Bibliometric analysis on the top one hundred cited studies on gastrointestinal endoscopy

Abstract

BACKGROUND

Gastrointestinal endoscopy has been widely used in the diagnosis and treatment of gastrointestinal diseases. A great many of studies on gastrointestinal endoscopy have been done.

AIM

To analyze the characteristics of top 100 cited articles on gastrointestinal endoscopy.

METHODS

A bibliometric analysis was conducted. The publications and their features were extracted from the Web of Science Core Collection, Science Citation Index-Expanded database. Excel, Web of Science database and SPSS software were used to perform the statistical description and analysis. VOSviewer software and MapChart were responsible for the visualizations.

RESULTS

The top 100 cited articles were published between 1976 and 2022. The guidelines (52%) and clinical trials (37%) are the main article types, and average publication year of the guidelines is much later than that of the clinical trials (2015 vs 1998). Among the clinical trials, diagnostic study (27.0%), cohort study (21.6%), case series (13.5%) and cross-sectional study (10.8%) account for a large proportion. Average citations of different study types and designs of the enrolled studies are of no significant differences. Most of the 100 articles were published by European authors and recorded by the endoscopic journals (65%). Top journals in medicine, such as the Lancet, New England Journal of Medicine and JAMA, also reported studies in this field. The hot spots of involved diseases include neoplasm or cancer-related diseases, inflammatory diseases, obstructive diseases, gastrointestinal hemorrhage and ulcer. Endoscopic surgery, endoscopic therapy and stent placement are frequently studied.

CONCLUSION

Our research contributes to delineating the field and identifying the characteristics of the most highly cited articles. It is noteworthy that there is a significantly smaller number of clinical trials included compared to guidelines, indicating potential areas for future high-quality clinical trials.

Key Words: Gastrointestinal endoscopy; Guideline; Clinical trial; Bibliometric analysis; Quality of study

Core Tip: The present study was conducted to analyze the characteristics of top 100 cited articles in this field. The primary findings include: (1) The guidelines (52%) and clinical trials (37%) are the main article types, and average publication year of the guidelines is much later than that of the clinical trials (2015 vs 1998); (2) Top journals in medicine, such as the Lancet, New England Journal of Medicine and JAMA, also reported studies in this field; and (3) The hot spots of involved diseases include neoplasm or cancer-related diseases, inflammatory diseases, obstructive diseases, gastrointestinal hemorrhage and ulcer. Researchers should attach importance to the unbalance of the highly cited guidelines and clinical trials on gastrointestinal endoscopy.

INTRODUCTION

According to the International Agency for Research on Cancer, the global number of cancer diagnoses reached nearly 20 million in 2022, and cancers derived from esophagus, stomach and colorectum account for 17.1%[1]. Moreover, the global population with benign gastrointestinal diseases, such as gastroesophageal reflux and peptic ulcer, has been increasing[2,3]. Gastrointestinal endoscopy is an effective tool for early diagnosis and treatment of gastrointestinal diseases[4,5]. In the past two centuries, digestive endoscopy has gone through the development process of early rigid endoscopy, semi-flexible endoscopy, fiber endoscopy, electronic endoscopy and capsule endoscopy. Recently, artificial intelligence has been added to gastrointestinal endoscopy to help the endoscopists make diagnostic decisions[6,7]. An inspiring study published in the Lancet Oncology demonstrated that the application of artificial intelligence to gastrointestinal endoscopy has higher accuracy than ordinary endoscopists in cancer detection[8]. Absolutely, remarkable achievements have been made in the field of gastrointestinal endoscopy and there are a large number of key studies that guide clinical practice. However, the characteristics of these studies have not been elucidated.

To be cited is the most common, direct and effective influence of a study, and it’s also the embodiment of recognition in the academic field. Therefore, it is necessary to analyze the highly cited articles. The typical features of these articles can be summarized in two aspects: The quality and visibility[9]. Although there are various motivations for citing particular articles, the primary one is still that the target papers are cognized by the authors and can be cited to support opinions. In this course, quality of the articles may be a priority in most circumstances. Another main point is the visibility of publications. Generally, studies published in top journals or written by well-known professors can gain more awareness and, thereafter, more citations[10,11]. Consequently, these articles may become more popular. In other words, popularity of the authors or journals would raise the evaluation of study quality. However, it is not always this way. In some academic areas, such as gastric diseases, articles published in journals with low impact factor (IF) can be highly cited[12].

