Endoscopists and endoscopic assistants’ qualifications, but not their biopsy rates, improve gastric precancerous lesions detection rate

Abstract

BACKGROUND

Detecting gastric precancerous lesions (GPLs) is critical for the early diagnosis and treatment of gastric cancer. Endoscopy combined with tissue examination is an important method for detecting GPLs. However, negative biopsy results often increase patients’ risks, economic burdens, and lead to additional healthcare costs. Improving the detection rate of GPLs and reducing the rate of negative biopsies is currently a key focus in endoscopic quality control.

AIM

To explore the relationships between the endoscopist biopsy rate (EBR), qualifications of endoscopists and endoscopic assistants, and detection rate of GPLs.

METHODS

EBR, endoscopists, and endoscopic assistants were divided into four groups: Low, moderate, high, and very high levels. Multivariable logistic regression analysis was used to analyze the relationships between EBR and the qualifications of endoscopists with respect to the detection rate of positive lesions. Pearson and Spearman correlation analyses were used to evaluate the correlation between EBR, endoscopist or endoscopic assistant qualifications, and the detection rate of positive lesions.

RESULTS

Compared with those in the low EBR group, the odds ratio (OR) values for detecting positive lesions in the moderate, high, and very high EBR groups were 1.12 [95% confidence interval (CI): 1.06-1.19, P < 0.001], 1.22 (95%CI: 1.14-1.31, P < 0.001), and 1.38 (95%CI: 1.29-1.47, P < 0.001), respectively. EBR was positively correlated with the detection rate of gastric precancerous conditions (atrophic gastritis/intestinal metaplasia) (ρ = 0.465, P = 0.004). In contrast, the qualifications of the endoscopists were positively correlated with GPLs detection (ρ = 0.448, P = 0.005). Compared to endoscopists with low qualification levels, those with moderate, high, and very high qualification levels endoscopists demonstrated increased detection rates of GPLs by 13% (OR = 1.13, 95%CI: 0.98-1.31), 20% (OR = 1.20, 95%CI: 1.03-1.39), and 32% (OR = 1.32, 95%CI: 1.15-1.52), respectively. Further analysis revealed that the qualifications of endoscopists were positively correlated with the detection rates of GPLs in the cardia (ρ = 0.350, P = 0.034), angularis (ρ = 0.396, P = 0.015) and gastric body (ρ = 0.453, P = 0.005) but not in the antrum (ρ = 0.292, P = 0.079). Moreover, the experience of endoscopic assistants was positively correlated with the detection rate of precancerous lesions by endoscopists with low or moderate qualifications (ρ = 0.427, P = 0.015).

CONCLUSION

Endoscopists and endoscopic assistants with high/very high qualifications, but not EBR, can improve the detection rate of GPLs. These results provide reliable evidence for the development of gastroscopic quality control indicators.

Key Words: Endoscopist biopsy rate; Endoscopist qualifications; Gastric precancerous conditions; Gastric precancerous lesions; Gastric cancer

Core Tip: This study demonstrates that endoscopists and endoscopic assistants with high qualifications, rather than the endoscopist biopsy rate, significantly improve the detection of gastric precancerous lesions (GPLs). Qualified endoscopists showed higher GPLs detection rates, particularly in the gastric cardia, angularis, and body regions. This suggests that a high endoscopist biopsy rate alone is not sufficient for detecting GPLs. Instead, greater emphasis should be placed on improving the qualifications of endoscopists and appropriate collaboration between endoscopists and assistants to perform accurate biopsies. These findings provide valuable insights for developing gastroscopic quality control standards.

INTRODUCTION

Gastric cancer (GC) remains a critical global public health challenge. Despite advancements in therapeutic strategies, the 5-year relative survival rate for GC remains suboptimal at 36%[1]. Current therapeutic paradigms integrate multimodal strategies, however, challenges including tumor heterogeneity and therapeutic resistance necessitate biomarker-driven precision strategies to achieve clinically meaningful improvements in survival[2]. In China, GC remains a common malignant tumor, with incidence and mortality rates ranking at the forefront among malignant tumors[3], imposing significant economic pressure on the health care system. The development of intestinal-type GC typically follows the Correa cascade model, from normal gastric mucosa to nonatrophic gastritis (NAG), atrophic gastritis (AG), intestinal metaplasia (IM), gastric dysplasia/intraepithelial neoplasia (Dys/IN), and ultimately, GC[4]. Gastric precancerous conditions (GPCs) include AG and IM[5], whereas gastric precancerous lesions (GPLs) include Dys/IN, which play crucial roles in malignant transformation of the gastric mucosa[6]. Therefore, the early identification of GPCs and GPLs is essential for improving patients’ quality of life and enhancing treatment outcomes.

By regularly undergoing gastroscopy screening, Japan and South Korea have successfully increased the diagnosis and treatment rates of GC and GPLs, improved patient prognosis, and effectively lowered the incidence and mortality of GC[7,8]. Gastroscopy combined with biopsy is the main method for detecting and monitoring GC and precancerous lesions. However, there is significant heterogeneity among different examiners, and the diagnosis rate often depends on the examiner[9]. A meta-analysis revealed a misdiagnosis rate of 9.4% for GC[10], highlighting the necessity to improve the quality of gastroscopy examinations and implement strict biopsy strategies in clinical practice.

The complex mucosal background of the stomach makes the diagnosis of GPLs and early GC (EGC) challenging[9]. The sensitivity of white light endoscopy in diagnosing GPLs is only 51%-74%[11,12]. Although new endoscopic technologies such as chromoendoscopy, narrow-band imaging, and magnifying endoscopy have shown certain advantages in identifying precancerous lesions in the stomach, advanced imaging techniques often require a certain level of expertise[13], and the diagnostic rate is limited by the endoscopist’s experience. Unnecessary biopsies not only cause trauma to patients but also increase the workload of pathologists. Therefore, the ability of endoscopists to identify lesions and obtain accurate samples is crucial[14-17], though there is currently limited research regarding the relationships between different endoscopist biopsy rates (EBRs) and the qualifications of endoscopists and lesion detection.

In China, endoscopic assistants are essential collaborators during gastroscopy examinations. Endoscopic assistants are required not only to master fundamental theoretical knowledge and technical skills, but also to demonstrate proficiency in operating endoscopic instruments and critical thinking capabilities, enabling them to deliver holistic care and comprehensive emergency nursing interventions[18]. Globally, endoscopic assistants play a pivotal role in gastrointestinal endoscopic examinations. The adaptability and operational autonomy of endoscopic assistants significantly contribute to optimizing healthcare resource utilization efficiency. Their performance has been shown to achieve patient satisfaction levels comparable to those of physicians[19]. However, there is very little research on the correlation between the qualifications of endoscopy assistants and the positive lesion detection rate of endoscopists.

This study analyzed 5 years of gastroscopy data from our center to explore the relationships between EBR, endoscopist qualifications, and positive lesion detection rate. We further analyzed the relationship between endoscopic assistants with different qualifications and the positive lesion detection rate. These findings provide reliable clinical evidence for improving the positive rate of gastroscopic biopsies and the development of endoscopic quality control indicators.

