{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,18]],"date-time":"2026-06-18T13:52:49Z","timestamp":1781790769475,"version":"3.54.5"},"reference-count":34,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,6,18]],"date-time":"2026-06-18T00:00:00Z","timestamp":1781740800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,6,18]],"date-time":"2026-06-18T00:00:00Z","timestamp":1781740800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Comprehensive Cancer Center (CCC) Hannover"},{"DOI":"10.13039\/501100002347","name":"Bundesministerium f\u00fcr Forschung, Technologie und Raumfahrt","doi-asserted-by":"publisher","award":["01ZZ2021"],"award-info":[{"award-number":["01ZZ2021"]}],"id":[{"id":"10.13039\/501100002347","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Medizinische Hochschule Hannover (MHH)"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Arch Gynecol Obstet"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>This prospective multicenter study aimed to compare the decision-making abilities of board-certified colposcopists and two commercially available large language models (LLM), ChatGPT-4o and ChatGPT-5, in cervical dysplasia management.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Twenty-three anonymized real-life patient cases with multiple-choice (MC) questions regarding treatment decisions were used to assess answer quality. Ten board-certified colposcopists and the two LLMs addressed the MC\u00a0questions. The gold standard was defined by two guideline authors. LLMs were prompted to justify their responses. Concordance rates were calculated and compared across all questions and histopathological subgroups, including cervical intraepithelial neoplasia (CIN), unspecific histopathological results, and cervical cancer cases.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Clinicians and LLMs achieved similar overall concordance rates compared to the gold standard (69.6% for clinicians, 69.6% for ChatGPT-4o, and 65.2% for ChatGPT-5). ChatGPT-5 outperformed clinicians in precancerous lesions (81.8% vs. 66.4%), while clinicians excelled in complex cases with unspecific histopathology (86% vs. 60%). Clinicians showed a tendency to overtreat low-grade lesions (CIN I), opting for more intensive surveillance. ChatGPT-4o performed better than ChatGPT-5 in cervical cancer cases, though both models struggled with these scenarios.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>This study highlights the potential of LLMs as decision support tools in cervical dysplasia management, particularly for straightforward cases like precancerous lesions. However, clinicians remain superior in handling complex or ambiguous cases. The tendency of clinicians to overtreat low-grade lesions may offer the potential to test the implementation of a decision support tool for those cases. While LLMs show promise, exploring open-ended clinical scenarios and integrating retrieval-augmented generation could enhance their practical application.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1007\/s00404-026-08498-w","type":"journal-article","created":{"date-parts":[[2026,6,18]],"date-time":"2026-06-18T12:41:53Z","timestamp":1781786513000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Comparing clinical decision-making between colposcopists and large language models in cervical dysplasia management: a pilot prospective multicenter study"],"prefix":"10.1007","volume":"313","author":[{"given":"Jan Lennart","family":"Stalp","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Juliane Alexandra","family":"Schneider","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lena","family":"Steinkasserer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jens","family":"Hachenberg","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Matthias","family":"Jentschke","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter","family":"Hillemanns","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dominik","family":"Wolff","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Agnieszka","family":"Denecke","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,6,18]]},"reference":[{"key":"8498_CR1","unstructured":"Deutsche Krebsgesellschaft (DKG) e.V. (2025) Kennzahlenauswertung 2025 Jahresbericht der zertifizierten Gyn\u00e4kologischen Dysplasie-Einheiten und Gyn\u00e4kologischen Dysplasie-Sprechstunden"},{"key":"8498_CR2","unstructured":"Leitlinienprogramm Onkologie (Deutsche Krebsgesellschaft, Deutsche Krebshilfe, AWMF): S3-Leitlinie Pr\u00e4vention des Zervixkarzinoms, Langversion 1.1, 2020, AWMF Registernummer: 