UEG's educational platform for gastroenterology education bringing recordings, courses, guidelines, and more together in one place.

UEG Week Berlin 2025 Recordings Gut Guide online

Rate for better recommendations!

Your feedback helps us show you more of what you like and less of what you dislike!

More of this!

We will show you more similar content.

Not for me!

We won't suggest this again.

Bookmark!

Save this content for later!

New to myUEG? Create an account! Log In with myUEG account

UEG Poster

Standard Poster

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

REASONING CAPABILITIES OUTPERFORM LITERATURE EXPOSURE: EVALUATING LARGE LANGUAGE MODELS ON GASTROENTEROLOGY BOARD QUESTIONS

Burcu Gurbuz 1, Hakan Aydinli 1, Ibrahim Nahit Sendur 1, Petr Vanek 2, Cesare Hassan 3, Henriette Heinrich 4, Cem Simsek 1

Affiliations

1 Hacettepe University, Ankara, Türkiye

2 Palacký University Olomouc, Olomouc, Czechia

3 Humanitas University, Pieve Emanuele MI, Italy

4 Stadtspital Triemli, Zürich, Switzerland

Summary

AI Generated

Reasoning-optimized large language models outperform non-reasoning models on European Board of Gastroenterology and Hepatology examination questions, with performance gains attributable to explicit reasoning capabilities rather than larger literature exposure.

Reasoning-enhanced ChatGPT-o1 answered 171 of 203 EBGH questions correctly compared to 164 for ChatGPT-4o, rescuing 16 items that the non-reasoning model missed.
Non-reasoning model accuracy showed statistically significant correlations with publication metrics (total publications r=0.837, p=0.026; full-text count r=0.823, p=0.033), while reasoning models showed weaker correlations (total publications r=0.312, p=0.566).
Among previously failed questions retested in January 2025, ChatGPT-o1 Pro solved 15 of 17 remaining items, with 6 uniquely answered by this model and only 2 stumping all tested models.
Reasoning models consistently detected tacit clinical cues and offered broader diagnostic or therapeutic reasoning in qualitative review.
The authors conclude that explicit reasoning capabilities, rather than deep-search over larger corpora, drive performance on challenging gastroenterology board questions and may support domain-agnostic clinical decision-making.

Download PDF

Was this helpful?

Thanks for your feedback.

This summary was generated by an AI large language model based on the content transcript. It is for informational purposes only and should not be considered a substitute for clinical judgment. Always rely on your professional expertise and the full clinical context when making clinical decisions.

Abstract

Introduction

LLMs (large language models) already clear recall-focused tests, yet their capacity for nuanced clinical reasoning is still in doubt (1,2). Recently, two innovations have sought to close this gap: reasoning-optimized models that reveal a step-by-step “chain of thought,” and deep-search variants that iteratively comb vast text corpora to build evidence-based answers. Whether explicit reasoning, rather than sheer data volume, actually boosts performance on demanding assessments such as the EBGH (European Board of Gastroenterology and Hepatology) examination remains to be demonstrated.

Aims & Methods

We aimed to evaluate whether reasoning capabilities represent a significant advancement beyond increasing literature exposure during training.For this purpose, we administered 203 EBGH multiple-choice questions to five LLMs: non-reasoning models (ChatGPT-4o, Gemini-1.5-Pro, Llama-3.1-405B, DeepSeek-R1) and reasoning-optimized models (ChatGPT-o1, ChatGPT-o1 Pro). To standardize the querying process for each LLM, we employed the following prompt for every single question:"Act as a gastroenterology doctor. Please read the following question and select the single correct answer choice.”. Each item was run three times; accuracy and answer consistency were recorded. PubMed abstract counts, full texts, free full texts, and total papers across ten GI domains served as proxies for literature density.All tests used official API versions current between January 2024 and 15 February 2025.