MATERIALS AND METHODS

Ethics

This study was reviewed and approved by the Ethics Committee of Wuxi People’s Hospital of Nanjing Medical University on July 17, 2023 (No. KY23001) and was conducted in strict accordance with the ethical principles of the Helsinki Declaration. This study has been registered in the Chinese Clinical Trial Registry with the registration number ChiCTR2400082985.

Inclusion criteria

This single-center retrospective cohort study included all patients who underwent gastroscopic and histopathological examinations of the cardia and stomach at Wuxi People’s Hospital from January 2018 to April 2023. All the endoscopists involved in the study received specialized training in upper gastrointestinal endoscopy prior to performing gastroscopy and were capable of independently conducting gastroscopic examinations and making diagnoses. All the endoscopic assistants involved in the study received specialized training in nursing care for upper gastrointestinal endoscopy prior to assisting in gastroscopic procedures and were capable of independently assisting endoscopists in completing gastroscopic examinations. Both Olympus CV290 gastroscope machines and Olympus GIF-HQ290 gastroscopes were used in this study.

Exclusion criteria

The exclusion criteria for this study were as follows: (1) Age younger than 18 years; (2) Incomplete gastroscopy examination data; (3) Esophageal or duodenal biopsy; (4) Endoscopic treatment such as endoscopic submucosal dissection, endoscopic mucosal resection, or surgical operation; and (5) Endoscopic ultrasonography examination.

Patient information

The research data were obtained from the endoscopic workstation of our unit and included the following.

General information: Patient identification number, hospital registration number, sex, and age.

Gastroscopy information: Examination date, findings, diagnosis, endoscopist, and endoscopic assistant.

Pathological information: Pathological examination number, name of requesting endoscopist, name of pathological reviewing endoscopist, site and quantity of specimens submitted, gross findings of submitted specimens, microscopic findings of submitted specimens, and pathological diagnosis.

Pathological diagnosis, disease definition and grouping

All gastric biopsy pathologies were reviewed by two senior pathologists. In the case of a disagreement regarding the diagnosis, a third, more senior pathologist made the final diagnosis. Diagnoses were defined as follows.

NAG: An inflammatory reaction of the surface layer of the gastric mucosa, without accompanying mucosal atrophy or other epithelial lesions, that endoscopically manifests as redness or edema of the gastric mucosa[20].

AG: A disease characterized by a gradual reduction in and atrophy of the intrinsic glands of the gastric mucosa. Endoscopically, the gastric mucosa appears with a red and white alternating color, with white predominance. The gastric mucosa becomes thinner, with some mucosal blood vessels exposed, and the folds may become flat or disappear, accompanied by mucosal granules or nodular manifestations[21].

IM: Pathological changes in which the gastric mucosal epithelial cells are replaced by intestinal-type epithelial cells[22] are classified as late changes in AG, “light blue crests” can be observed via high-definition staining and magnifying endoscopy[23], and metaplastic atrophy confirmed by pathology is a reliable indicator for diagnosing gastric mucosal atrophy[5].

Dys/IN: A key stage before the occurrence of GC, Dys/IN is characterized by cells of varying sizes and shapes, enlarged cell nuclei, increased nuclear-cytoplasmic ratio, and coarse chromatin. Endoscopically, it may present as densely packed glands, increasingly distorted structures, irregular microvascular patterns, and “acanthosis nigricans appearance”[5,24]. Dys includes low-grade Dys, moderate-grade Dys, and high-grade Dys. IN includes low-grade IN and high-grade IN.

GC: Malignant tumor originating from the epithelium of the gastric mucosa. The main pathological type is adenocarcinoma, where cancer cells can form gland-like structures of varying sizes, irregular shapes, and arrangements with varying degrees of nuclear atypia. It includes EGC limited to the mucosal layer and submucosal layer, as well as advanced GC that extends beyond the submucosal layer. Histologically, it can be classified into tubular adenocarcinoma, mucinous adenocarcinoma, poorly cohesive carcinoma, and other rare types[25].

Other tumors: Epithelial tumors and nonepithelial tumors. Epithelial tumors include lymphomas, and nonepithelial tumors include neuroendocrine tumors, stromal tumors, and smooth muscle tumors.

Research parameters and definitions

Total number of endoscopist examinations: The total number of gastroscopies completed by endoscopists from 2006 (endoscopy workstation data traceable to 2006) to April 2023.

Qualifications evaluation and grouping: Endoscopists are required to meet certain procedural thresholds to achieve specific objective skill criteria in order to obtain the qualification for independent endoscopic procedures[26-28]. The technical skills and experience of endoscopists often require extensive hands-on practice[28]. Previous studies have also used the number of endoscopic procedures performed to assess the qualifications of endoscopists[29-32]. For example, in Yuan et al’s study, performing over 7000 endoscopic procedures was considered high experience, while performing over 1000 procedures was considered low experience[29]. Januszewicz et al[33] reported using quartiles to grade the biopsy rates of endoscopists. Therefore, in this study, the number of procedures by endoscopists/assistants was chosen as the standard for assessing the qualifications of endoscopists/assistants, and the quartile method was used to categorize qualifications into four levels: Low-, moderate-, high-, and very high.

EBR: The proportion of gastroscopies in which endoscopists perform biopsies during gastroscopy.

EBR grouping: Endoscopists are divided into low-, moderate-, high-, and very high-EBR groups on the basis of the quartile distribution of the endoscopists’ EBR values[33].

Positive lesions: Pathological diagnoses include AG, IM, Dys/IN, GC, and other tumor lesions.

Positive detection rate/positive biopsy rate: The proportion of patients with positive lesions among those who underwent biopsy to the total number of patients who were biopsied.

The negative biopsy rate: The number of cases in which pathological diagnosis did not detect AG/IM, Dys/IN, GC, or other tumors as a proportion of the total number of biopsy cases[33].

Detection rate of lesions in different areas: During gastroscopy, when biopsies are taken in different areas, such as the cardia, gastric fundus, gastric body, gastric angle, gastric antrum, and pylorus, the number of positive lesions detected is the proportion of the total number of biopsies in that area.

Statistical analysis

Baseline data are described using medians, interquartile ranges, and contingency tables in this study. The distributions of various pathological diagnoses and biopsy sites are depicted using percentage pie charts. The distribution of pathological diagnoses for each biopsy site is described using percentage stacked bar charts. The Shapiro-Wilk test was used for normality testing, followed by the Pearson correlation coefficient and Spearman correlation coefficient to measure the relationships between EBR, qualifications, and the detection rates of various lesions and negative biopsy rates. A multivariate logistic regression model was used to assess the association between the EBR group and GPCs and GPLs, adjusting for patient gender and age at diagnosis. The model included these variables as covariates to estimate the odds ratios (ORs) and 95% confidence intervals (CIs), reflecting the independent relationship between the EBR group and the outcomes, while controlling for potential confounding effects of gender and age. All statistical tests were two-tailed, and P < 0.05 was considered statistically significant. All analyses were performed using SPSS 27.0 software.

