{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T23:55:09Z","timestamp":1781654109139,"version":"3.54.5"},"reference-count":14,"publisher":"Georg Thieme Verlag KG","issue":"CP","license":[{"start":{"date-parts":[[2026,6,3]],"date-time":"2026-06-03T00:00:00Z","timestamp":1780444800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Endosc Int Open"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Introduction Large language models (LLMs) are increasingly used for clinical decision support and show high adherence to clinical practice guidelines. However, LLM-generated responses may translate guideline recommendations into definitive clinical advice without adequately reflecting the underlying strength of recommendations or the quality of supporting evidence. This study aimed to evaluate the risk of guideline overselling in GPT-generated responses based on guidelines.<\/jats:p>\n                  <jats:p>Methods Four ESGE colorectal guidelines were selected. All guideline statements were extracted together with their strength of recommendation and the quality of evidence. For each statement, a clinical vignette and related question were submitted to GPT-5 using both generic and guideline-oriented prompts. Three independent experts assessed each response for adherence to guideline content and calibration of wording.<\/jats:p>\n                  <jats:p>Results Overall, GPT-generated responses showed high adherence to guideline content with generic prompts (54\/59, 91.5%), whereas calibration to recommendation strength and evidence quality was lower (44\/59, 74.6%). Calibration was significantly reduced for weak versus strong recommendations (33.3% vs. 86.4%, P &lt; 0.001) and for statements supported by low-quality versus high- or moderate-quality evidence (44.4% vs. 90.2%, P &lt; 0.001). After prompt calibration experiment, adherence increased to 96.6% (57\/59), while calibration markedly improved to 94.9% (56\/59), mainly through reduced overly assertive wording for low-certainty recommendations.<\/jats:p>\n                  <jats:p>Conclusions GPT-5 demonstrated high guideline adherence but limited calibration to the quality of evidence underlying individual recommendations, with the potential to amplify the perceived certainty of low-quality evidence. Explicit incorporation of recommendation strength and evidence quality within prompts substantially improves calibration and may mitigate this risk in clinical use.<\/jats:p>","DOI":"10.1055\/a-2889-0169","type":"journal-article","created":{"date-parts":[[2026,6,3]],"date-time":"2026-06-03T22:57:17Z","timestamp":1780527437000},"source":"Crossref","is-referenced-by-count":0,"special_numbering":"S","title":["Risk of Guideline Overselling and Certainty Inflation in GPT-5 Responses: A Statement-Based Analysis of Four ESGE Colorectal Guidelines"],"prefix":"10.1055","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4992-9289","authenticated-orcid":false,"given":"Marcello","family":"Maida","sequence":"additional","affiliation":[{"name":"Kore University of Enna, Sicily, Italy","place":["Enna"],"department":["Department of Medicine and Surgery"]},{"name":"Umberto I Hospital, Sicily, Italy","place":["Enna"],"department":["Gastroenterology Unit"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2107-2156","authenticated-orcid":false,"given":"Antonio","family":"Facciorusso","sequence":"additional","affiliation":[{"name":"Universit\u00e0 del Salento, Italy","place":["Lecce"],"department":["Gastroenterology Unit, Department of Experimental Medicine"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lorenzo","family":"Fuccio","sequence":"additional","affiliation":[{"name":"University of Bologna, Emilia-Romagna, Italy","place":["Bologna"],"department":["Department of Gastroenterology"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3304-0271","authenticated-orcid":false,"given":"Tony","family":"Tham","sequence":"additional","affiliation":[{"name":"Ulster Hospital, Northern Ireland, United Kingdom of Great Britain and Northern Ireland","place":["Dundonald"],"department":["Division Of Gastroenterology"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alessandro","family":"Vitello","sequence":"additional","affiliation":[{"name":"Kore University of Enna, Sicily, Italy","place":["Enna"],"department":["Department of Medicine and Surgery"]},{"name":"Umberto I Hospital, Sicily, Italy","place":["Enna"],"department":["Gastroenterology Unit"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0802-3173","authenticated-orcid":false,"given":"Giulio","family":"Calabrese","sequence":"additional","affiliation":[{"name":"University of Naples Federico II, Campania, Italy","place":["Naples"],"department":["Clinical Medicine and Surgery"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"194","published-online":{"date-parts":[[2026,6,16]]},"reference":[{"issue":"19","key":"ref1","doi-asserted-by":"publisher","first-page":"3328","DOI":"10.3390\/cancers16193328","article-title":"The application of large language