015\/027OL, http:\/\/www.leitlinienprogramm-onkologie.de\/leitlinien\/zervixkarzinom-praevention\/"},{"key":"8498_CR3","unstructured":"Leitlinienprogramm Onkologie (Deutsche Krebsgesellschaft, Deutsche Krebshilfe, AWMF): S3-Leitlinie Diagnostik, Therapie und Nachsorge der Patientin mit Zervixkarzinom, Langversion, 2.2, 2022, AWMF-Registernummer: 032\/033OL, https:\/\/www.leitlinienprogramm-onkologie.de\/leitlinien\/zervixkarzinom\/"},{"key":"8498_CR4","doi-asserted-by":"publisher","first-page":"503","DOI":"10.1177\/030089160208800614","volume":"88","author":"M Palazzi","year":"2002","unstructured":"Palazzi M, De Tomasi D, D\u2019Affronto C et al (2002) Are international guidelines for the prescription of adjuvant treatment for early breast cancer followed in clinical practice? Results of a population-based study on 1547 patients. Tumori Journal 88:503\u2013506. https:\/\/doi.org\/10.1177\/030089160208800614","journal-title":"Tumori Journal"},{"key":"8498_CR5","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0279116","volume":"17","author":"M Bierbaum","year":"2022","unstructured":"Bierbaum M, Rapport F, Arnolda G et al (2022) Clinical practice guideline adherence in oncology: a qualitative study of insights from clinicians in Australia. PLoS ONE 17:e0279116. https:\/\/doi.org\/10.1371\/journal.pone.0279116","journal-title":"PLoS ONE"},{"key":"8498_CR6","doi-asserted-by":"publisher","first-page":"860","DOI":"10.1055\/s-0034-1383042","volume":"74","author":"M Beckmann","year":"2014","unstructured":"Beckmann M, Quaas J, Bischofberger A et al (2014) Establishment of the certification system \u201cGynaecological dysplasia\u201d in Germany. Geburtshilfe Frauenheilkd 74:860\u2013867. https:\/\/doi.org\/10.1055\/s-0034-1383042","journal-title":"Geburtshilfe Frauenheilkd"},{"key":"8498_CR7","doi-asserted-by":"publisher","DOI":"10.2196\/69672","volume":"11","author":"T Wu","year":"2025","unstructured":"Wu T, Wang Y, Cui X et al (2025) AI-based identification method for cervical transformation zone within digital colposcopy: development and multicenter validation study. JMIR Cancer 11:e69672. https:\/\/doi.org\/10.2196\/69672","journal-title":"JMIR Cancer"},{"issue":"23","key":"8498_CR8","doi-asserted-by":"publisher","first-page":"e703","DOI":"10.1016\/S2589-7500(23)00153-X","volume":"7500","author":"K Dembrower","year":"2023","unstructured":"Dembrower K, Crippa A, Col\u00f3n E et al (2023) Artificial intelligence for breast cancer detection in screening mammography in Sweden: a prospective, population-based, paired-reader, non-inferiority study. The Lancet Digital Health 7500(23):e703\u2013e711. https:\/\/doi.org\/10.1016\/S2589-7500(23)00153-X","journal-title":"The Lancet Digital Health"},{"key":"8498_CR9","doi-asserted-by":"publisher","first-page":"89","DOI":"10.1038\/s41586-019-1799-6","volume":"577","author":"SM McKinney","year":"2020","unstructured":"McKinney SM, Sieniek M, Godbole V et al (2020) International evaluation of an AI system for breast cancer screening. Nature 577:89\u201394. https:\/\/doi.org\/10.1038\/s41586-019-1799-6","journal-title":"Nature"},{"key":"8498_CR10","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.222639","author":"JH Yoon","year":"2023","unstructured":"Yoon JH, Strand F, Baltzer PAT et al (2023) Standalone AI for breast cancer detection at screening digital mammography and digital breast tomosynthesis: a systematic review and meta-analysis. Radiology. https:\/\/doi.org\/10.1148\/radiol.222639","journal-title":"Radiology"},{"key":"8498_CR11","doi-asserted-by":"publisher","DOI":"10.1080\/10872981.2025.2592430","author":"P Ros-Arlanz\u00f3n","year":"2025","unstructured":"Ros-Arlanz\u00f3n P, Gutarra-\u00c1vila R, Arrarte-Esteban V et al (2025) When AI models take the exam: large language models vs medical students on multiple-choice course exams. Med Educ Online. https:\/\/doi.org\/10.1080\/10872981.2025.2592430","journal-title":"Med Educ Online"},{"key":"8498_CR12","doi-asserted-by":"publisher","DOI":"10.7759\/cureus.90212","author":"R Zhang","year":"2025","unstructured":"Zhang R, Cai Q, Sartori A et al (2025) Comparing artificial intelligence large language models in medical training: a performance analysis of ChatGPT and DeepSeek on United States Medical Licensing Examination (USMLE) style questions. Cureus. https:\/\/doi.org\/10.7759\/cureus.90212","journal-title":"Cureus"},{"key":"8498_CR13","doi-asserted-by":"publisher","first-page":"3845","DOI":"10.3390\/curroncol31070284","volume":"31","author":"JL Stalp","year":"2024","unstructured":"Stalp JL, Denecke A, Jentschke M et al (2024) Quality of ChatGPT-generated therapy recommendations for breast cancer treatment in gynecology. Curr Oncol 31:3845\u20133854. https:\/\/doi.org\/10.3390\/curroncol31070284","journal-title":"Curr Oncol"},{"key":"8498_CR14","doi-asserted-by":"publisher","first-page":"1502","DOI":"10.3390\/jpm13101502","volume":"13","author":"S Griewing","year":"2023","unstructured":"Griewing S, Gremke N, Wagner U et al (2023) Challenging ChatGPT 3.5 in Senology\u2014an assessment of concordance with breast cancer tumor board decision making. J Pers Med 13:1502. https:\/\/doi.org\/10.3390\/jpm13101502","journal-title":"J Pers Med"},{"key":"8498_CR15","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1038\/s41523-023-00557-8","volume":"9","author":"V Sorin","year":"2023","unstructured":"Sorin V, Klang E, Sklair-Levy M et al (2023) Large language model (ChatGPT) as a support tool for breast tumor board. NPJ Breast Cancer 9:44\u201344. https:\/\/doi.org\/10.1038\/s41523-023-00557-8","journal-title":"NPJ Breast Cancer"},{"key":"8498_CR16","doi-asserted-by":"publisher","first-page":"1831","DOI":"10.1007\/s00404-023-07130-5","volume":"308","author":"S Lukac","year":"2023","unstructured":"Lukac S, Dayan D, Fink V et al (2023) Evaluating ChatGPT as an adjunct for the multidisciplinary tumor board decision-making in primary breast cancer cases. Arch Gynecol Obstet 308:1831\u20131844. https:\/\/doi.org\/10.1007\/s00404-023-07130-5","journal-title":"Arch Gynecol Obstet"},{"key":"8498_CR17","doi-asserted-by":"publisher","unstructured":"Roy S, Khatua A, Ghoochani F, et al (2024) Beyond Accuracy: Investigating Error Types in GPT-4 Responses to USMLE Questions. In: Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval. ACM, New York, NY, USA, pp 1073\u20131082. https:\/\/doi.org\/10.1145\/3626772.3657882","DOI":"10.1145\/3626772.3657882"},{"key":"8498_CR18","doi-asserted-by":"publisher","first-page":"37","DOI":"10.1177\/001316446002000104","volume":"20","author":"J Cohen","year":"1960","unstructured":"Cohen J (1960) A coefficient of agreement for nominal scales. Educ Psychol Meas 20:37\u201346. https:\/\/doi.org\/10.1177\/001316446002000104","journal-title":"Educ Psychol Meas"},{"key":"8498_CR19","doi-asserted-by":"publisher","first-page":"379","DOI":"10.1002\/j.1538-7305.1948.tb01338.x","volume":"27","author":"CE Shannon","year":"1948","unstructured":"Shannon CE (1948) A mathematical theory of communication. Bell Syst Tech J 27:379\u2013423. https:\/\/doi.org\/10.1002\/j.1538-7305.1948.tb01338.x","journal-title":"Bell Syst Tech J"},{"key":"8498_CR20","doi-asserted-by":"publisher","unstructured":"Waskom M (2021) seaborn: statistical data visualization. J Open Source Softw 6:3021. https:\/\/doi.org\/10.21105\/joss.03021","DOI":"10.21105\/joss.03021"},{"key":"8498_CR21","doi-asserted-by":"publisher","first-page":"90","DOI":"10.1109\/MCSE.2007.55","volume":"9","author":"JD Hunter","year":"2007","unstructured":"Hunter JD (2007) Matplotlib: a 2D graphics environment. Comput Sci Eng 9:90\u201395. https:\/\/doi.org\/10.1109\/MCSE.2007.55","journal-title":"Comput Sci Eng"},{"key":"8498_CR22","doi-asserted-by":"crossref","unstructured":"Wilcox AR (1967) Indices of Qualitative Variation. Oak Ridge National Laboratory. https:\/\/www.osti.gov\/servlets\/purl\/4167340. Accessed 21 Apr 2026.","DOI":"10.2172\/4167340"},{"key":"8498_CR23","doi-asserted-by":"publisher","first-page":"819","DOI":"10.1056\/NEJMoa2308900","volume":"390","author":"M Plante","year":"2024","unstructured":"Plante M, Kwon JS, Ferguson S et al (2024) Simple versus radical hysterectomy in women with low-risk cervical cancer. N Engl J Med 390:819\u2013829. https:\/\/doi.org\/10.1056\/NEJMoa2308900","journal-title":"N Engl J Med"},{"key":"8498_CR24","doi-asserted-by":"publisher","first-page":"1336","DOI":"10.1093\/jnci\/djg037","volume":"95","author":"NF Schlecht","year":"2003","unstructured":"Schlecht NF (2003) Human papillomavirus infection and time to progression and regression of cervical intraepithelial neoplasia. CancerSpectrum Knowl Environ 95:1336\u20131343. https:\/\/doi.org\/10.1093\/jnci\/djg037","journal-title":"CancerSpectrum Knowl Environ"},{"key":"8498_CR25","volume-title":"Best care at lower cost","author":"Institute of Medicine","year":"2013","unstructured":"Institute of Medicine (2013) Best care at lower cost. National Academies Press, Washington"},{"key":"8498_CR26","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0181970","volume":"12","author":"H Lyu","year":"2017","unstructured":"Lyu H, Xu T, Brotman D et al (2017) Overtreatment in the United States. PLoS ONE 