RESULTS

Research process diagram

This study included a total of 200910 patients who underwent gastroscopy at our center from January 2018 to April 2023. Among them, 903 patients (0.45%) were younger than 18 years old; 5485 patients (2.73%) had incomplete gastroscopy data; 23561 patients (11.73%) underwent endoscopic submucosal dissection, endoscopic mucosal resection, or surgical treatment; 4750 patients (2.36%) underwent endoscopic ultrasonography; and 2279 patients (1.13%) underwent biopsy histopathology of the esophagus and duodenum. Therefore, total of 36981 patients (18.41%) were excluded, and 169417 patients were included in the final analysis. Biopsy specimen from 45805 patients underwent histopathological examination. The study flowchart is shown in Figure 1.

Figure 1 Study population flowchart. A total of 200910 patients underwent gastroscopy at our center from January 2018 to April 2023 Among them, 903 patients were younger than 18 years old; 5485 patients had incomplete gastroscopy data; 23561 patients underwent endoscopic submucosal dissection, endoscopic mucosal resection, or surgical treatment; 4750 patients underwent endoscopic ultrasonography; and 2279 patients underwent biopsy histopathology of the esophagus and duodenum. Exclusion criteria and reasons: (1) This study focuses on adult patients, so individuals under 18 years old were excluded; (2) To ensure the completeness and reliability of the data for analysis, patients with incomplete gastroscopy examination data were excluded; (3) This study only focuses on gastric biopsy-related issues, so esophageal or duodenal biopsies were excluded; (4) Gastroscopic treatments such as endoscopic submucosal dissection, endoscopic mucosal resection, or gastric surgery were excluded, as these procedures may alter the normal anatomical structure of the stomach and affect the data analysis; and (5) Endoscopic ultrasonography was excluded to avoid introducing additional variables that could interfere with the study’s findings. A total of 36981 patients were excluded, and 169417 patients were included in the analysis. The biopsy specimen of 45805 patients underwent histopathological examination. EUS: Endoscopic ultrasonography; EBR: Endoscopist biopsy rate; GPCs: Gastric precancerous conditions; GPLs: Gastric precancerous lesions; GC: Gastric cancer; NAG: Nonatrophic gastritis.

Baseline data, biopsy pathological diagnosis, and distribution of lesion sites

The median (quartile) age of the patients was 55 (46, 64) years, with an age range of 18-95 years. The proportion of males was greater than that of females (males = 53.39%, females = 46.61%), and the largest population undergoing gastroscopy was the 50-69 years age group. There were 1696 male (3.7%) and 886 (1.93%) female patients with GPLs, and the number of male patients with GPLs was 1.9 times greater than the number of female patients. A total of 872 male (1.9%) and 386 female (0.84%) patients had GC, and the number of male patients with GC was 2.26 times greater than the number of female patients (Table 1). A total of 53584 biopsies were performed (Table 2), including 2135 in the Cardia (3.98%), 1730 in the gastric fundus (3.23%), 13599 in the gastric body (25.38%), 8583 in the gastric angle (16.02%), 27201 in the gastric antrum (50.76%), and 336 in the pylorus (0.63%). The highest number of biopsies was performed in the gastric antrum (Figure 2A). Pathology revealed that NAG accounted for 28607 cases (62.45%), AG accounted for 13274 cases (28.92%), GPLs accounted for 2582 cases (5.64%), GC accounted for 1258 cases (2.93%), and other tumor lesions accounted for 84 cases (0.18%) (Figure 2B). The detection rates of GC at each site were as follows: Cardia, 6.79%; pylorus, 3.57%; gastric angle, 3.39%; gastric body, 3.33%; gastric antrum, 1.57%; and gastric fundus, 1.16%, with the highest incidence in the cardia and the lowest incidence in the gastric fundus. The detection rates of GPLs at each site were as follows: Gastric angle, 9.04%; cardia, 6.28%; gastric antrum, 5.41%; pylorus, 4.17%; gastric body, 2.37%; and gastric fundus, 0.75%, with the highest incidence in the gastric angle and the lowest incidence in the gastric fundus. The detection rates of AG at each site were as follows: Gastric angle, 48.56%; pylorus, 31.25%; gastric antrum, 30.47%; cardia, 24.96%; gastric body, 13.63%; and gastric fundus, 1.73%, with the highest incidence in the gastric angle and the lowest incidence in the gastric fundus (Figure 2C).

Figure 2 Proportion of biopsies from different sites and distribution of gastric diseases. A: The proportions of biopsies from different sites are shown; B: The proportions of different diseases are shown. The horizontal axis represents different diseases, and the vertical axis represents different proportions (%); C: The proportions of different diseases in different regions of the stomach are shown. The horizontal axis represents different regions of the stomach, and the vertical axis represents different proportions (%). NAG: Nonatrophic gastritis; AG/IM: Atrophic gastritis/intestinal metaplasia; GPLs: Gastric precancerous lesions; GC: Gastric cancer.

Table 1 Baseline data and proportions of pathological diagnoses by biopsy site, n (%).

	Gastroscopes	NAG, %	AG/IM, %	GPLs, %	GC, %	Other tumors, %
Total	45805 (100)	62.45	28.98	5.64	2.75	0.18
Age groups (median age, 55 years; range: 18-95 years)
18-49 years	15005 (32.76)	25.13	6.38	0.94	0.28	0.03
50-69 years	25037 (54.66)	31.69	18.02	3.53	1.32	0.10
≥ 70 years	5763 (12.58)	5.64	4.58	1.17	1.15	0.05
Sex
Male	24455 (53.39)	31.56	16.16	3.70	1.90	0.06
Female	21350 (46.61)	30.89	12.82	1.93	0.84	0.12

NAG: Nonatrophic gastritis; AG/IM: Atrophic gastritis/intestinal metaplasia; GPLs: Gastric precancerous lesions; GC: Gastric cancer.

Table 2 Distribution of diseases in different regions of the stomach, n (%).

	Total	NAG, %	AG/IM, %	GPLs, %	GC, %	Other tumors, %
Total	53584 (100)	64.22	27.94	5.09	2.71	0.17
Cardia	2135 (3.98)	2.47	0.99	0.25	0.27	0.00
Gastric body	13599 (25.38)	20.36	3.46	0.60	0.96	0.10
Antrum	27201 (50.76)	31.71	15.47	2.75	0.84	0.04
Angularis	8583 (16.02)	6.23	7.78	1.45	0.57	0.02
Fundus	1730 (3.23)	3.09	0.06	0.02	0.05	0.01
Pylorus	336 (0.63)	0.38	0.20	0.03	0.02	0.00

NAG: Nonatrophic gastritis; AG/IM: Atrophic gastritis/intestinal metaplasia; GPLs: Gastric precancerous lesions; GC: Gastric cancer.

