with an increased risk of cardiopulmonary and sedation-related complications, especially hypoxemia and greater operator fatigue.¹⁷^,¹⁸ Additionally, the risk of complications, such as aspiration pneumonia, may increase with longer procedure duration, particularly among older patients and those with comorbidities.¹⁸

The strengths of this study include its characterization of a stepwise relation between procedural duration and CSFs across distinct time intervals, offering a more detailed understanding of the association between time and diagnostic yield. Unlike previous studies that focused exclusively on neoplastic lesions, this analysis encompassed a broader range of clinically significant pathologies relevant to routine diagnostic EGDs, enhancing clinical applicability. Both univariable and multivariable logistic regression models were used to assess independent predictors of CSF detection, with adjustment for key procedural and clinical confounders such as indication and operator type. The study also examined procedural quality behaviors, including photo documentation, landmark identification, and biopsy acquisition, in parallel with time, providing insights into how duration may act as an integrative proxy for endoscopic meticulousness.

In conclusion, the findings of this study support incorporating procedural time as a key process-based quality measure in diagnostic EGD. A longer procedural duration may serve as an indicator of a comprehensive evaluation, including adequate photo documentation, landmark identification, and mucosal inspection. These factors may not individually capture the full scope of procedural quality. Thus, incorporating time as a key quality indicator may enhance the detection of clinically meaningful pathologies.

Conflicts of Interest

The authors have no potential conflicts of interest.

Funding

None.

Author Contributions

Conceptualization: MT, DG, SS, DA; Data curation: DA, CF, KJ; Formal analysis: DA, PJF; Investigation: DA, CF, KJ; Methodology: MT, DG, SS, DA; Project administration: MT; Writing–original draft: MT, DA; Writing–review & editing: all authors.

Fig. 1.

Distribution of clinically significant findings (CSFs) by endoscopy duration category. CSFs were more frequently identified in procedures lasting longer than 3 minutes, with the highest detection rate observed in those exceeding 6 minutes. This trend supports a positive association between procedural time and CSFs.

Fig. 2.

Distribution of clinically significant findings (CSFs) across different procedural indications. The highest proportion of CSFs was observed in procedures performed for bleeding, while lower proportions were seen in reflux/gastroesophageal reflux disease and nausea. This visual trend parallels the regression findings, which identified bleeding as a statistically significant predictor of CSF detection.

Table 1.

Definitions of CSFs and indications

Variable	Included definitions
CSFs	Esophagitis (e.g., erosive or eosinophilic esophagitis)
	Barrett’s esophagus
	Gastritis
	Peptic ulcer disease
	Gastrointestinal malignancies
	Stomach atrophy
	Intestinal metaplasia
	Strictures
	Varices
	Polyps
Included classifications
Indications	Anemia
	Bleeding (hematemesis, hematochezia, melena)
	Dysphagia
	Nausea
	Pain (e.g., epigastric, abdominal)
	Reflux/gastroesophageal reflux disease screening

Definitions of clinically significant findings (CSFs) identified during esophagogastroduodenoscopy, along with the classifications for procedural indications, are listed.

Table 2.

Descriptive statistics of procedural variables

Variable	n	Mean±SD	Median (range)
Age (yr)	120	50.6±20.5	54.5 (13–84)
Procedure duration (min)	119^a)	9.0±7.5	6.8 (1.5–36.6)
Photographs	120	8.6±3.6	8 (0–19)
Landmarks	120	4.9±2.0	5.0 (0–10)
Visualization clarity	120	2.6±0.4	2.7 (1.6–3.0)
Biopsies	120	1.6±1.7	1.0 (0–7)

Descriptive statistics for key procedural and clinical variables among all included esophagogastroduodenoscopy cases (n=120) are presented.

SD, standard deviation.

^a)One case is missing.

Table 3.

