You are here

  • Home
  • Archive
  • Volume 11, Issue 1
  • Learning curve of achieving competency in emergency endoscopy in upper gastrointestinal bleeding: how much experience is necessary?

Article Text

Article menu
Gastrointestinal bleeding
Learning curve of achieving competency in emergency endoscopy in upper gastrointestinal bleeding: how much experience is necessary?
  1. Gabriel Allo1,
  2. Sonja Lang1,
  3. Anna Martin1,
  4. Martin Bürger1,
  5. Xinlian Zhang2,
  6. Seung-Hun Chon3,
  7. Dirk Nierhoff1,
  8. Ulrich Töx1,
  9. Tobias Goeser1,
  10. Philipp Kasper1
  1. 1Department of Gastroenterology and Hepatology, Faculty of Medicine and University Hospital Cologne, University of Cologne, Cologne, Germany
  2. 2Division of Biostatistics and Bioinformatics, Herbert Wertheim School of Public Health and Human Longevity Science, University of California San Diego, La Jolla, California, USA
  3. 3Department of General, Visceral and Cancer and Transplant Surgery, University of Cologne, Cologne, Germany
  1. Correspondence to Dr Gabriel Allo; gabriel.allo{at}uk-koeln.de

Abstract

Objectives The management of upper gastrointestinal bleeding (UGIB) has seen rapid advancements with revolutionising innovations. However, insufficient data exist on the necessary number of emergency endoscopies needed to achieve competency in haemostatic interventions.

Design We retrospectively analysed all oesophagogastroduodenoscopies with signs of recent haemorrhage performed between 2015 and 2022 at our university hospital. A learning curve was created by plotting the number of previously performed oesophagogastroduodenoscopies with signs of recent haemorrhage against the treatment failure rate, defined as failed haemostasis, rebleeding and necessary surgical or radiological intervention.

Results The study population included 787 cases with a median age of 66 years. Active bleeding was detected in 576 cases (73.2%). Treatment failure occurred in 225 (28.6%) cases. The learning curve showed a marked decline in treatment failure rates after nine oesophagogastroduodenoscopies had been performed by the respective endoscopists followed by a first plateau between 20 and 50 procedures. A second decline was observed after 51 emergency procedures followed by a second plateau. Endoscopists with experience of <10 emergency procedures had higher treatment failure rates compared with endoscopists with >51 emergency oesophagogastroduodenoscopies performed (p=0.039) or consultants (p=0.041).

Conclusions Our data suggest that a minimum number of 20 oesophagogastroduodenoscopies with signs of recent haemorrhage is necessary before endoscopists should be considered proficient to perform emergency procedures independently. Endoscopists might be considered as advanced-qualified experts in managing UGIB after a minimum of 50 haemostatic procedure performed. Implementing recommendations on minimum numbers of emergency endoscopies in education programmes of endoscopy trainees could improve their confidence and competency in managing acute UGIB.

  • endoscopy
  • gastrointestinal bleeding
  • surgical training

Data availability statement

Data are available on reasonable request.

https://creativecommons.org/licenses/by/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

https://doi.org/10.1136/bmjgast-2023-001281

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Limited data suggest that patients with acute upper gastrointestinal bleeding (UGIB) have improved outcomes, if they receive endoscopy by more experienced endoscopists; however, it is unknown how many emergency procedures are necessary for an endoscopist to achieve competence in managing acute UGIB.

    WHAT THIS STUDY ADDS

    • Here for the first time, we describe a clear learning curve for endoscopists treating acute UGIB.

    • Endoscopists with <10 oesophagogastroduodenoscopies with signs of haemorrhage had significantly higher treatment failure rates than experienced endoscopists or consultants.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Implementing recommendations on minimum numbers of emergency endoscopies in education programmes of endoscopy trainees could improve their confidence and competency in managing UGIBs.

    Introduction

    Acute upper gastrointestinal bleeding (UGIB) is a life-threatening event and remains a leading indication for emergency endoscopy. In recent years, the management of UGIB has seen rapid advancements with substantial innovations in triage, timing of endoscopy, medication, transfusion management and haemostatic procedures.1–8 Although research efforts to optimise patient care are advancing on all fronts, one crucial aspect of treatment has been almost completely neglected: the endoscopist.

    While the benefits of the aforementioned innovations are evident, mastering the necessary skills to confidently and successfully apply these techniques requires intensive training and education to achieve profound understanding of bleeding pathophysiology, gastrointestinal anatomy and the intricacies of haemostatic devices. Besides accurate hands-on skills, endoscopists are constantly faced with challenges of accurate patient triage, initiation of medical and haemostatic procedures as well as real-time decision-making during life-threatening emergencies.