Results

In December 2024, ChatGPT-4o answered 164 of 203 EBGH questions correctly, while the reasoning-enhanced o1 reached 171, rescuing 16 items that 4o had missed.When the 39 previously incorrect questions were retested in January 2025, 4o solved 22, leaving 17 that remained unsolved; o1 then solved seven of these,DeepResearch variant solved 5, and o1 Pro solved 15,with 6 items uniquely answered by o1 Pro and 2 stumping every model.Qualitative review showed that reasoning LLMs consistently detected tacit clinical cues and offered broader diagnostic or therapeutic reasoning.Total publication count showed the strongest correlation with overall accuracy(r:0.647, R²:0.418, p:0.167),followed by abstract count and full-text count(both r:0.618,R²:0.382,p:0.196), and free full-text availability(r:0.563, R²:0.317, p=0.255).With updated publication volumes, we found statistically significant correlations between publication metrics and non-reasoning model accuracy, with total publication count(r=0.837, R²:0.701, p:0.026), full-text count(r:0.823, R²:0.677, p:0.033), and abstract count(r=0.819,R²=0.670,p:0.035) all showing significant correlations. Reasoning models demonstrated weaker correlations with publication metrics compared to non-reasoning models, with total publication volume correlations of r:0.312(R²:0.097, p:0.566).We observed weak negative correlations between publication metrics and reasoning model consistency, with free full-text availability showing the strongest negative correlation(r:-0.352,R²:0.124, p:0.514). Interestingly, non-reasoning models showed weak positive correlations with publication metrics, with total publication count(r=0.388,R²=0.151,p=0.466) demonstrating the strongest relationship. However, none of these consistency correlations reached statistical significance.

Conclusion

Explicit reasoning, not deep-search over larger corpora, is the key driver of LLM success on challenging GI board questions. Reasoning-optimized models generalize across sparsely studied domains and may provide safer, domain-agnostic decision support for gastroenterologists.

References

1- Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepaño C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digit Health. 2023;2(2):e0000198.
2- Lee P, Bubeck S, Petro J. Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine. N Engl J Med. 2023;388(13):1233-9.

Conference

UEG Week Berlin 2025

Submission format

Abstract

Session

Quality, diversity and sustainability (Posters)

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG Presentation

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino 1

Affiliations

1 Kohnodai Hospital, National Center for Global Health and Medicine, Kōnodai, Japan

Summary

AI Generated

Summary is not available for this content yet.

Download PDF

Was this helpful?

Thanks for your feedback.

Abstract

Event

UEG Week Berlin 2025

Topics

Hepatobiliary

Session

Rising Stars from Europe and Japan: Metabolic liver diseases

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG Presentation

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

Time management: How can we save time?

Deepa Jegadish Gandhi 1

Affiliations

1 Mater Misericordiae University Hospital, Dublin, Ireland

Summary

AI Generated

Summary is not available for this content yet.

Download PDF

Was this helpful?

Thanks for your feedback.

Abstract

Event

UEG Week Berlin 2025

Topics

Nurses

Session

Staff safety and management

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG Presentation

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

European survey on physical and mental health in endoscopy

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

European survey on physical and mental health in endoscopy

Ute Pfeifer 1

Affiliations

1 Evangel. Krankenhaus Düsseldorf Abt. Endoskopie, Düsseldorf, Germany

Summary

AI Generated

Summary is not available for this content yet.

Download PDF

Was this helpful?

Thanks for your feedback.

Abstract

Event

UEG Week Berlin 2025

Topics

Nurses

Session

Staff safety and management

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG Presentation

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

Aletta Kata Kiss 1, Arpad Varga 1, Marietta Görög 2, Tamara Madacsy 1, Woo Young Chung 3, Petra Pallagi 1, Viktória Szabó 1, Petra Susánszki 2, Enikő Kúthy-Sutus 1, Szintia Barnai 2, Dániel Varga 4, Péter Bíró 4, Ingrid Hegnes Sendstad 1, Tim Crul 1, Boldizsár Jójárt 1, Bálint Tél 5, Anita Balazs 6, György Lázár 4, Miklós Erdélyi 4, Shmuel Muallem 3, József Maléth 1

Affiliations

1 University of Szeged, Szeged, Hungary|||University of Szeged, Szeged, Hungary|||University of Szeged, Szeged, Hungary

2 University of Szeged, Szeged, Hungary|||University of Szeged, Szeged, Hungary

3 National Institutes of Health, Bethesda, United States

4 University of Szeged, Szeged, Hungary

5 University of Szeged, Szeged, Hungary|||University of Szeged, Szeged, Hungary|||Semmelweis University, Budapest, Hungary

6 Charité Universitätsmedizin Berlin, Berlin, Germany

Summary

AI Generated

Summary is not available for this content yet.