Bibliometrics entails the quantitative analysis of scientific literature. Previous efforts have been made to compile and publish lists of the most frequently cited articles in various fields and journals. For instance, an early bibliometric analysis summarized the characteristics of highly cited clinical studies on digestive diseases[13]. In this paper, clinical trials account for 84% of the top 100 articles, whereas studies propose or refine relevant methodology make up 16%. However, our review of the literature indicates a lack of dedicated bibliometric analysis for the highly cited articles on gastrointestinal endoscopy at present. In our study, we intend to demonstrate the global trend and characteristics of the 100 most cited articles focused on gastrointestinal endoscopy.

MATERIALS AND METHODS

Eligibility of the publications

The target of this study is the studies that focus on gastrointestinal endoscopy. Thus, synonyms for gastrointestinal endoscopy should present in the title, be the main objects or tools. Besides, study types other than research paper and review paper (including systematic review and meta-analysis) were not taken into analysis, such as case report, letter, editorial, etc.

Literature acquisition

The publications were searched in the Web of Science Core Collection (WoSCC), Science Citation Index-Expanded (SCI-EXPANDED) database on April 11, 2024. Gastrointestinal endoscopy, intestinal endoscopy, gastric endoscopy, digestive system endoscopy and gastroenterological endoscopy were combined by a Boolean operator “OR”. The document types were restricted to “article” and “review article”. Then, the records were sorted by citations in a descending order. The investigators reviewed these records and determined the eligibility of the studies. Finally, the top 100 articles were filtered out.

Feature extraction

The 100 records were exported to a plain text file. Basic bibliometric data were automatically extracted from WoSCC. The data were saved as an Excel file. Afterwards, two investigators manually obtained some other parameters, including IF, category rank and quartile of the journals, study type and study design.

Feature analysis

Basic bibliometric analyses, such as the publication number and citations of the authors, institutions and countries/regions, and different ages, were performed using Excel and VOSviewer. The keywords were classified and counted by VOSviewer. We selected the keywords related to diseases and operations respectively. Hence, two sets of keywords were constructed and they were arranged by the publication year. Visualization of keywords was conducted using VOSviewer, while visualization of global distribution of the articles was based on MapChart, an online tool. Manually extracted data, including IF, category rank and quartile of the journals, study type and study design, were analyzed with descriptive statistics. Average citations of different study type and study design were compared using one-way analysis of variance test or independent t-test, and the pairwise comparisons were performed using LSD test via SPSS 26.0. Correlation analysis was applied to determine the relationship between IF and average citations. The significance level was set as 0.05.

RESULTS

Publication trend at different ages

In the WoSCC SCI-EXPADED database, the top 100 cited articles were published between 1976 and 2022. The citations ranged from 134 to 755, with a mean of 268, and a median (inter-quartile) of 219 (167, 317). There are 7 publication years in which the publication number was 5 or over. The most productive publication year was 2017 (n = 8), followed by 2012 (n = 7), 2020 (n = 7), 2019 (n = 6), 2010 (n = 5), 2014 (n = 5), and 2016 (n = 5). Generally, the publications increased at different ages (time interval: 5 years), and over a half of the highly cited articles were published between 2010 and 2020 (Table 1).

Table 1 The publication trend of the top 100 cited articles at different ages.

Publication year	Publication number	Total citations	Average citations
1976-1980	2	737	369
1981-1985	3	549	183
1986-1990	6	1029	172
1991-1995	7	1855	265
1996-2000	4	741	185
2001-2005	12	4860	405
2006-2010	11	2166	197
2011-2015	23	6532	284
2016-2020	29	7746	267
2021-2022	3	599	200

Study type analysis

The study types of the top 100 cited articles included clinical trial (37.0%), review (6.0%), meta-analysis (3.0%), guideline (52.0%) and database analysis (2.0%) (Table 2). Interestingly, 90.4% of the guidelines were issued by the European Society of Gastrointestinal Endoscopy, and another 2 organizations were the American Society for Gastrointestinal Endoscopy and the Japan Gastroenterological Endoscopy Society. Average publication year of the clinical trials was 1998, while it was 2015 for the guidelines (Table 2). Average citations of the meta-analysis studies were the highest (n = 287), followed by the guidelines (n = 276), clinical trials (n = 265), reviews (n = 246) and database studies (n = 151) (Table 2). However, the differences in average citations of different study types were not significant (P = 0.785).

Table 2 Publication trend of different study types.