    Considering these challenges, a profound education of gastroenterologists in the field of UGIB appears indispensable. However, recent analysis showed that only 60% of final year gastroenterology trainees from the UK expressed confidence in managing UGIB, with 82% desiring further training.9 The trainees reported a serious lack of exposure to haemostatic procedures and alarmingly, 18.6% of certified gastroenterologists had no exposure to any type of endotherapy.10 11 This alarming lack of trainee exposure to acute UGIB warrants a fundamental reassessment of current training programmes.12

    It is well known from colonoscopy studies that the experience of the endoscopist and the number of successfully performed procedures improve the examination results.13 Similarly, establishing a minimum number of haemostatic procedures required before certification may significantly enhance trainees’ confidence and skill set. However, current guidelines lack specific recommendation in the context of UGIB.14–16

    Therefore, this study aims to investigate how the endoscopists’ experience can impact outcomes in UGIB and to identify a minimum number of emergency procedures required to achieve competence in managing UGIB.

    Methods

    We reviewed all oesophagogastroduodenoscopies (EGDs) which were performed at the University Hospital Cologne, between 1 January 2015 and 31 December 2022 because of suspected UGIB from our endoscopy documentation system. From this dataset, we identified cases with stigmata of recent haemorrhage (SRH), defined as spurting or oozing lesions, fresh blood in the upper gastrointestinal tract with or without detectable lesions, adherent clot as well as visible vessels as previously described.17 In case of variceal haemorrhage, active bleeding from varices, red whale or white nipple signs, a clot overlying the varices or oesophageal varices without any other obvious source of bleeding upper gastrointestinal tract were also defined as SRH. We included EGDs with adherent clots detected since the underlying bleeding sources demonstrated high rebleeding rates and guidelines recommend considering clot removal followed by endoscopic treatment of underlying high-risk stigmata.14 We hypothesised that increasing experience translates into higher rates of clot removal and endoscopic treatment resulting in lower rebleeding rates.

    Cases were then included in our study analysis, if EGDs with SRH were performed by a member of our 24-hour on-call endoscopy team. Exclusion criteria were age <18 years, no follow-up 7 days after index EGD or initial presentation at another hospital.

    During the study period, 18 physicians participated in emergency on-call duties. Seven were highly experienced consultants with >100 emergency procedures performed. Three of them had >5 years of experience in emergency endoscopies and four of them had >15 years of experience partly gained at prior employments at other hospitals. All analysed trainees started training at our tertiary centre. There were three trainees, who started endoscopy training within 3 years before the study period and eight trainees started endoscopy training during the study period. In our department, trainees start on-call duties after completing 4 years of general internal medicine and gastroenterology training and at least 6 months of supervised full-time endoscopy training. They are considered proficient for on-call duties by the discretion of the chief of our endoscopy unit when a proven independent ability to endoscopically treat bleeding in the context of EGD and colonoscopies is achieved. A virtual reality simulator or ex vivo animal models for training of both basic and advanced endoscopic techniques are not available at our centre.

    Our on-call emergency endoscopy service is operating on a 365 days/24 hours basis from 17:00 hours to 08:00 hours on weekdays and 24 hours on weekends and holidays. On-call duties are performed by all endoscopists after they are considered sufficiently trained as described above. In addition, there is a backup team of three highly experienced consultants who can be called for on-demand supervision in case of unmanageable endoscopic emergencies during after-hours, weekends and holidays. In case of emergency bleedings and unstable patients, endoscopies are usually performed at the intensive care unit or emergency department with anaesthetic support or an experienced physician responsible for sedation.

    To evaluate the endoscopic performance of each endoscopist, we chose the composite of treatment failure defined as unsuccessful primary endoscopic haemostasis (and necessary surgical or radiological intervention to achieve haemostasis) and/or rebleeding within 7 days after index EGD as the primary end point. Rebleeding was defined as new haematemesis or bloody nasogastric aspirate after initial successful endoscopic intervention, new melena or haematochezia after normalisation of stool colour or a drop of haemoglobin ≥20 g/L in the absence of an alternative explanation and SRH on repeated endoscopy. We did not include mortality in the composite end point since the majority of patients with UGIB die of non-bleeding-related causes, such as multiorgan failure, pulmonary conditions and terminal malignancies, which are not influenceable by the endoscopists’ procedural skills.18

    For each EGD, we determined the number of EGDs with SRH ever performed by the respective endoscopist prior to the new-onset bleeding event by reviewing our endoscopy documentation system and defined this number as a surrogate marker for experience in emergency endoscopy. Supervision was defined as attendance of a more experienced endoscopist during endoscopy.