Correlation between EBR and the detection rate of positive lesions

The total number of examinations performed by 37 endoscopists ranged from 1466 to 132205, with a median of 17985 examinations. According to the quartiles of the total number of examinations, the endoscopists were divided into groups as follows: Low-level experience group (1466-7379 examinations, 10 endoscopists), moderate-level experience group (7380-17985 examinations, 9 endoscopists), high-level experience group (17986-25385 examinations, 9 endoscopists), and very high-level experience group (25386-132205 examinations, 9 endoscopists). The negative biopsy rate of each endoscopist ranged from 40.50%-72.41%, with a median negative biopsy rate of 63.09%. The EBR values of the 37 endoscopists ranged from 13.94%-40.17%, with a median EBR value of 25.98%. According to the quartiles of EBR values, the endoscopists were grouped as follows: Low-EBR group (13.94%-23.44%, 10 endoscopists), moderate-EBR group (23.45%-25.98%, 9 endoscopists), high-EBR group (25.99%-31.39%, 9 endoscopists), and very high-EBR group (31.40%-40.17%, 9 endoscopists). Compared with those in the low-EBR group, the OR values for detecting positive lesions in the moderate-, high-, and very high-EBR groups were 1.12 (95%CI: 1.06-1.19, P < 0.001), 1.22 (95%CI: 1.14-1.31, P < 0.001), and 1.38 (95%CI: 1.29-1.47, P < 0.001), respectively, indicating that as the EBR increased, the detection rate of positive lesions also increased. Compared with those in the low-EBR group, the OR values for detecting GPCs in the moderate-, high-, and very high-EBR groups were 1.16 (95%CI: 1.09-1.24, P < 0.001), 1.28 (95%CI: 1.19-1.38, P < 0.001), and 1.48 (95%CI: 1.38-1.58, P < 0.001), respectively, indicating that as the EBR increased, the detection rate of GPCs also increased. In contrast, the OR values for detecting GPLs in the moderate-, high-, and very high-EBR groups were 0.98 (95%CI: 0.87-1.11, P = 0.79), 1.11 (95%CI: 0.96-1.27, P = 0.159), and 1.10 (95%CI: 0.97-1.25, P = 0.138), respectively, indicating no significant difference in the detection rate of GPLs as the EBR increased. The OR values for negative biopsies in the moderate-, high-, and very high-EBR groups were 0.89 (95%CI: 0.84-0.95, P < 0.001), 0.82 (95%CI: 0.76-0.88, P < 0.001), and 0.73 (95%CI: 0.68-0.77, P < 0.001), respectively, indicating that as the EBR increased, the negative biopsy rate decreased (Table 3).

Table 3 Relationship between endoscopist biopsy rate and positive/negative lesion detection, n (%).

Endoscopist	EBR	Gastroscopes	GPCs	GPLs	Negative biopsies	EBR group	OR for GPCs, 95%CI	OR for GPLs, 95%CI	OR for negative biopsies, 95%CI
1	13.94%	1466	19 (21.84)	0 (0)	63 (72.41)	Low EBR (13.94%-23.44%)	1.00, -	1.00, -	1.00, -
2	17.06%	1668	29 (23.58)	3 (2.44)	86 (69.92)
3	18.95%	108865	527 (29.28)	136 (7.56)	1052 (58.44)
4	21.23%	5268	118 (25.93)	22 (4.84)	303 (66.59)
5	21.53%	32126	347 (24.47)	52 (3.67)	968 (68.27)
6	21.96%	19484	238 (24.64)	52 (5.38)	645 (66.77)
7	22.15%	16212	280 (23.08)	66 (5.44)	820 (67.6)
8	22.30%	7349	139 (25.23)	37 (6.72)	357 (64.79)
9	23.16%	35092	277 (21.9)	64 (5.06)	883 (69.8)
10	23.44%	39896	487 (28.15)	113 (6.53)	1083 (62.6)
11	23.48%	10006	267 (30.34)	37 (4.2)	551 (62.61)	Moderate EBR (23.45%-25.98%)	1.16, 1.09-1.24	0.95, 0.86-1.05	0.89, 0.84-0.94
12	23.67%	6820	189 (28.29)	31 (4.64)	429 (64.22)
13	23.68%	16172	311 (23.52)	46 (3.48)	924 (69.89)
14	23.72%	3992	103 (29.26)	15 (4.26)	222 (63.07)
15	24.18%	18608	350 (29.79)	52 (4.43)	737 (62.72)
16	24.42%	27558	415 (27.36)	103 (6.79)	957 (63.09)
17	24.80%	9816	279 (31.89)	51 (5.83)	518 (59.2)
18	24.96%	132205	1638 (27.42)	358 (5.99)	3717 (62.23)
19	25.98%	27800	423 (29.27)	131 (9.07)	846 (58.55)
20	26.35%	4666	171 (36.85)	24 (5.17)	255 (54.96)	High EBR (25.99%-31.39%)	1.28, 1.19-1.38	0.94, 0.84-1.05	0.82, 0.76-0.88
21	27.49%	5890	185 (29.55)	21 (3.35)	398 (63.58)
22	28.15%	22349	409 (30.16)	61 (4.5)	855 (63.05)
23	28.37%	17985	383 (27.81)	71 (5.16)	887 (64.42)
24	28.90%	12956	412 (27.93)	86 (5.83)	947 (64.2)
25	29.65%	18638	426 (30.45)	82 (5.86)	864 (61.76)
26	30.16%	18413	292 (26.67)	50 (4.57)	732 (66.85)
27	31.08%	19148	397 (33.5)	64 (5.4)	698 (58.9)
28	31.39%	25385	379 (29.61)	57 (4.45)	829 (64.77)
29	32.55%	18142	421 (43.49)	126 (13.02)	392 (40.5)	Very high EBR (31.40%-40.17%)	1.48, 1.38-1.58	0.9, 0.81-1	0.73, 0.68-0.77
30	32.58%	26041	460 (31.68)	65 (4.48)	888 (61.16)
31	32.69%	7379	244 (28.91)	50 (5.92)	519 (61.49)
32	33.09%	12300	418 (31.19)	74 (5.52)	812 (60.6)
33	33.29%	35746	748 (36.65)	129 (6.32)	1121 (54.92)
34	33.83%	8756	410 (33.61)	71 (5.82)	718 (58.85)
35	34.89%	23507	488 (29.24)	81 (4.85)	1055 (63.21)
36	35.98%	1642	72 (24.83)	5 (1.72)	207 (71.38)
37	40.17%	12100	523 (27.32)	96 (5.02)	1269 (66.3)

EBR: Endoscopist biopsy rate; GPCs: Gastric precancerous conditions; GPLs: Gastric precancerous lesions; OR: Odds ratio; CI: Confidence interval.