12:e0181970. https:\/\/doi.org\/10.1371\/journal.pone.0181970","journal-title":"PLoS ONE"},{"key":"8498_CR27","doi-asserted-by":"publisher","first-page":"110351","DOI":"10.1016\/j.compbiomed.2025.110351","volume":"192","author":"J Vrdoljak","year":"2025","unstructured":"Vrdoljak J, Boban Z, Males I et al (2025) Evaluating large language and large reasoning models as decision support tools in emergency internal medicine. Comput Biol Med 192:110351\u2013110351. https:\/\/doi.org\/10.1016\/j.compbiomed.2025.110351","journal-title":"Comput Biol Med"},{"key":"8498_CR28","doi-asserted-by":"publisher","first-page":"51","DOI":"10.15586\/aei.v53i4.1353","volume":"52","author":"ME Gerek","year":"2025","unstructured":"Gerek ME, \u00d6nalan T, \u00c7\u00f6lkesen F (2025) Evaluating large language models for WAO\/EAACI guideline compliance in hereditary angioedema management. Allergol Immunopathol (Madr) 52:51\u201359. https:\/\/doi.org\/10.15586\/aei.v53i4.1353","journal-title":"Allergol Immunopathol (Madr)"},{"key":"8498_CR29","doi-asserted-by":"publisher","first-page":"625","DOI":"10.1038\/s41586-024-07421-0","volume":"630","author":"S Farquhar","year":"2024","unstructured":"Farquhar S, Kossen J, Kuhn L, Gal Y (2024) Detecting hallucinations in large language models using semantic entropy. Nature 630:625\u2013630. https:\/\/doi.org\/10.1038\/s41586-024-07421-0","journal-title":"Nature"},{"issue":"1","key":"8498_CR30","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1097\/lgt.0000000000000788","volume":"28","author":"RB Perkins","year":"2024","unstructured":"Perkins RB, Guido RS, Castle PE et al (2024) 2019 ASCCP risk-based management consensus guidelines: updates through 2023. J Low Genit Tract Dis 28(1):3\u20136. https:\/\/doi.org\/10.1097\/lgt.0000000000000788","journal-title":"J Low Genit Tract Dis"},{"key":"8498_CR31","unstructured":"Lewis P, Perez E, Piktus A, et al (2021) Retrieval-augmented generation for knowledge-intensive NLP tasks"},{"key":"8498_CR32","doi-asserted-by":"publisher","first-page":"e78393","DOI":"10.2196\/78393","volume":"27","author":"JYM Tung","year":"2025","unstructured":"Tung JYM, Le Q, Yao J et al (2025) Performance of retrieval-augmented generation large language models in guideline-concordant prostate-specific antigen testing: comparative study with junior clinicians. J Med Internet Res 27:e78393\u2013e78393. https:\/\/doi.org\/10.2196\/78393","journal-title":"J Med Internet Res"},{"issue":"3","key":"8498_CR33","doi-asserted-by":"publisher","first-page":"100943","DOI":"10.1016\/j.patter.2024.100943","volume":"5","author":"V Li\u00e9vin","year":"2024","unstructured":"Li\u00e9vin V, Hother CE, Motzfeldt AG, Winther O (2024) Can large language models reason about medical questions? Patterns 5(3):100943\u2013100943.\u00a0https:\/\/doi.org\/10.1016\/j.patter.2024.100943","journal-title":"Patterns"},{"key":"8498_CR34","unstructured":"DeepL SE DeepL Translator: DeepL SE. https:\/\/www.deepl.com\/de\/translator"}],"container-title":["Archives of Gynecology and Obstetrics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00404-026-08498-w","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00404-026-08498-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00404-026-08498-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,18]],"date-time":"2026-06-18T13:14:57Z","timestamp":1781788497000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00404-026-08498-w"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,18]]},"references-count":34,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["8498"],"URL":"https:\/\/doi.org\/10.1007\/s00404-026-08498-w","relation":{},"ISSN":["1432-0711"],"issn-type":[{"value":"1432-0711","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,18]]},"assertion":[{"value":"21 April 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"8 June 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"18 June 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"This study protocol was reviewed and approved by the ethics committee of the Hannover Medical School, approval number 10903. Individual patient informed consent was waived due to complete data anonymization.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval"}},{"value":"Statement: DeepL [\n                      \n                      ] was used for the translation of the LLM answers from German to English and for wording of the manuscript itself. After using this service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declaration of generative AI and AI-assisted technologies in the writing process"}}],"article-number":"206"}}