    Descriptive analysis was conducted using Statistical Package for the Social Sciences, V.29 (IBM, Armonk, New York, USA) and R statistical software, V.3.5.1 (2018 the R Foundation for Statistical Computing). Categorical variables were presented as absolute and relative frequencies. Continuous variables were expressed as median and IQR. Univariate and multivariate analyses using the logistic regression model were performed to identify factors associated with treatment failure. Covariates were included in the multivariate analysis if their p value was <0.05 in the univariate analysis.

    A learning curve was produced by plotting the number of prior EGDs with SRH performed by the respective endoscopist against the cumulative treatment failure rate until this level of experience. This approach gives a visual representation of the learning phase of endoscopists. We identified the coordinates with the most significant drop in treatment failure and thereby divided the study population into four groups. These groups were compared using one-way analysis of variance for parametric variables or Kruskall-Wallis test for non-parametric data, respectively. Additionally, we performed a moving average analysis by calculating the mean treatment failure rate for each EGD with SRH performed by the respective endoscopist over blocks of five endoscopies, mirroring the methods of a previously published study.13 A graph was created by plotting the mean treatment failure rate for all endoscopists against the number of EGDs with SRH performed. For categorical variables, χ2 test was used. Post hoc p values were obtained from Fisher’s exact test with adjustment for multiple comparisons using the false discovery rate. P values <0.05 were regarded as statistically significant.

    Results

    Flow chart of patient selection is shown in figure 1. Patients’ baseline characteristics are described in table 1. 787 cases were included in the final analysis of which 249 (31.6%) were women and the median age was 66 years. Outwith regular hours, 334 (42.4%) EGDs were performed and 108 (13.7) procedures were supervised. Duodenal as well as gastric ulcers were the most frequent bleeding source with 246 (31.3%) and 132 (16.8%) cases, respectively. Spurting and oozing bleeding was detected in 47 (6.0%) and 437 (55.5%) cases, respectively. Gastric and oesophageal varices were the source of bleeding in 116 (14.8%) cases and active bleeding from varices was detected in 84 EGDs (72.4% of gastric and oesophageal varices).

    Figure 1
    Figure 1

    Study flow. EGD, oesophagogastroduodenoscopy; SRH, stigmata of recent haemorrhage; UGIB, upper gastrointestinal bleeding.

    View this table:
    Table 1

    Patient demographics, endoscopic findings and outcomes

    Endoscopic haemostatic interventions were performed in 643 EGDs (81.7%) and the most frequently applied procedures were haemoclips (297; 37.7%). Ongoing bleeding at the end of endoscopy was documented in 75 procedures (9.5%). Surgical or radiological interventions were necessary to achieve haemostasis in 14 (1.8%) and 48 (6.1%) of the patients, respectively. Treatment failure occurred in 225 (28.6%) cases while 203 (25.8) patients died during their hospital stay.

    Factors associated with treatment failure

    In the underlying study, age over 65 years, a history of prior gastrointestinal bleeding, the need for more than two packed red blood cells, EGD during on-call duty, active bleeding detected during endoscopy, gastric ulcers, oesophageal varices, gastric varices as well as the experience level were significantly associated with treatment failure on univariate analysis. Multivariate analysis identified age over 65 years, the need for more than two packed red blood cells, EGD during on-call duty, active bleeding, gastric ulcers, oesophageal varices and the experience level as factors that are significantly associated with treatment failure (table 2).

    View this table:
    Table 2

    Factors associated with treatment failure

    Learning curve

    Figure 2 displays the change of the cumulative treatment failure rate depending on the experience of the endoscopist. The curve shows a clear learning curve with a marked decline in endoscopic treatment failure rates after at least nine EGDs with SRH have been performed (39.1%). Furthermore, after reaching a first plateau phase between 20 and 50 EGDs performed, the learning curve demonstrates another decline after 51 EGDs (33.1%) followed by a second plateau phase up to 100 procedures. The moving average analysis also showed a clear decline in treatment failure rates with growing experience (online supplemental figure 1).