We further analyzed the correlation between EBR and endoscopist qualifications. There was no statistically significant correlation between EBR and endoscopist qualifications (ρ = 0.044, P = 0.796), indicating that the EBR was similar among endoscopists with different qualifications (Figure 3A). Moreover, the EBR value was positively correlated with the GPC detection rate (ρ = 0.465, P = 0.004), though the correlation with the GPL detection rate was not statistically significant (ρ = 0.141, P = 0.406). In contrast, the EBR value was negatively correlated with the GC detection rate (r = -0.728, P < 0.001) and negative biopsy rate (ρ = -0.389, P = 0.017) (Figure 3B).

Figure 3 Relationships between endoscopist biopsy rate, endoscopist qualifications and endoscopic detection. A: The relationship between endoscopist biopsy rate (EBR) and endoscopist qualifications is shown. The horizontal axis represents different groups of EBR (low, moderate, high/very high), and the vertical axis represents different groups of endoscopist qualifications (low, moderate, high/very high); B: The relationship between EBR and the positive detection rate is shown. The horizontal axis represents different EBR groups (low, moderate, high/very high), and the vertical axis represents endoscopic detection rates for different diseases (%). ρ: Spearman’s correlation coefficient. P < 0.05 indicates statistical significance. EBR: Endoscopist biopsy rate; GPCs: Gastric precancerous conditions; GPLs: Gastric precancerous lesions.

Correlation between endoscopist qualifications and the detection rate of positive lesions

According to the multivariable logistic regression analysis data, compared with those in the low-seniority endoscopist group, the OR values for detecting GPCs in the moderate-, high-, and very high-seniority endoscopist groups were 0.92 (95%CI: 0.84-1.00, P = 0.059), 1.03 (95%CI: 0.94-1.12, P = 0.57), and 0.98 (95%CI: 0.91-1.07, P = 0.663), respectively, indicating no significant difference in the detection rate of GPCs among endoscopists with different seniority levels. In contrast, the OR values for detecting GPLs in the moderate-, high-, and very high-seniority groups were 1.13 (95%CI: 0.98-1.31, P = 0.03), 1.20 (95%CI: 1.03-1.39, P = 0.003), and 1.32 (95%CI: 1.15-1.52, P < 0.001), respectively. The OR values for negative biopsies in the moderate-, high-, and very high-seniority groups were 1.04 (95%CI: 0.95-1.13, P = 0.387), 0.92 (95%CI: 0.85-1.00, P = 0.059), and 0.93 (95%CI: 0.86-1.00, P = 0.054), respectively (Table 4). Compared with low qualification level endoscopists, those with moderate, high, and very high qualification levels increased detection rates of precancerous lesions by 13% (OR = 1.13, 95%CI: 0.98-1.31), 20% (OR = 1.20, 95%CI: 1.03-1.39), and 32% (OR = 1.32, 95%CI: 1.15-1.52), respectively.

Table 4 Relationship between endoscopists qualifications and positive/negative lesion detection, n (%).

Endoscopist	EBR	Gastroscopes	GPCs	GPLs	Negative biopsies	Qualification group	OR for GPCs, 95%CI	OR for GPLs, 95%CI	OR for negative biopsies, 95%CI
1	13.94%	1466	19 (21.84)	0 (0)	63 (72.41)	Low qualification (1466-7379)	1.00, -	1.00, -	1.00, -
36	35.98%	1642	72 (24.83)	5 (1.72)	207 (71.38)
2	17.06%	1668	29 (23.58)	3 (2.44)	86 (69.92)
14	23.72%	3992	103 (29.26)	15 (4.26)	222 (63.07)
20	26.35%	4666	171 (36.85)	24 (5.17)	255 (54.96)
4	21.23%	5268	118 (25.93)	22 (4.84)	303 (66.59)
21	27.49%	5890	185 (29.55)	21 (3.35)	398 (63.58)
12	23.67%	6820	189 (28.29)	31 (4.64)	429 (64.22)
8	22.30%	7349	139 (25.23)	37 (6.72)	357 (64.79)
31	32.69%	7379	244 (28.91)	50 (5.92)	519 (61.49)
34	33.83%	8756	410 (33.61)	71 (5.82)	718 (58.85)	Moderate qualification (7380-17985)	0.92, 0.84-1	1.13, 0.98-1.31	1.04, 0.95-1.13
17	24.80%	9816	279 (31.89)	51 (5.83)	518 (59.2)
11	23.48%	10006	267 (30.34)	37 (4.2)	551 (62.61)
37	40.17%	12100	523 (27.32)	96 (5.02)	1269 (66.3)
32	33.09%	12300	418 (31.19)	74 (5.52)	812 (60.6)
24	28.90%	12956	412 (27.93)	86 (5.83)	947 (64.2)
13	23.68%	16172	311 (23.52)	46 (3.48)	924 (69.89)
7	22.15%	16212	280 (23.08)	66 (5.44)	820 (67.6)
23	28.37%	17985	383 (27.81)	71 (5.16)	887 (64.42)
29	32.55%	18142	421 (43.49)	126 (13.02)	392 (40.5)	High qualification (17986-25385)	1.03, 0.94-1.12	1.20, 1.03-1.39	0.92, 0.85-1.00
26	30.16%	18413	292 (26.67)	50 (4.57)	732 (66.85)
15	24.18%	18608	350 (29.79)	52 (4.43)	737 (62.72)
25	29.65%	18638	426 (30.45)	82 (5.86)	864 (61.76)
27	31.08%	19148	397 (33.5)	64 (5.4)	698 (58.9)
6	21.96%	19484	238 (24.64)	52 (5.38)	645 (66.77)
22	28.15%	22349	409 (30.16)	61 (4.5)	855 (63.05)
35	34.89%	23507	488 (29.24)	81 (4.85)	1055 (63.21)
28	31.39%	25385	379 (29.61)	57 (4.45)	829 (64.77)
30	32.58%	26041	460 (31.68)	65 (4.48)	888 (61.16)	Very high qualification (25386-132205)	0.98, 0.91-1.07	1.32, 1.15-1.52	0.93, 0.86-1.00
16	24.42%	27558	415 (27.36)	103 (6.79)	957 (63.09)
19	25.98%	27800	423 (29.27)	131 (9.07)	846 (58.55)
5	21.53%	32126	347 (24.47)	52 (3.67)	968 (68.27)
9	23.16%	35092	277 (21.9)	64 (5.06)	883 (69.8)
33	33.29%	35746	748 (36.65)	129 (6.32)	1121 (54.92)
10	23.44%	39896	487 (28.15)	113 (6.53)	1083 (62.6)
3	18.95%	108865	527 (29.28)	136 (7.56)	1052 (58.44)
18	24.96%	132205	1638 (27.42)	358 (5.99)	3717 (62.23)

EBR: Endoscopist biopsy rate; GPCs: Gastric precancerous conditions; GPLs: Gastric precancerous lesions; OR: Odds ratio; CI: Confidence interval.