Study type	Publication number	Total citations	Average citations	Average publication year
Clinical trial	37	9818	265	1998
Review	6	1474	246	2007
Guideline	52	14359	276	2015
Database study	2	302	151	2009
Meta-analysis	3	861	287	2007

Study design analysis

According to the JAMA classifications for clinical trials[14], the 37 clinical trials were divided into 9 categories, including survey study (n = 2, 5.4%), randomized controlled trial (n = 2, 5.4%), quality improvement study (n = 2, 5.4%), parallel-design trial (n = 3, 8.1%), diagnostic study (n = 10, 27.0%), cross-sectional study (n = 4, 10.8%), comparative effectiveness research (n = 1, 2.7%), cohort study (n = 8, 21.6%), and case series (n = 5, 13.5%) (Table 3). Average citations of the cross-sectional study studies were the highest (n = 327), followed by the diagnostic studies (n = 288) and case series (n = 276) (Table 3). However, the differences of average citations of different study designs were not significant (P = 0.967). We further divided the studies into single-center (n = 25, 67.6%) and multi-center (n = 12, 32.4%). Average citations of the single-center studies were comparable with those of the multicenter-studies (252 vs 293, P = 0.479).

Table 3 Publication trend of different study designs.

Study design	Publication number	Total citations	Average citations	Average publication year
Survey study	2	514	257	1993
Randomized controlled trial	2	405	203	1994
Quality improvement study	2	301	151	2000
Parallel-design trial	3	809	270	2006
Diagnostic study	10	2883	288	2003
Cross-sectional study	4	1308	327	1984
Comparative effectiveness research	1	171	171	1984
Cohort study	8	2060	258	1998
Case series	5	1382	276	1998

Analyses of the authors, institutions and countries/regions

The 100 articles were signed by 874 authors of 493 affiliations from 43 countries or regions. Thirty-one authors published 5 papers or over. Table 4 shows the top 10 authors in terms of publication number. Hassan Cesare was the most productive author (n = 27), followed by Dumonceau Jean-Marc (n = 23) (Table 4). The citations of Hassan Cesare, Dumonceau Jean-Marc and Dinis-Ribeiro Mario were above 5000 (n = 8102, 6650, and 5085, respectively). Ninety percent of the top 10 authors had a European nationality. Fifty-nine affiliations created at least 5 articles. Table 5 displays the top 10 prolific institutions. University of Amsterdam (n = 28) and IRRCCS Policlinico Gemelli (n = 27) took the leading position. Ten (90.9%) of them were European countries. Twenty-three countries or regions produced 5 publications or over. Most of the countries or regions belonged to Europe (Figure 1). Moreover, almost all prolific countries or regions, except the United States and Argentina, were concentrated in Europe (Figure 1).

Figure 1  Global distribution of the 100 articles.

Table 4 Top 10 authors in terms of publication number.

Author	Publication number	Total citations	Institution	Country
Hassan Cesare	27	8102	Nuovo Regina Margherita Hosp	Italy
Dumonceau Jean-Marc	23	6650	Gedyt Endoscopy Ctr	Argentina
Van Hooft Jeanin E.	14	3621	Leiden Univ	Netherlands
Dinis-Ribeiro Mario	14	5085	Portuguese Institute of Oncology	Portugal
Dumonceau Jean-Marc	12	3591	Charleroi Univ Hosp	Belgium
Vanbiervliet Geoffroy	9	2119	CHU Nice	France
Bisschops Raf	8	2249	Univ Hosp Leuven	Belgium
Tringali Andrea	8	2225	Fdn Policlin Univ A Gemelli IRCCS	Italy
Spada Cristiano	8	2398	Fdn Poliambulanza	Italy
Ponchon Thierry	8	3125	CHU Lyon	France

Table 5 Top 10 institutions in terms of publication number.

Institution	Publication number	Total citations	Country
University of Amsterdam	28	7717	Netherlands
IRRCCS Policlinico Gemelli	27	6995	Italy
Gedyt Endoscopy CTR	20	6322	Argentina
Academic Medical Center Amsterdam	19	5114	Netherlands
Poliambulatorio Nuovo Regina Margherita	17	5637	Italy
Univ Libre Bruxelles	15	4217	Belgium
Aix Marseille University	14	3943	France
University of Oslo	14	3865	Norway
Centre of Postgraduate Medical Education	13	3883	Poland
Maria Sklodowska-Curie National Research Institute of Oncology	12	3619	Poland
University of Geneva	12	3215	Switzerland