    Figure 2
    Figure 2

    Endoscopists’ learning curve. Experience level=cumulative number of previous oesophagogastroduodenoscopies with stigmata of recent haemorrhage performed by endoscopists. The red line marks the point with the highest treatment failure rate followed by a steep drop. The green line marks the last point of the first plateau followed by a marked drop in treatment failure rates. The straight line displays the curve of best fit for the plot. Goodness of fit: R2=0.46, F-statistics=85.18, p<0.001.

    Comparison of experience groups

    After demonstrating the association between the experience level and the treatment failure rate, we built experience level groups based on clear cuts on the learning curve. A total number of 64 (8.1%) emergency EGDs have been performed by endoscopists with an experience of <10 EGDs with SRH (low experience group) in the present study. 219 EGDs (27.8%) have been performed by endoscopist with an experience of 10–50 EGDs (intermediate experience group) and 152 (19.3%) by endoscopist with an experience of 51–100 EGDs (high experience group). 352 (44.7%) EGDs have been performed by experienced consultants with >100 EGDs of experience (consultants). After starting their endoscopy training, 5 of 11 trainees performed their first 50 EGDs with SRH within the study period and the median duration to reach this cut-off was 2.5 years (range 1.68–2.94 years). The treatment failure rates of the four groups were 39.1%, 31.5%, 25.0% and 26.4%, respectively (figure 3 and table 2). The group of endoscopists with low experience had a significantly higher treatment failure rate than the highly experienced group (p=0.039) and consultants (p=0.041), while the treatment failure rate did not differ significantly compared with the intermediate experienced group. Results were consistent in the multivariate analysis.

    Figure 3
    Figure 3

    Treatment failure rates of experience groups. Experience=cumulative number of previous oesophagogastroduodenoscopies with stigmata of recent haemorrhage performed by endoscopists. Superscripted symbols indicate significant differences (*p=0.039; #p=0.041). Treatment failure rate did not differ significantly between group <10 and 10–50 (p=0.26).

    The high experience group performed emergency EGDs more frequently during on-call duty compared with the low experience group (p=0.002), the intermediate experience group (p=0.012) and the consultants (p=0.01). The low experience group performed their emergency EGDs more frequently under supervision compared with the intermediate experience group (p=0.003), the high experience group (p<0.001) and the consultants (p<0.001) (online supplemental table 1). Interestingly, there was no significant difference in the choice of endoscopic treatment methods. Although innovative haemostatic devices such as haemostatic spray and over-the-scope clips (OTSCs) were used more frequently by more experienced endoscopists compared with the low experience group without reaching statistical significance (figure 4 and online supplemental table 1). Furthermore, in a subgroup analysis of all ulcers with adherent clots, the low experience group received endoscopic haemostatic therapy in only 9.1%, differing significantly to 41.9%, 58.3% and 37.3% in the intermediate experience group, the high experience group and the consultants (p=0.049), respectively (data not shown).

    Figure 4
    Figure 4

    Endoscopic haemostatic interventions according to experience groups. Experience=cumulative number of previous oesophagogastroduodenoscopies with stigmata of recent haemorrhage performed by endoscopists. OTS, over-the-scope.

    Discussion

    In this retrospective study, we found that the experience of endoscopists was a significant factor for a successful treatment in acute UGIB. The treatment failure rate declined with increasing experience and endoscopists with experience of <10 EGDs with SRH performed had a higher treatment failure rate compared with those with >51 procedures.

    The importance of the endoscopists’ experience in managing UGIB has not been extensively studied, and due to missing data, most guidelines on the management of UGIB do not recommend a specific number of emergency EGDs before considering an endoscopist as competent to perform emergency procedures independently and without supervision.14 15 To date, there have been only few studies examining the role of the endoscopists’ experience on the success of the haemostatic emergency procedures.

    An Italian study from 2005 compared the performance of ‘experienced’ and ‘less experienced’ endoscopists in managing non-variceal UGIB. Seven endoscopists were divided into two groups based on the number of haemostatic endoscopic procedures performed before the study. Patients with EGDs performed by endoscopists with >100 emergency procedures had lower rebleeding rates and transfusion requirements than patients treated by endoscopists with 40–70 emergency procedures.19 The authors suggest that the lower appropriateness of applied endoscopic treatment by less experienced endoscopists might have led to higher rebleeding rates, particularly in cases of Forrest IIb ulcers, where the less experienced endoscopists were often not following guidelines, fearing uncontrollable bleeding after recommended clot removal.