Correlation analysis revealed that the qualifications of endoscopists were positively correlated with the detection rate of GPLs (ρ = 0.448, P = 0.005), with no statistically significant correlation with the detection rate of GPCs (ρ = 0.288, P = 0.084) or GCs (ρ = -0.064, P = 0.709) or with the negative biopsy rate (ρ = -0.293, P = 0.079) (Figure 4A). The data above indicate that the greater the degree of seniority of the endoscopist, the higher the detection rate of GPLs.

Figure 4 Relationships between endoscopists’ and endoscopic assistants’ qualifications and endoscopic detection. A: The relationship between endoscopist qualifications and endoscopic detection is shown. The horizontal axis represents different qualifications of the endoscopist (low, moderate, high/very high), and the vertical axis represents endoscopic detection rate of different diseases; B: The relationship between the endoscopists’ qualifications and the detection rates of gastric precancerous lesions in different anatomical areas of the stomach is shown. The horizontal axis represents different qualifications of the endoscopist (low, moderate, high/very high), and the vertical axis represents gastroscopic detection rate of gastric precancerous lesions; C: The relationship between endoscopic assistant qualifications and endoscopic detection is shown. The horizontal axis represents different qualifications of the endoscopist (low, moderate, high/very high), and the vertical axis represents gastroscopic detection by low- and moderate-level qualified endoscopists. ρ: Spearman’s correlation coefficient. P < 0.05 indicates statistical significance. GPCs: Gastric precancerous conditions; GPLs: Gastric precancerous lesions.

We further analyzed the relationship between the qualifications of endoscopists and the detection rates of GPLs in different areas of the stomach. The results showed that the qualifications of endoscopists were positively correlated with the detection rates of GPLs in the cardia (ρ = 0.350, P = 0.034), angularis (ρ = 0.396, P = 0.015), and gastric body (ρ = 0.453, P = 0.005). However, no statistically significant correlation was found between the qualifications of endoscopists and the detection rates of GPLs in the antrum (ρ = 0.292, P = 0.079) (Figure 4B). Due to the very low biopsy rates in the fundus and pylorus, which were 3.23% and 0.63%, respectively, and the even fewer cases of GPLs positivity, no further correlation analysis was conducted for these sites in this study. These data indicate that the endoscopists qualifications were particularly correlated with the detection rates of GPLs in the cardia, angularis and gastric body.

Relationship between endoscopic assistant qualifications and gastroscopic detection by endoscopists with low- and moderate-level qualifications

An additional 752 cases lacking endoscopic assistant data were excluded from this analysis. A total of 45381 examinations were assisted by 32 endoscopic assistants (ranging from 445 to 45381 per assistant), with a median examination volume of 12264.5 examinations. The correlation analysis revealed that in gastroscopic examinations conducted by endoscopic assistants with different levels of qualification, the qualifications of endoscopic assistants were closely related to the detection rates of low- and moderate-quality endoscopists for GPLs (ρ = 0.427, P = 0.015). However, there was no statistically significant correlation with the negative biopsy rate of endoscopists (ρ = -0.306, P = 0.088), the detection rate of GPCs (ρ = 0.148, P = 0.419), or the detection rate of GC (ρ = 0.047, P = 0.799) (Figure 4C). In contrast, there was no correlation with the detection rate of GPCs (ρ = 0.187, P = 0.305), GPLs (ρ = 0.254, P = 0.161), GC (ρ = -0.169, P = 0.356), or the negative biopsy rate (ρ = 0.034, P = 0.855) of highly/very highly qualified endoscopists (data not shown). These findings indicate that high/very high-quality endoscopic assistants may increase the detection rate of GPLs by low- and moderate-quality endoscopists.

DISCUSSION

GPCs and GPLs are independent risk factors for GC, providing a potential pathological basis for GC[34]. A follow-up study in South Korea in 2017 involving 3714 patients diagnosed with AG for up to 6.9 years revealed that the incidence of GC progression in patients with mild, moderate, and severe AG was 1.6%, 5.2%, and 12.0%, respectively, with the presence of IM further increasing the risk of GC[35]. A meta-analysis in 2018 also revealed a greater risk of GC development in patients with IM[35]. The risks of progression to GC in patients with confirmed low-grade IN and high-grade IN are 2.8%-11.5% and 10%-68.8%, respectively[36-40]. Therefore, early identification and intervention of GPCs and GPLs are highly clinically important. Patients with GPCs and GPLs often lack specific symptoms, and gastroscopy combined with pathological biopsy is the gold standard for diagnosis. However, the correlation between histological changes and endoscopic findings in the diagnosis and monitoring of GPCs and GPLs is often poor, with the sensitivity of white light endoscopy in diagnosing AG being only 42%[41] and even lower at 24% for IM[42] and 51%-74% for GPLs[11,12]. In addition, there is a certain degree of pathological improvement between the pathology results of endoscopic biopsy and those of endoscopic resection of GPLs[14-17]. Unnecessary biopsies not only increase patient trauma[43] and economic burden but also add to the workload of pathologists. Systematic training can significantly improve the ability of endoscopists to diagnose and grade lesions[44]. As endoscopists use discretion and subjectivity to decide whether and where to biopsy, their ability to identify lesions and accurately sample them is crucial for the accurate detection of lesions[14-17,45]. Therefore, understanding the relationships between the EBR and the experience of endoscopists; between EBR and the experience of endoscopists and the detection rate of positive lesions, and the role of endoscopist assistants in the detection rate of positive lesions by endoscopists is highly valuable for guiding rational biopsies and improving strategies for increasing the detection rate of positive lesions during gastroscopy.

This study revealed that AG accounted for the highest proportion (28.92%) of histopathological diagnoses. A cross-sectional survey in 2014 revealed that the pathological detection rate of AG was 25.8%[41], whereas a retrospective analysis in 2016 reported that the detection rate of CAG in patients over 35 years of age who underwent gastroscopy was 22.4%[46]. The detection rate of AG in this study was slightly higher than that in previous studies, which may be related to the increasing incidence of chronic gastritis in our country[46,47]. In addition, AG is more common in elderly individuals in different regions of the world, and its prevalence gradually increases with age[41]. In this study, the prevalence of AG in patients aged 18-49, 50-69, and ≥ 70 years was 19.47%, 32.98%, and 36.39%, respectively, showing a clear trend of increasing with age. The detection rate of GC was 2.26 times greater and that of GPLs was 1.9 times greater in males than in females. These findings are consistent with the results of several previous studies that revealed that the incidence of GC in males is 2-2.7 times greater than that in females[41,48,49].