Multivariable logistic regression predictors of CSFs

Variable	OR (95% CI)	p-value
3 to 6 min (vs. <3)	8.49 (0.89–80.88)	NS
>6 min (vs. <3)	17.52 (1.70–180.71)	0.02
Landmarks	1.13 (0.90–1.42)	NS
Biopsies	1.23 (0.87–1.74)	NS
Photographs	1.04 (0.91–1.19)	NS
Fellow	0.54 (0.17–1.69)	NS
Anemia	8.41 (0.88–99.4)	NS
Bleeding	10.00 (1.40–71.43)	0.02
Dysphagia	3.26 (0.40–31.1)	NS
Nausea	2.18 (0.07–38.9)	NS
Other	8.42 (1.20–59.19)	0.03
Pain	2.35 (0.48–14.0)	NS

Values are presented as odds ratios (ORs) and 95% confidence intervals (CIs) for predictors of clinically significant findings (CSFs) during EGD. Significant predictors are bolded and italicized including procedure duration >6 minutes, bleeding, and “other” clinical indications. All other variables, including number of photographs, biopsies, landmarks identified, and operator type, did not reach statistical significance.

NS, not significant.

REFERENCES

1. Ahlawat R, Hoilat GJ, Ross AB. Esophagogastroduodenoscopy. StatPearls Publishing; 2025.
2. Rizk MK, Sawhney MS, Cohen J, et al. Quality indicators common to all GI endoscopic procedures. Gastrointest Endosc 2015;81:3–16.Article PubMed
3. Yadlapati R, Early D, Iyer PG, et al. Quality indicators for upper GI endoscopy. Am J Gastroenterol 2025;120:290–312.Article PubMed
4. Ferrari C, Tadros M. Enhancing the quality of upper gastrointestinal endoscopy: current indicators and future trends. Gastroenterol Insights 2024;15:1–18.Article
5. Nagula S, Parasa S, Laine L, et al. AGA clinical practice update on high-quality upper endoscopy: expert review. Clin Gastroenterol Hepatol 2024;22:933–943.Article PubMed
6. Romańczyk M, Romańczyk T, Lesińska M, et al. The relation of esophagogastroduodenoscopy time and novel upper gastrointestinal quality measures. Eur J Gastroenterol Hepatol 2022;34:763–768.Article PubMed
7. Gao Y, Cai MX, Tian B, et al. Setting 6-minute minimal examination time improves the detection of focal upper gastrointestinal tract lesions during endoscopy: a multicenter prospective study. Clin Transl Gastroenterol 2023;14:e00612.Article PubMed PMC
8. Teh JL, Tan JR, Lau LJ, et al. Longer examination time improves detection of gastric cancer during diagnostic upper gastrointestinal endoscopy. Clin Gastroenterol Hepatol 2015;13:480–487.Article PubMed
9. Park JM, Huo SM, Lee HH, et al. Longer observation time increases proportion of neoplasms detected by esophagogastroduodenoscopy. Gastroenterology 2017;153:460–469.Article PubMed
10. Romańczyk M, Ostrowski B, Kozłowska-Petriczko K, et al. Scoring system assessing mucosal visibility of upper gastrointestinal tract: The POLPREP scale. J Gastroenterol Hepatol 2022;37:164–168.Article PubMed PDF
11. Katz PO, Dunbar KB, Schnoll-Sussman FH, et al. ACG clinical guideline for the diagnosis and management of gastroesophageal reflux disease. Am J Gastroenterol 2022;117:27–56.Article PubMed PMC
12. Areia M, Esposito G, Leclercq P, et al. Performance measures for upper gastrointestinal endoscopy: a European Society of Gastrointestinal Endoscopy (ESGE) quality improvement initiative: update 2025. Endoscopy 2025;57:1268–1297.Article PubMed
13. Kawamura T, Wada H, Sakiyama N, et al. Examination time as a quality indicator of screening upper gastrointestinal endoscopy for asymptomatic examinees. Dig Endosc 2017;29:569–575.Article PubMed PDF
14. Lee SH, Park YK, Cho SM, et al. Technical skills and training of upper gastrointestinal endoscopy for new beginners. World J Gastroenterol 2015;21:759–785.Article PubMed PMC
15. Kim SY, Park JM. Quality indicators in esophagogastroduodenoscopy. Clin Endosc 2022;55:319–331.Article PubMed PMC PDF
16. Emura F, Sharma P, Arantes V, et al. Principles and practice to facilitate complete photodocumentation of the upper gastrointestinal tract: World Endoscopy Organization position statement. Dig Endosc 2020;32:168–179.Article PubMed PDF
17. Supan W, Ongiem A, Noitasaeng P, et al. Incidence of complications in high-risk patients undergoing combined esophagogastroduodenoscopy and colonoscopy with total intravenous anesthesia: a prospective cohort study. Ther Clin Risk Manag 2025;21:1395–1408.Article PubMed PMC PDF
18. Levy I, Gralnek IM. Complications of diagnostic colonoscopy, upper endoscopy, and enteroscopy. Best Pract Res Clin Gastroenterol 2016;30:705–718.Article PubMed