    Our results align with the above-mentioned study, as we also found an association of lower experience and higher treatment failure rates. However, our study findings extend beyond the simple comparison of young and highly experienced endoscopists by revealing a clear learning curve. The given curve indicates an initial learning phase of about 9 EGDs with SRH, followed by a phase of proficiency until 51 procedures and thereafter, a third phase of competence. The risk factors associated with treatment failure identified in our study and patient characteristics did not differ between experience groups except that group 51–100 performed EGDs more frequently during on-call service. Interestingly, there was a trend towards more experienced endoscopists using innovative haemostatic procedures such as OTSCs or haemostatic spray more frequently. This observation of choosing more specific haemostatic devices might have impacted outcomes, as recent studies showed a significant benefit of OTSCs in high-risk lesions and in case of rebleeding compared with standard endoscopic therapy.4 20 21 Additionally, the results of recent studies indicate that the use of haemostatic spray might result in more favourable outcomes in gastrointestinal tumour bleeding, which is associated with striking rebleeding and mortality rates.22 23 However, it is likely that appropriate training would increase confidence in applying these modern techniques and improve a patient’s outcome.

    Similar to the aforementioned Italian study,19 endoscopists with lower experience also did not remove adherent clots in the vast majority of cases in the present study, likely due to fear of subsequent uncontrollable bleeding after clot removal. This underlines the importance of sufficient training to establish proficiency and confidence in all available endoscopic haemostatic techniques as well as in complication management.

    Recent surveys from the UK displayed a lack of confidence among gastroenterology trainees in managing UGIB. Here, only about 60% of gastroenterology trainees expressed confidence in managing UGIB in their final year and 82% desired further training in haemostatic procedures.9 When analysing the portfolios of certified gastroenterologist in the UK, 18.6% had no exposure to any type of endotherapy, while only 37.1% performed band ligation and 50.7% placed a metal clip.10 Furthermore, a significant decline of trainee exposure to acute UGIB from 76% to 15% was observed in the UK, with about one-quarter of trainees doubting to become confident until the end of specialty training.12

    These alarming results further highlight the need for comprehensive training programmes, where trainees are able to perform challenging UGIB cases under supervision. Several training initiatives have been postulated in recent years, such as individual endoscopy lists and tailored training programmes, which aim to increase trainees’ exposure to UGIB without risking patient safety.24 However, although recommendations exist, these are not sufficiently implemented and applied in everyday clinical practice.

    Our study results support the idea of a requirements catalogue for endoscopic examinations during the endoscopy training programme, including minimum numbers of endoscopic emergency procedures (eg, haemostatic interventions) required before training is regarded sufficient, since our learning curve revealed a clear association of experience in haemostatic procedures and patient outcome. Training methods to improve trainees’ exposure to acute UGIB in real-life setting might be the implementation of protected and supervised morning slots for stable overnight patients with UGIB. Besides collecting experience during regular practice, implementing gastrointestinal haemostatic training courses in the regular education programme can significantly improve endoscopists’ skills and confidence in managing emergency situation.25 However, participants should perform hands-on-training to improve their skills, because mere knowledge-based training had no effect on the endoscopists’ performance.26 Since patient-based training can cause discomfort or even harm patients, simulation-based training modalities such as ex vivo animal models or computerised virtual reality simulators are commonly used to improve trainees skills in a safe environment and represent promising approaches to train basic and advanced endoscopic procedures.27 28

    Since our learning curve shows a first plateau of treatment failure rates after about 20 EGDs with SRH were performed, we recommend this as a minimum for an endoscopists to be considered proficient to perform emergency procedures independently. Endoscopist might be considered highly qualified experts in managing UGIB after a minimum of 50 haemostatic procedure performed. Our results also imply that hospitals should focus on maintaining a sufficiently trained on-call endoscopy service team to improve clinical management of patients with UGIB.

    Initially, the treatment failure rates of 28.6% in our study might appear high, however, the rebleeding rate was not higher than in other studies on endoscopic high-risk stigmata.29 Furthermore, the composite primary end point also included need for surgical and radiological haemostatic intervention as well as ongoing bleeding at the end of endoscopy, which occur naturally more frequently in cases with high-risk stigmata than in unselected study populations with UGIB.