Lesions associated with gastric mucosal atrophy are often located in the gastric antrum or corpus[5], with the corpus and antrum being common sites for GC. A consensus[5,47,50] on the biopsy strategy for gastric mucosal lesions recommends biopsies at the antrum, corpus, and angle of the stomach to ensure a comprehensive evaluation of the gastric mucosa. In the present study, most of the biopsies were obtained from the gastric antrum (50.76%), corpus (25.38%), and angle (16.02%), accounting for 92.16% of all biopsy sites. In terms of the detection rate of GC at different biopsy sites, the corpus (453 biopsies, 0.85%) and antrum (426 biopsies, 0.80%) had the highest rates; in terms of the detection rate of GPLs, the angle (1472 biopsies, 2.75%) had the highest rate; and in terms of the detection rate of GC, the cardia region had the highest rate (6.79%). Despite recommended biopsy strategies, the selection of biopsy sites in clinical practice may be influenced by several factors, including the experience of the endoscopist and the visual assessment of lesions. Therefore, there are significant differences in biopsy site proportions, emphasizing the importance of standardizing biopsy strategies to ensure that all patients are treated using a standardized diagnostic process. A study of patients with GC who were treated within 25 years revealed that the cardia was the second most common site for GC after the antrum, with an increasing trend in the incidence of cardia cancer[51]. This study also revealed that, compared with other gastric regions, the cardia had the highest detection rate of GC. Therefore, even though the cardia is not a primary biopsy site recommended by consensus, it should be emphasized in observation and biopsy during gastroscopy procedures.

Upper gastrointestinal endoscopy, the earliest digestive endoscopy technique that is used clinically, plays an essential role in tumor screening and GPL follow-up. This approach is highly important for ensuring the quality control of upper gastrointestinal endoscopy. In recent years, quality control indicators for upper gastrointestinal endoscopy examinations have received widespread attention. However, unlike colonoscopy, which has multiple established quality control indicators, high-quality validation of quality control indicators for upper gastrointestinal endoscopy remains lacking. EBR, as an important indicator in the quality assessment system for upper gastrointestinal endoscopy, has been the subject of several studies examining its relationship with the detection rates of GC and precancerous lesions. A multicenter study found that EBR is closely related to the detection rates of CAG, IM, and Dys, and it can reduce the rate of missed diagnoses[33]. Another study also demonstrated a close correlation between EBR and the detection rate of GC[52]. Furthermore, the expertise of endoscopists serves as another pivotal factor in ensuring the quality of upper gastrointestinal endoscopic examinations. Experienced endoscopists are capable of executing essential procedural steps during gastroscopy, including comprehensive mucosal visualization, precise lesion identification, and appropriate biopsy sampling. Such proficiency enables them to conduct higher-quality examinations, thereby effectively minimizing the occurrence of missed lesions[53-55]. In this study, the overall EBR at our center was 26.68%, with significant differences between the EBR achieved by different endoscopists, ranging from 13.94%-40.17%. A retrospective study conducted in Japan in 1998 reported an average gastric EBR of 55%[56]; a study conducted in South Korea in 2017 reported that the EBR among endoscopists ranged from 6.9%-27.8%[52]; and a multicenter study conducted in Poland in 2019 reported significant differences in EBR among different endoscopists in two high-volume outpatient centers (22.4%-65.8%)[33]. A study evaluating the quality of upper gastrointestinal endoscopy nationwide in Italy in 2023 reported that only 32.7% of included patients did not undergo biopsy, and 50.5% of patients had adequate biopsy sampling (at least two biopsy samples from both the gastric antrum and corpus)[57]. Therefore, there are certain differences in the EBR among endoscopists in different countries, which may be related to the incidence of diseases, endoscopists’ understanding of diseases, and differences in gastroscopy techniques. In reports from Japan, Poland, and Italy, patients are mostly referred to or evaluated by general practitioners for upper gastrointestinal symptoms, unlike the asymptomatic screening population in South Korea; therefore, the EBR may be greater. The study population in our cohort mainly consisted of patients in outpatient/inpatient settings who required evaluation of upper gastrointestinal symptoms with a subset of the population undergoing asymptomatic screening; therefore, the overall EBR includes the EBR for asymptomatic screening and the EBR for the evaluation of upper gastrointestinal symptoms. Gastroscopy, an examination that is dependent on the operator’s ability, results in significant differences in examination quality among different operators. Only endoscopists who have undergone proper training and possess relevant capabilities can independently perform upper gastrointestinal endoscopy[58,59].

A previous study reported a significant correlation between EBR and the tumor detection rate (R² = 0.76; P = 0.0015)[52]. A multicenter study in 2019 revealed the importance of EBR for diagnosing GPCs and GPLs, especially with a significant correlation with the total detection rates of AG, IM, and GPLs (ρ = 0.83, P < 0.001), and that endoscopists with a higher EBR have a lower risk of missing cancer during gastroscopy[33]. This study revealed a positive correlation between EBR and the detection rate of positive lesions. Further analysis revealed that EBR was positively correlated with only the detection rate of GPCs, whereas the correlation with the detection rate of GPLs was not statistically significant. This may be related to the higher prevalence of GPCs in our study population. However, biopsies are more important for detecting and diagnosing GPLs and GC. Therefore, clinicians should not only emphasize EBR but also focus on the experience of the endoscopist for identifying high-risk lesions and optimizing EBR to perform rational biopsies that improve the detection rate of GPLs and GC. We also observed a negative correlation between EBR and the rate of negative biopsies, indicating that an increase in EBR may reduce the rate of negative biopsies and the risk of missed diagnoses. These results indicate that EBR is a parameter worthy of attention in gastroscopy and that clinicians should not emphasize only EBR. However, further research is needed to determine the optimal EBR practice standards to improve the detection rate of GPLs and GC.

The qualifications of endoscopists are often evaluated on the basis of whether they have reached specific standards in terms of the number of upper gastrointestinal endoscopies or various surgical procedures performed. In addition, key performance indicators, such as the adenoma detection rate in colonoscopy and the selective duct cannulation rate in endoscopic retrograde cholangiopancreatography, are also important supplementary measures of evaluation. These assessment methods have been widely recognized and applied both domestically and internationally[58]. The present study also adopted this method and used the total number of examinations completed by endoscopists as the basis for assessing endoscopist qualifications. In this study, there were differences in the qualifications of gastroscopy endoscopists, and further analysis revealed that there was a positive correlation between the qualifications of gastroscopy endoscopists and the detection rate of GPLs. Therefore, the importance of gastroscopy endoscopists’ qualifications is evident. When performing gastroscopy, the endoscopist must possess a high level of professional skill to evaluate lesions comprehensively and perform accurate biopsies. This requires endoscopists to enhance their practice and accumulate rich diagnostic experience to address complex gastric mucosal backgrounds.

In this study, we also studied the relationship between the endoscopists’ qualifications and the detection rates of GPLs in different anatomical areas of the stomach. It was found that the detection rates of GPLs in the cardia, angularis, and gastric body, but not the antrum were positively correlated with the qualifications of endoscopists. Considering that the antrum is relatively easy to observe during gastroscopy and is one of the biopsy sites recommended by several consensus[5,47,50], the likelihood of detecting diseases in this area is relatively higher. Endoscopists with different levels of experience are able to identify suspicious lesions in the antrum and perform biopsies or routine screening biopsies. Compared to the antrum, the cardia is located at the junction of the esophagus and the stomach, a position that is relatively deep and has a complex angle. The upper part of the stomach, especially the anterior and posterior walls of the cardia, may be obscured by structures of the esophagus and other areas of the stomach, making it difficult to accurately locate and observe during endoscopic procedures[60]. In addition, observation of the cardia is also influenced by factors such as the filling status of the stomach, the physiological function of the cardia area, and patient discomfort[61,62]. Therefore, when performing biopsies for GPLs in these areas, there is a greater reliance on the endoscopist’s recognition of the lesions and their ability to accurately sample. These findings still require validation through more cases or multicenter data.