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

PubReader
ePub Link

Cite

CITE: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

Time as a marker of quality in diagnostic upper endoscopy: a retrospective study of its association with clinically significant findings

DOI: https://doi.org/10.5946/ce.2025.342

XML Download

Figure

Time as a marker of quality in diagnostic upper endoscopy: a retrospective study of its association with clinically significant findings

Fig. 1. Distribution of clinically significant findings (CSFs) by endoscopy duration category. CSFs were more frequently identified in procedures lasting longer than 3 minutes, with the highest detection rate observed in those exceeding 6 minutes. This trend supports a positive association between procedural time and CSFs.

Fig. 2. Distribution of clinically significant findings (CSFs) across different procedural indications. The highest proportion of CSFs was observed in procedures performed for bleeding, while lower proportions were seen in reflux/gastroesophageal reflux disease and nausea. This visual trend parallels the regression findings, which identified bleeding as a statistically significant predictor of CSF detection.

Graphical abstract

Fig. 1.

Fig. 2.

Graphical abstract

Time as a marker of quality in diagnostic upper endoscopy: a retrospective study of its association with clinically significant findings

Variable	Included definitions
CSFs	Esophagitis (e.g., erosive or eosinophilic esophagitis)
	Barrett’s esophagus
	Gastritis
	Peptic ulcer disease
	Gastrointestinal malignancies
	Stomach atrophy
	Intestinal metaplasia
	Strictures
	Varices
	Polyps
Included classifications
Indications	Anemia
	Bleeding (hematemesis, hematochezia, melena)
	Dysphagia
	Nausea
	Pain (e.g., epigastric, abdominal)
	Reflux/gastroesophageal reflux disease screening

Variable	n	Mean±SD	Median (range)
Age (yr)	120	50.6±20.5	54.5 (13–84)
Procedure duration (min)	119^a)	9.0±7.5	6.8 (1.5–36.6)
Photographs	120	8.6±3.6	8 (0–19)
Landmarks	120	4.9±2.0	5.0 (0–10)
Visualization clarity	120	2.6±0.4	2.7 (1.6–3.0)
Biopsies	120	1.6±1.7	1.0 (0–7)

Variable	OR (95% CI)	p-value
3 to 6 min (vs. <3)	8.49 (0.89–80.88)	NS
>6 min (vs. <3)	17.52 (1.70–180.71)	0.02
Landmarks	1.13 (0.90–1.42)	NS
Biopsies	1.23 (0.87–1.74)	NS
Photographs	1.04 (0.91–1.19)	NS
Fellow	0.54 (0.17–1.69)	NS
Anemia	8.41 (0.88–99.4)	NS
Bleeding	10.00 (1.40–71.43)	0.02
Dysphagia	3.26 (0.40–31.1)	NS
Nausea	2.18 (0.07–38.9)	NS
Other	8.42 (1.20–59.19)	0.03
Pain	2.35 (0.48–14.0)	NS

Table 1. Definitions of CSFs and indications

Definitions of clinically significant findings (CSFs) identified during esophagogastroduodenoscopy, along with the classifications for procedural indications, are listed.

Table 2. Descriptive statistics of procedural variables

Descriptive statistics for key procedural and clinical variables among all included esophagogastroduodenoscopy cases (n=120) are presented.

SD, standard deviation.

^a)One case is missing.

Table 3. Multivariable logistic regression predictors of CSFs

NS, not significant.