    Compared with other studies on non-variceal UGIB, mortality rates were higher in our study. This might be explained by the significant number of cases with in-hospital bleeding, variceal bleeding and bleeding from malignancies included in this study population, which are associated with high mortality rates of up to 20%, 32% and 47.5%, respectively.22 30 31

    Despite higher treatment failure rates of endoscopists with low experience, the mortality rates tended to be lower in this group. However, these results have to be interpreted with caution, since endoscopists with low experience performed their endoscopies more frequently during regular hours under supervision. Similar to our study, a meta-analysis identified admission during off-hours as a risk factor for unfavourable outcome32 and the more favourable circumstances during regular hours might have improved patient’s survival. Moreover, urgent endoscopy within 6 hours is not recommended for stable patients with UGIB1 33 and thereby stable patients receive endoscopy more frequently during regular hours compared with critically ill and instable patients, who require urgent endoscopy directly after haemodynamic stabilisation.14 15

    The limitations of our study derive from its retrospective and monocentric design. Since this study was conducted at a tertiary centre with 24 hours/day endoscopy service with an experienced consultant as an on-demand back-up, the transferability of our results might be limited towards other institutions with less experience and resources. Additionally, due to the retrospective design, we were unable to clarify to which extent a supervising endoscopist supported the main endoscopist during the procedure. Furthermore, we were unable to retrospectively obtain the exact timing of endoscopy after presentation of bleeding. Therefore, we were not able to analyse the impact of experience level on timing of endoscopy and its influence on patients’ outcomes.

    In conclusion, our results demonstrate a clear learning curve in endoscopists’ performance in acute UGIB management. Treatment failure rates were significantly higher during the early learning phase. Based on our data, we recommend a minimum of 20 EGDs with SRH for an endoscopists to be considered proficient to perform emergency endoscopy independently. Endoscopist might be considered highly qualified experts in managing UGIB after a minimum of 50 haemostatic procedure performed. Implementing individual endoscopy lists in the education of trainees could improve their competence in managing acute UGIB.

    Data availability statement

    Data are available on reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    In accordance with the German law (paragraph 15, sentence 1, North Rhine Medical Association’s professional code of conduct from 14 November 1998 as amended on 19 November 2011, and paragraph 6, sentence 1, Health Data Protection Act of North Rhine-Westphalia), approval by a local Ethics Committee and written informed consent from the participants were not required because of the strictly retrospective design of our study. This study was performed in line with the principles of the Declaration of Helsinki.