Journal analysis

In total, 24 journals documented the 100 papers, 16 (66.7%) of them had only 1 publication (Table 6). The Endoscopy (50 publications) was the leading journal in this field, followed by Gastrointestinal Endoscopy (11 publications) (Table 6). It must be noted that 80.4% of the publications in Endoscopy are guidelines and clinical trials account for 12%. Most of the journals are top journals, the Lancet, New England Journal of Medicine and JAMA included (Table 6). The average IF (2022) is 30.2, and the median is 9.5 (Table 6). These journals had stable performance of IF in the latest 5 editions. The average 5-year IF is 24.6, with a median of 14.3 (Table 6). Eighteen (75.0%) journals were located in the first quartile (Q1) of the categories, 5 (20.8%) in the second (Q2) and 1 (4.2%) in the fourth (Q4) (Table 6). Correlation analysis indicated that the average citations was not correlated to the IF (Pearson correlation = 0.146, P = 0.496).

Table 6 Characteristics of journals recording the 100 articles.

Source title	IF, 2022	IF, 5-year	Category rank	Quartile	Publications	Citations	Average citations
Endoscopy	9.3	9.6	6	Q1	50	13907	278
Gastrointestinal Endoscopy	7.7	7.9	16	Q1	11	3389	308
Gut	24.5	25.2	5	Q1	7	1627	232
American Journal of Gastroenterology	10.2	11.5	10	Q1	6	1592	265
New England Journal of Medicine	158.5	115.7	2	Q1	4	706	177
Annals of Internal Medicine	39.2	35.3	6	Q1	2	528	264
Clinical Gastroenterology and Hepatology	12.6	11.4	8	Q1	2	708	354
Digestive Endoscopy	5.5	5.3	16	Q1	2	490	245
Annual Review of Biomedical Engineering	9.7	18.9	9	Q1	1	140	140
Clinical Microbiology Reviews	36.8	37.4	3	Q1	1	305	305
European Radiology	5.9	5.5	16	Q1	1	144	144
Gastroenterology	29.4	25.8	3	Q1	1	699	699
Jama-Journal of The American Medical Association	120.7	81.4	3	Q1	1	599	599
Journal of Clinical Pathology	3.4	3.1	28	Q2	1	136	136
Journal of Gastroenterology and Hepatology	4.1	4	38	Q2	1	187	187
Journal of Pediatric Gastroenterology and Nutrition	2.9	3.4	42	Q2	1	140	140
Lancet	168.9	118.1	1	Q1	1	138	138
Lancet Oncology	51.1	41.6	4	Q1	1	209	209
Photochemistry and Photobiology	3.3	3.2	28	Q2	1	281	281
Scandinavian Journal of Gastroenterology	1.8	2.2	86	Q4	1	283	283
Scientific Data	9.8	10.8	11	Q1	1	158	158
Surgical Endoscopy and Other Interventional Techniques	3.1	3.5	50	Q1	1	135	135
Surgical Endoscopy-Ultrasound and Interventional Techniques	3.1	3.5	50	Q1	1	149	149
World Journal of Gastroenterology	4.3	5.3	33	Q2	1	164	164

IF: Impact factor.

Keywords analysis

We analyzed 2 sets of keywords. The one is related to the diseases. As shown in Figure 2, gastrointestinal endoscopy was most frequently conducted to detect neoplasms or cancer-related diseases in the top 100 articles. Inflammatory diseases (pancreatitis, esophagitis, inflammatory bowel disease and gastritis), obstructive diseases (strictures, stones and obstructions), gastrointestinal hemorrhage and ulcer were also highly concerned topics (Figure 2). Other diseases, such as coronary artery disease, anemia and pneumonia, were mentioned in the top 100 studies (Figure 2). The other set of keywords is associated with the application of gastrointestinal endoscope. Routine gastrointestinal endoscopy with various endoscopes had the most occurrence, followed by endoscopic surgery, endoscopic therapy, stent placement and sedation (Figure 3). Other high-frequency operations included endoscopic retrograde cholangiopancreatography, bowel preparation and drainage (Figure 3).

Figure 2  Disease profile involved in the 100 articles.

Figure 3  Operations and procedures involved in the 100 articles.