Studies have shown that there are no significant differences in clinical outcomes, examination costs, or complications between endoscopic assistants who have received systematic training and doctors performing endoscopy[63]. This finding highlights the importance of endoscopic assistants in gastroscopic examinations. As essential assistants in endoscopic examinations, the appropriate combination of endoscopic assistants with different qualifications and endoscopists may have a positive impact on the smooth progress of examinations and the detection rate of lesions. This study revealed that the ability of endoscopic assistants was positively correlated with the detection rate of GPLs by low- and mid-level endoscopists, indicating that the use of highly qualified endoscopic assistants can increase the detection rate of GPLs by endoscopists with low- and mid-level experience. This finding highlights the importance of the professional competence and experience of endoscopic assistants in improving the quality of diagnosis and treatment in medical teams. Highly qualified endoscopic assistants typically accumulate clinical knowledge and skills over time, enabling them to more effectively assist endoscopists with low- and mid-level experience when performing precise endoscopic operations to increase the detection rate of lesions.

This study revealed that the detection rate of EBR was negatively correlated with that of GC. The analysis revealed that advanced GC typically presents as obvious masses or ulcers, making it easier to identify and diagnose than GPLs and EGC. A higher EBR may indicate increased detection of NAG, GA or IM, which leads to a reduced relative detection rate of GC. With accumulated experience, endoscopists have gained more accurate lesion recognition skills, suggesting that endoscopists may reduce unnecessary biopsies by more accurately identifying easily recognizable lesions, leading to a decrease in the demand for EBR in these lesions while maintaining the detection rate. Therefore, as the identification of advanced GC is relatively intuitive, endoscopists may not need to rely on a higher EBR to improve the detection rate. A previous study revealed that after systematic training, endoscopists can significantly improve the detection rate of GC[64]. The endoscopists at this center had performed endoscopy for at least 1 year prior to the start of the study, with a median number of 7379 gastroscopies per endoscopist. These data indicate that the endoscopists included in this study received adequate training and may have sufficient recognition skills for advanced GC, resulting in no significant difference in the detection rates of GC among endoscopists with different qualifications.

In summary, gastric endoscopy, it is important to perform biopsies after improving the understanding of GPLs and GC rather than solely emphasizing EBR. Endoscopists with higher qualifications are more likely to identify GPLs; therefore, it is crucial to enhance the training and practical training of endoscopists in clinical practice. Furthermore, to improve the quality of gastroscopy examinations, medical institutions should also focus on the professional development and continuing education of endoscopic assistants to increase their professional skills and work efficiency. It is also important to consider the rational allocation of endoscopic assistants with different qualifications in endoscopy work to maximize the effectiveness of gastroscopy examinations and provide every patient with high-quality care and examination services.

This study has several limitations. First, as a single-center retrospective study, our findings may be limited by the specific medical environment and patient characteristics. Additionally, as our data do not include biopsy information from the esophagus or duodenum, our conclusions can only reflect the observations of patients who underwent gastric biopsies at our center during the study period and may not represent the actual situation of upper gastrointestinal endoscopy in the entire region. In terms of endoscopy usage, differences in the endoscopic techniques used by different endoscopists (such as narrow band imaging or magnifying endoscopy) may indirectly affect the endoscopists’ careful observation of the mucosa and decisions regarding biopsies. We found it difficult to fully consider the impact of this variable in our study. Helicobacter pylori (H. pylori) infection is closely associated with the occurrence of GC and can affect the metastasis of GC cells and the clinical prognosis of patients[65]. After H. pylori infection, the gastric mucosa typically exhibits characteristics such as thick mucus attachment, diffuse or punctate redness, mucosal swelling, and enlarged folds with a serpentine appearance[66]. After H. pylori eradication, the gastric mucosa may display a gastritis-like appearance[67]. These endoscopic features can influence the endoscopist’s ability to recognize and assess EGC. Therefore, H. pylori infection should also be considered a factor affecting endoscopic quality control and should be included in related studies. As this study is retrospective, the data on H. pylori infection were incomplete, and therefore this factor was not included in the data analysis. Future prospective studies could be designed for more in-depth research on this topic. In addition to the EBR and endoscopist qualifications, the duration of endoscopic procedures has also been identified as a crucial quality metric for esophagogastroduodenoscopy[68]. However, our research data lack records of examination time, which may reflect the thoroughness of endoscopists in mucosal examination, thus affecting the detection of lesions and the selection of biopsies. Therefore, we could not evaluate the potential impact of examination time on the biopsy rate or lesion detection rate. Our study population included patients with upper gastrointestinal symptoms and asymptomatic screening populations. This diversity makes it difficult for us to determine whether endoscopists are adopting targeted biopsies or multipoint random biopsies when performing biopsies, therefore, we could assess the impact of different biopsy strategies on outcomes. Finally, similar to studies on other endoscopic quality control indicators, endoscopists may manipulate the results by adjusting the number of biopsies or modifying gastric endoscopy reports, and the potential bias caused by this behavior[33,69] is beyond our control. Subsequent studies should focus on improving the quality of endoscopic workstation data and adopting prospective designs while recruiting researchers from multiple centers to increase the representativeness and generalizability of the study so that our findings can be extended to a wider range of regions and patient populations. Patient-derived xenograft (PDX) models have been widely used, and some studies have utilized PDX models to investigate the effects of different formulations of anesthetic drugs on breast cancer metastasis[70]. In future research, we can use PDX animal models to validate our findings. In addition, the number of patients with missed diagnoses of GC in our study cohort was small; therefore, we did not include relevant analyses in this study, though we will conduct further research in future studies. Additionally, due to the low number of patients with GC, we did not further investigate missed diagnoses. Furthermore, the proportion of other tumors was also very low; therefore, we did not conduct any related further analysis.

CONCLUSION

In summary, this study revealed that the detection rate of GPLs is positively correlated with endoscopist qualifications. The detection rate of GPLs by endoscopists with low to moderate qualifications is lower, but the cooperation of highly-qualified endoscopic assistants can improve the detection rate. Future research should focus on establishing and validating quality control standards for upper gastrointestinal endoscopy. In addition, exploring optimal EBR practices for different patient populations with varying GC risk factors to ensure the specificity and efficiency of biopsies and further increase the detection rates of GPLs and EGCs.

ACKNOWLEDGEMENTS

We would like to express our special thanks to Professor Xiao-Shan Li from the Affiliated Wuxi People’s Hospital of Nanjing Medical University for his guidance on statistical methodology.