    References

    1. ↵
      1. Lau JYW,
      2. Yu Y,
      3. Tang RSY, et al
      . Timing of endoscopy for acute upper gastrointestinal bleeding. N Engl J Med 2020;382:1299–308. doi:10.1056/NEJMoa1912484
      OpenUrlCrossRefPubMed
    2. ↵
      1. Blatchford O,
      2. Murray WR,
      3. Blatchford M
      . A risk score to predict need for treatment for uppergastrointestinal haemorrhage. Lancet 2000;356:1318–21. doi:10.1016/S0140-6736(00)02816-6
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Jairath V,
      2. Kahan BC,
      3. Gray A, et al
      . Restrictive versus liberal blood transfusion for acute upper gastrointestinal bleeding (TRIGGER): a pragmatic, open-label, cluster randomised feasibility trial. Lancet 2015;386:137–44. doi:10.1016/S0140-6736(14)61999-1
      OpenUrlCrossRefPubMed
    4. ↵
      1. Meier B,
      2. Wannhoff A,
      3. Denzer U, et al
      . Over-the-scope-clips versus standard treatment in high-risk patients with acute non-variceal upper gastrointestinal bleeding: a randomised controlled trial (STING-2). Gut 2022;71:1251–8. doi:10.1136/gutjnl-2021-325300
      OpenUrlAbstract/FREE Full Text
    5. ↵
      1. Pittayanon R,
      2. Khongka W,
      3. Linlawan S, et al
      . Hemostatic powder vs standard endoscopic treatment for gastrointestinal tumor bleeding: a multicenter randomized trial. Gastroenterology 2023;165:762–72. doi:10.1053/j.gastro.2023.05.042
      OpenUrl
    6. ↵
      1. Shung DL,
      2. Au B,
      3. Taylor RA, et al
      . Validation of a machine learning model that outperforms clinical risk scoring systems for upper gastrointestinal bleeding. Gastroenterology 2020;158:160–7. doi:10.1053/j.gastro.2019.09.009
      OpenUrlPubMed
    7. ↵
      1. Kanno T,
      2. Yuan Y,
      3. Tse F, et al
      . Proton pump inhibitor treatment initiated prior to endoscopic diagnosis in upper gastrointestinal bleeding. Cochrane Database Syst Rev 2022;1:CD005415. doi:10.1002/14651858.CD005415.pub4
    8. ↵
      1. HALT-IT Trial Collaborators
      . Effects of a high-dose 24-H infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT-IT): an international randomised, double-blind, placebo-controlled trial. Lancet 2020;395:1927–36. doi:10.1016/S0140-6736(20)30848-5
      OpenUrlCrossRefPubMed
    9. ↵
      1. Segal J,
      2. Siau K,
      3. Kanagasundaram C, et al
      . Training in Endotherapy for acute upper gastrointestinal bleeding: a UK-wide Gastroenterology Trainee survey. Frontline Gastroenterol 2020;11:430–5. doi:10.1136/flgastro-2019-101345
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Siau K,
      2. Dunckley P,
      3. Anderson J
      . PTU-010 exposure to endotherapy for upper gastrointestinal bleeding at the point of gastroscopy certification – is it sufficient? Gut 2017:A55. doi:10.26226/morressier.59a6b34ad462b80290b55688
    11. ↵
      1. Siau K,
      2. Morris AJ,
      3. Murugananthan A, et al
      . Variation in exposure to endoscopic haemostasis for acute upper gastrointestinal bleeding during UK gastroenterology training. Frontline Gastroenterol 2020;11:436–40. doi:10.1136/flgastro-2019-101351
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Penny HA,
      2. Kurien M,
      3. Wong E, et al
      . Changing trends in the UK management of upper GI bleeding: is there evidence of reduced UK training experience Frontline Gastroenterol 2016;7:67–72. doi:10.1136/flgastro-2014-100537
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Ward ST,
      2. Mohammed MA,
      3. Walt R, et al
      . An analysis of the learning curve to achieve competency at colonoscopy using the JETS database. Gut 2014;63:1746–54. doi:10.1136/gutjnl-2013-305973
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Gralnek IM,
      2. Stanley AJ,
      3. Morris AJ, et al
      . Endoscopic diagnosis and management of nonvariceal upper gastrointestinal hemorrhage (NVUGIH): European Society of Gastrointestinal Endoscopy (ESGE) guideline – update 2021. Endoscopy 2021;53:300–32. doi:10.1055/a-1369-5274
      OpenUrlPubMed
    15. ↵
      1. Laine L,
      2. Barkun AN,
      3. Saltzman JR, et al
      . ACG clinical guideline: upper gastrointestinal and ulcer bleeding. Am J Gastroenterol 2021;116:899–917. doi:10.14309/ajg.0000000000001245
      OpenUrlPubMed
    16. ↵
      1. Siau K,
      2. Hearnshaw S,
      3. Stanley AJ, et al
      . British Society Of Gastroenterology (BSG)-Led multisociety consensus care bundle for the early clinical management of acute upper gastrointestinal bleeding. Frontline Gastroenterol 2020;11:311–23. doi:10.1136/flgastro-2019-101395
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Rockall TA,
      2. Logan RF,
      3. Devlin HB, et al
      . Risk assessment after acute upper gastrointestinal haemorrhage. Gut 1996;38:316–21. doi:10.1136/gut.38.3.316
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Sung JJY,
      2. Tsoi KKF,
      3. Ma TKW, et al
      . Causes of mortality in patients with peptic ulcer bleeding: a prospective cohort study of 10,428 cases. Am J Gastroenterol 2010;105:84–9. doi:10.1038/ajg.2009.507
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Parente F,