DISCUSSION

As far as we know, this is the first bibliometric analysis to elucidate the characteristics of the top 100 cited publications focused on gastrointestinal endoscopy. Our results show a typical trend in the document types of the 100 publications in this field. Generally, the guidelines and clinical trials are the main article types. However, in recent years, the guidelines with high citations were intensively published, while very few clinical trials rank top 100 in terms of citations. The first guideline among the top 100 articles on gastrointestinal endoscopy provides recommendations on conscious sedation and monitoring[15]. This guideline was published in 2003 by the American Society for Gastrointestinal Endoscopy. The first clinical trial on sedation and gastrointestinal endoscopy in the WoS database was published in 1976 in the British Medical Journal[16]. In this study, diazepam was the sedative. Midazolam was proved to be preferred over diazepam by patients in 1983[17]. Then, propofol was used in gastrointestinal endoscopy and had advantages in rapid recovery over midazolam, but reduced patient satisfaction[18]. However, with the increasing number of endoscopic cases, it is vital to provide rapid-onset and rapid-recovery sedation. Therefore, propofol became the primary sedative for gastrointestinal endoscopy. Recently, remimazolam has been recommended for gastrointestinal endoscopy[19,20]. The efficiency and safety of ciprofol have also been demonstrated[21,22]. Anesthesiologists now have many choices to induce sufficient and safe sedation for gastrointestinal endoscopy. The latest highly cited guideline gives detailed strategies of endoscopic submucosal dissection for superficial tumors of the digestive tract[23]. Among the 52 highly cited guidelines in this field, this is the third update of edition 2015 and edition 2019[24,25], which indicates the early diagnosis and treatment of gastrointestinal neoplasms.

The first clinical trial among the top 100 articles on gastrointestinal endoscopy, published in 1976, evaluated the value of emergency upper gastrointestinal fiber-endoscopy and modified Sengstaken tube for patients with portal hypertension and varicosity[26]. According to the keywords analysis of this study, gastrointestinal hemorrhage is one of the hot spots in this field. There are 16 studies focused on gastrointestinal hemorrhage in the 100 articles. The most popular one, indicating the value of capsule endoscopy for obscure gastrointestinal hemorrhage, was cited 699 times[27]. The most cited clinical trial on gastrointestinal endoscopy, however, is about the influence of bowel preparation on colonoscopy[28]. This study shows that cleansing is highly correlated to quality, difficulty and speed of colonoscopy and that hospitalized patients and patients with higher levels of comorbidities had lower cleanliness quality. Thus, a guideline for bowel preparation was published the next year (2006) and became one of the top 100 cited articles[29]. Another two updates (2013 and 2019) of guidelines for bowel preparation also raised many citations[30,31].

Since clinical trials are the basis of meta-analyses and guidelines, the quality of clinical trials determines the reliability of meta-analyses and guidelines. Mismatch of the number of guidelines and clinical trials in the top 100 must be paid attention. Many recommendations in the guidelines have low-level evidence, even in the most cited guideline[24]. For instance, “If the horizontal margin is positive and no other high risk criteria are met, endoscopic surveillance/re-treatment is an option” and “European Society of Gastrointestinal Endoscopy recommends regular endoscopic follow-up after excision/ablation of Barrett esophagus-associated high-grade dysplasia or mucosal cancer, but more research is needed to determine the appropriate short and long intervals” are strongly recommended, but the evidence level is low[24]. Thus, further clinical trials should be conducted to improve the weaknesses.

There may be some reasons for the different citations between guidelines and clinical trials. Constructing a guideline usually includes raising problem, reviewing literature, evaluating research evidence, expert consensus, drafting, external review and revisions. Therefore, guidelines are considered as the golden standards in disciplines and naturally are more likely to be cited. Researchers in this field should pay more attention to the quality rather than the number of studies. Despite the various factors influencing citation, basic features of high-quality trials, such as focusing on the clinical problems, rational study design, standardized study procedure and scientific result interpretation, should be more considered.

There are some limitations in this bibliometric analysis. In order to retrieve well-recognized articles, we did the search in WoSCC SCI-EXPANDED database. This may omit some publications and impact cited times and ranks of the studies. However, the SCI-EXPANDED is a well-known database that records high-quality English articles, which is in accordance with our research target. The choice of time interval (5 years) to count the number of publications and citations has no scientific basis. We referred to the WoS system where a 5-year IF is used to evaluate journals. Besides, we did not merge synonyms. So, there are some repeated keywords in Figure 2 and Figure 3. Nevertheless, it does not impact the results.

CONCLUSION

In the present study, we analyzed the characteristics of the 100 most frequently cited articles on gastrointestinal endoscopy. Authors from Europe made significant contributions to the 100 hot papers. Journals that recorded the articles have high IF. The hotspots are endoscopic diagnosis and treatment of tumor-related diseases. The guidelines are the most popular and recognized articles in this field, followed by clinical trials. The differences of citations among study types and designs are not significant for the top 100 articles. However, highly cited clinical trials are rare for more than a decade. High-quality clinical trials are urgently needed to support the guidelines, and further, the clinical practice.