Download PDF

Was this helpful?

Thanks for your feedback.

Abstract

Introduction

Cystic fibrosis transmembrane conductance regulator (CFTR) is essential for epithelial ion and fluid secretion. While CFTR is primarily activated via cAMP/PKA signaling, the role of intracellular Ca²⁺ signaling—particularly through store-independent mechanisms—remains incompletely understood.

Aims & Methods

We aimed to elucidate how store-independent Ca²⁺ influx via Orai1 regulates basal CFTR activity in primary polarized epithelial cells.
To investigate this we used mouse and human pancreatic, hepatic and airway organoids, we combined RNA-sequencing, immunofluorescence, Fura-2 and MQAE dye-based Ca²⁺/Cl⁻ imaging, FLIM-FRET, dSTORM microscopy, and in vivo fluid secretion assays. Genetic silencing and pharmacological inhibition (CM5480) were employed to dissect Orai1-dependent signaling pathways.

Results

We discovered that Orai1 mediates a constitutive, store-independent Ca²⁺ influx sustained by SPCA2. This influx occurs independently of ER Ca²⁺ depletion but requires STIM1. Super-resolution imaging revealed Orai1 clustering into discrete apical membrane nanodomains that colocalize with SPCA2, STIM1, PKA, and CFTR. Functional assays confirmed that this localized signaling complex maintains basal CFTR activity. Inhibiting Orai1 reduced basal Cl⁻ and HCO₃⁻ secretion, disrupted mitochondrial ATP production, and suppressed pancreatic ductal fluid secretion in vitro and in vivo. Further, the stimulatory effect of this Ca²⁺ influx on CFTR was mediated via Ca²⁺/calmodulin-sensitive adenylyl cyclases (AC1, 3, 8), which were co-clustered with Orai1 and CFTR. Their knockdown phenocopied Orai1 inhibition, and no additive effect was observed when both pathways were blocked.

Conclusion

Our results reveal a previously unrecognized SPCA2–STIM1–Orai1 nanodomain that drives store-independent Ca²⁺ influx, activating CFTR via local cAMP production in the absence of neurohormonal stimuli. This regulatory module is conserved across secretory epithelia and represents a key mechanism for maintaining basal fluid secretion and epithelial homeostasis.

Event

UEG Week Berlin 2025

Topics

Mechanisms & Personalised Medicine Pancreas

Submission format

Abstract

Session

Cogulation, lipids and bugs: All relevant for pancreatitis

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG Poster

Standard Poster

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Federica Borrelli de Andreis 1, Salvatore Francesco Vadalà di Prampero 2, Ilaria Simonelli 3, Francesca Motta 2, Silvana Fusha 2, Francesco Esposito 4, Alessia Mancini 2, Anna Di Gemma 2, Guido Costamagna 2, Milutin Bulajic 2

Affiliations

1 Catholic University of Sacred Heart, Rome, Italy|||Gemelli Isola - Isola Tiberina Hospital, Roma, Italy

2 Gemelli Isola - Isola Tiberina Hospital, Roma, Italy

3 Isola Tiberina - Gemelli Isola Hospital, Rome, Italy

4 Gemelli Isola - Isola Tiberina Hospital, Rome, Italy|||Sapienza University of Rome, Rome, Italy

Summary

AI Generated

Summary is not available for this content yet.

Download PDF

Was this helpful?

Thanks for your feedback.

Abstract

Introduction

Gastrointestinal (GI) endoscopy requires not only technical skills but also strong non-technical abilities like communication, teamwork, and decision-making. Traditional training is often individualized, with little focus on interdisciplinary collaboration. Role Swapping (RS) — where physicians and nurses exchange tasks during simulated procedures — may enhance teamwork and mutual understanding.

Aims & Methods

A prospective pilot study was conducted involving 22 participants (10 endoscopists, 12 nurses) using in-silico colonoscopy simulators. The training included a pre-simulation skills assessment, theoretical onsite classes, randomized role-swapping simulation sessions, and a post-simulation evaluation after two weeks. Technical skills, non-technical skills (assessed via NOTSS and SPLINTS frameworks), and satisfaction were evaluated using validated questionnaires. Wilcoxon tests compared pre- and post-training scores.