      2. Anderloni A,
      3. Bargiggia S, et al
      . Outcome of non-variceal acute upper gastrointestinal bleeding in relation to the time of endoscopy and the experience of the endoscopist: a two-year survey. World J Gastroenterol 2005;11:7122–30. doi:10.3748/wjg.v11.i45.7122
      OpenUrlPubMed
    20. ↵
      1. Jensen DM,
      2. Kovacs T,
      3. Ghassemi KA, et al
      . Randomized controlled trial of over-the-scope clip as initial treatment of severe nonvariceal upper gastrointestinal bleeding. Clin Gastroenterol Hepatol 2021;19:2315–23. doi:10.1016/j.cgh.2020.08.046
      OpenUrlPubMed
    21. ↵
      1. Schmidt A,
      2. Gölder S,
      3. Goetz M, et al
      . Over-the-scope clips are more effective than standard endoscopic therapy for patients with recurrent bleeding of peptic ulcers. Gastroenterology 2018;155:674–86. doi:10.1053/j.gastro.2018.05.037
      OpenUrlCrossRefPubMed
    22. ↵
      1. Allo G,
      2. Bürger M,
      3. Chon S-H, et al
      . Efficacy of endoscopic therapy and long-term outcomes of upper gastrointestinal tumor bleeding in patients with esophageal cancer. Scand J Gastroenterol 2023;58:1064–70. doi:10.1080/00365521.2023.2199439
      OpenUrl
    23. ↵
      1. Kim Y-I,
      2. Choi IJ,
      3. Cho S-J, et al
      . Outcome of endoscopic therapy for cancer bleeding in patients with unresectable gastric cancer. J Gastroenterol Hepatol 2013;28:1489–95. doi:10.1111/jgh.12262
      OpenUrl
    24. ↵
      1. Adu-Tei S,
      2. Raju SA,
      3. Marks LJS, et al
      . Letter: enhancing training opportunities for upper GI bleeding in Sheffield—a UK transferable model Aliment Pharmacol Ther 2021;53:1241–2. doi:10.1111/apt.16354
      OpenUrlPubMed
    25. ↵
      1. Siau K,
      2. Fazal W,
      3. Thoufeeq M, et al
      . PWE-114 upper GI Haemostasis course improves delegate confidence in theoretical and practical aspects of Haemostasis management. British Society of Gastroenterology Annual Meeting, 17–20 June 2019, Abstracts; Posters. BMJ Publishing Group Ltd and British Society of Gastroenterology, June 2019 doi:10.1136/gutjnl-2019-BSGAbstracts.485
    26. ↵
      1. Haycock AV,
      2. Youd P,
      3. Bassett P, et al
      . Simulator training improves practical skills in therapeutic GI Endoscopy: results from a randomized, blinded, controlled study. Gastrointest Endosc 2009;70:835–45. doi:10.1016/j.gie.2009.01.001
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Camus M,
      2. Marteau P,
      3. Pocard M, et al
      . Validation of a live animal model for training in endoscopic hemostasis of upper gastrointestinal bleeding ulcers. Endoscopy 2013;45:451–7. doi:10.1055/s-0032-1326483
      OpenUrl
    28. ↵
      1. Ferlitsch A,
      2. Schoefl R,
      3. Puespoek A, et al
      . Effect of virtual endoscopy simulator training on performance of upper gastrointestinal endoscopy in patients: a randomized controlled trial. Endoscopy 2010;42:1049–56. doi:10.1055/s-0030-1255818
      OpenUrlCrossRefPubMed
    29. ↵
      1. de Groot NL,
      2. van Oijen MGH,
      3. Kessels K, et al
      . Reassessment of the predictive value of the forrest classification for peptic ulcer rebleeding and mortality: can classification be simplified Endoscopy 2014;46:46–52. doi:10.1055/s-0033-1344884
      OpenUrlCrossRefPubMed
    30. ↵
      1. Horvatits T,
      2. Mahmud N,
      3. Serper M, et al
      . MELD-lactate predicts poor outcome in variceal bleeding in cirrhosis. Dig Dis Sci 2023;68:1042–50. doi:10.1007/s10620-022-07744-w
      OpenUrl
    31. ↵
      1. Jairath V,
      2. Thompson J,
      3. Kahan BC, et al
      . Poor outcomes in hospitalized patients with gastrointestinal bleeding: impact of baseline risk, bleeding severity, and process of care. Am J Gastroenterol 2014;109:1603–12. doi:10.1038/ajg.2014.263
      OpenUrlPubMed
    32. ↵
      1. Xia XF,
      2. Chiu PWY,
      3. Tsoi KKF, et al
      . The effect of off-hours hospital admission on mortality and clinical outcomes for patients with upper gastrointestinal hemorrhage: a systematic review and meta-analysis of 20 cohorts. United European Gastroenterol J 2018;6:367–81. doi:10.1177/2050640617732651
      OpenUrl
    33. ↵
      1. Guo CLT,
      2. Wong SH,
      3. Lau LHS, et al
      . Timing of Endoscopy for acute upper gastrointestinal bleeding: a territory-wide cohort study. Gut 2022;71:1544–50. doi:10.1136/gutjnl-2020-323054
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors GA conceived of the presented idea, designed the study and is acting as guarantor. GA, AM and PK retrieved data. GA, XZ and SL performed the analytic calculations. S-HC, MB, DN, UT and TG helped supervise the project. GA, SL, AM, XZ, MB, S-HC, DN, UT, TG and PK contributed to the interpretation of the results. GA and PK took the lead in writing the manuscript.

    • Funding We acknowledge support for the Article Processing Charge from the DFG (German Research Foundation, 491454339).

    • Competing interests MB obtained consulting fees from Janssen and travel support from Pfizer.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.