models in gastroenterology: a review of the literature","volume":"16","author":"M Maida","year":"2024","journal-title":"Cancers"},{"issue":"03","key":"ref2","doi-asserted-by":"publisher","first-page":"262","DOI":"10.1055\/a-2388-6084","article-title":"And the AI-CORE (artificial intelligence COlorectal cancer research) working group. The role of generative language systems in increasing patient awareness of colon cancer screening","volume":"57","author":"M Maida","year":"2025","journal-title":"Endoscopy"},{"issue":"13","key":"ref3","doi-asserted-by":"publisher","first-page":"a25689416","DOI":"10.1055\/a-2568-9416. PMID: 40376022; PMCID: PMC12080512","article-title":"Exploring ChatGPT effectiveness in addressing direct patient queries on colorectal cancer screening","volume":"12","author":"M Maida","year":"2025","journal-title":"Endosc Int Open"},{"issue":"09","key":"ref4","doi-asserted-by":"publisher","first-page":"1917.e17","DOI":"10.1016\/j.cgh.2024.04.022","article-title":"ChatGPT4 outperforms Endoscopists for determination of Postcolonoscopy rescreening and surveillance recommendations","volume":"22","author":"P W Chang","year":"2024","journal-title":"Clin Gastroenterol Hepatol"},{"issue":"01","key":"ref5","doi-asserted-by":"publisher","first-page":"81","DOI":"10.1111\/jgh.16375","article-title":"ChatGPT on guidelines: providing contextual knowledge to GPT allows it to provide advice on appropriate colonoscopy intervals","volume":"39","author":"D YZ Lim","year":"2024","journal-title":"J Gastroenterol Hepatol"},{"issue":"7650","key":"ref6","doi-asserted-by":"crossref","first-page":"924","DOI":"10.1136\/bmj.39489.470347.AD","article-title":"GRADE: an emerging consensus on rating quality of evidence and strength of recommendations","volume":"336","author":"G H Guyatt","year":"2008","journal-title":"BMJ"},{"key":"ref7","volume-title":"GRADE Handbook for Guideline Development","author":"H J Sch\u00fcnemann","year":"2023"},{"issue":"01","key":"ref8","doi-asserted-by":"publisher","first-page":"358","DOI":"10.1038\/s41746-024-01356-6","article-title":"Autonomous medical evaluation for guideline adherence of large language models","volume":"7","author":"D Fast","year":"2024","journal-title":"NPJ Digit Med"},{"issue":"08","key":"ref9","doi-asserted-by":"publisher","first-page":"687","DOI":"10.1055\/a-1185-3109","article-title":"Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy (ESGE) guideline - update 2020","volume":"52","author":"C Hassan","year":"2020","journal-title":"Endoscopy"},{"issue":"03","key":"ref10","doi-asserted-by":"publisher","first-page":"C1","DOI":"10.1055\/a-0854-5925","article-title":"Endoscopic surveillance after surgical or endoscopic resection for colorectal cancer: European Society of Gastrointestinal Endoscopy (ESGE) and European Society of Digestive Oncology (ESDO) guideline","volume":"51","author":"C Hassan","year":"2019","journal-title":"Endoscopy"},{"issue":"11","key":"ref11","doi-asserted-by":"publisher","first-page":"1082","DOI":"10.1055\/a-1016-4977","article-title":"Endoscopic management of lynch syndrome and of familial risk of colorectal cancer: European Society of Gastrointestinal Endoscopy (ESGE) guideline","volume":"51","author":"M E van Leerdam","year":"2019","journal-title":"Endoscopy"},{"issue":"07","key":"ref12","doi-asserted-by":"publisher","first-page":"516","DOI":"10.1055\/a-2304-3219","article-title":"Colorectal polypectomy and endoscopic mucosal resection: European Society of Gastrointestinal Endoscopy (ESGE) guideline - update 2024","volume":"56","author":"M Ferlitsch","year":"2024","journal-title":"Endoscopy"},{"key":"ref13","doi-asserted-by":"crossref","first-page":"159","DOI":"10.2307\/2529310","article-title":"The measurement of observer agreement for categorical data","volume":"33","author":"J R Landis","year":"1977","journal-title":"Biometrics"},{"issue":"08","key":"ref14","doi-asserted-by":"publisher","first-page":"1689178","DOI":"10.3389\/frai.2025.1689178","article-title":"In the search for the perfect prompt in medical AI queries","volume":"17","author":"F Berghea","year":"2025","journal-title":"Front Artif Intell"}],"container-title":["Endoscopy International Open"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/www.thieme-connect.de\/products\/ejournals\/pdf\/10.1055\/a-2889-0169.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T23:22:32Z","timestamp":1781652152000},"score":1,"resource":{"primary":{"URL":"http:\/\/www.thieme-connect.de\/DOI\/DOI?10.1055\/a-2889-0169"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,3]]},"references-count":14,"journal-issue":{"issue":"CP","published-online":{"date-parts":[[2026,1,13]]},"published-print":{"date-parts":[[2026]]}},"alternative-id":["a28890169"],"URL":"https:\/\/doi.org\/10.1055\/a-2889-0169","archive":["Portico","CLOCKSS"],"relation":{},"ISSN":["2364-3722","2196-9736"],"issn-type":[{"value":"2364-3722","type":"print"},{"value":"2196-9736","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,6,3]]},"published":{"date-parts":[[2026,6,3]]}}}