Results

RS in simulated GI endoscopy training significantly enhanced both technical and non-technical competencies among participants. Post-training assessments using validated tools showed improvements across all evaluated domains. Situational awareness, decision-making, communication, teamwork, and leadership skills all exhibited statistically significant gains. Participants improved their ability to gather information (p=0.02), understand data (p=0.008), and anticipate future scenarios (p=0.002). Decision-making skills were notably enhanced, with better option consideration (p<0.001), more transparent communication of decisions (p=0.001), and greater adaptability to dynamic conditions (p<0.001). Communication improvements included more effective information exchange (p=0.007), better shared understanding (p=0.035), and increased team coordination (p=0.001). Leadership skills were also significantly strengthened, particularly in setting standards (p=0.002) and supporting teammates (p=0.012). Regarding technical skills, endoscopists reported substantial confidence improvements in tasks traditionally performed by nurses. Significant progress was observed in preparing endoscopic equipment (p=0.016), medication dilution (p=0.016), patient discharge procedures (p=0.014), instrument arrangement during interventional colonoscopies (p=0.026), and setup of patient monitoring systems (p=0.041). Confidence also increased in tasks initially rated as low-confidence, such as proper waste disposal after GI procedures (p=0.01). Nurses, conversely, enhanced their technical capabilities related to GI endoscopy performance. They showed greater proficiency in advancing the colonoscope (p=0.004), reducing loops (p=0.007), navigating angled turns (p=0.007), executing ancillary maneuvers (p=0.038), and conducting thorough mucosal inspections during withdrawal (p=0.036). Improvements were also seen in colonoscope insertion (p=0.026), understanding of colonoscopy quality indicators (p=0.011), and lesion identification (p=0.046). Participant feedback revealed high satisfaction with the RS training experience. Participants appreciated the opportunity to engage in cross-disciplinary tasks, reporting that the simulation increased their mutual respect, role awareness, and confidence in collaborative practices.

Conclusion

Role swapping during simulated GI endoscopy training enhanced both technical and non-technical skills and fostered better collaboration between nurses and physicians. This innovative educational approach shows potential to improve clinical practice quality and patient safety.

References

1. Hedges AR, Johnson HJ, Kobulinsky LR, et al. Effects of Cross-Training on Medical Teams’ Teamwork and Collaboration: Use of Simulation. Pharmacy (Basel) 2019; 7: 13.
2. de Laurentis C, Pirillo D, Di Cristofori A, et al. Boosting teamwork between scrub nurses and neurosurgeons: exploring the value of a role-played hands-on, cadaver-free simulation and systematic review of the literature. Front Surg 2024; 11: 1386887.

Conference

UEG Week Berlin 2025

Submission format

Abstract

Session

Endoscopy: New techniques and training (Posters)

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG Presentation

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

Accreditation for a newly recruited endoscopy nurse

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

Accreditation for a newly recruited endoscopy nurse

Anne Vidil 1

Affiliations

1 None, None, France

Summary

AI Generated

Summary is not available for this content yet.

Download PDF

Was this helpful?

Thanks for your feedback.

Abstract

Event

UEG Week Berlin 2025

Topics

Nurses

Session

Education: Important tool for quality in endoscopy

Citation

United European Gastroenterology Journal 2025; 13 (Supplement 8)

Published

2025

More Like This:

Unraveling Steatotic Liver Diseases: Insights from Amino Acid and TCA Cycle Metabolomics

Masaaki Mino

Time management: How can we save time?

Deepa Jegadish Gandhi

European survey on physical and mental health in endoscopy

Ute Pfeifer

STORE-INDEPENDENT ORAI1 SIGNALING DEFINES BASAL CFTR FUNCTION IN SECRETORY EPITHELIA

József Maléth

ROLE SWAPPING BETWEEN NURSES AND PHYSICIANS IN SIMULATED ENDOSCOPY TRAINING ENHANCES COLLABORATION AND TEAMWORK: A PROSPECTIVE OBSERVATIONAL PILOT STUDY

Milutin Bulajic

Accreditation for a newly recruited endoscopy nurse

Anne Vidil

UEG's educational platform for gastroenterology education bringing recordings, courses, guidelines, and more together in one place.

Most popular content

Rate for better recommendations!