{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T09:44:33Z","timestamp":1785750273013,"version":"3.56.0"},"reference-count":15,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T00:00:00Z","timestamp":1785715200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T00:00:00Z","timestamp":1785715200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100001779","name":"Monash University","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100001779","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Neurosurg Rev"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Patients increasingly consult large language models (LLMs) before specialist review and may do so in widely differing emotional registers. We examined whether re-casting an unruptured intracranial aneurysm (UIA) case from a third-person vignette into a first-person patient account, with anxiety about treatment or non-treatment, alters the recommendations LLMs return. Sixty-seven UIA cases previously discussed at our multidisciplinary team (MDT) were presented to Claude Opus 4.6, ChatGPT-5.4 and Gemini 3 Pro Thinking under four conditions: third-person vignette (3P), neutral first-person (1PN), and first-person accounts anxious about treatment (1PAT) or non-treatment (1PANT). Each case was submitted five times (4,020 prompts). Majority recommendations were benchmarked against MDT consensus (Cohen\u2019s \u03ba, McNemar), tested for directional asymmetry (Bowker), classified as progressive or regressive, and linguistically analysed. Concordance with MDT was fair-to-moderate (71.6\u201380.0%; \u03ba 0.34\u20130.51). ChatGPT over-treated under all conditions (p\u2009=\u20090.001\u20130.013) and Gemini over-treated only at 3P baseline (p\u2009=\u20090.001), abolished under first-person framing. Claude showed no directional preference. First-person framing drove significant clipping-to-coiling migration in Gemini and ChatGPT (p\u2009=\u20090.001 and 0.015). Claude alone hedged under treatment-directed anxiety (p\u2009=\u20090.002). Anxiety-induced shifts were predominantly regressive. Gemini under rupture-anxiety combined the highest empathy-opener rate (85%), highest confidence-marker density and lowest hedging density of any cell, a sycophantic phenotype invisible to categorical analysis. Patient persona and affective framing measurably and model-specifically alter LLM recommendations for UIAs, with anxiety-induced shifts moving models away from rather than toward expert consensus. Clinicians should anticipate AI-shaped expectations varying with a patient\u2019s emotional state and counsel against framing-induced advice. This study was retrospectively registered on the Open Science Framework (https:\/\/doi.org\/10.17605\/OSF.IO\/HCU6D).<\/jats:p>","DOI":"10.1007\/s10143-026-04419-2","type":"journal-article","created":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T08:47:50Z","timestamp":1785746870000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["The Influence of Patient Persona and Affective Framing on Management Recommendations of ChatGPT, Gemini and Claude for Unruptured Intracranial Aneurysms: A Comparative Benchmarking Study"],"prefix":"10.1007","volume":"49","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-5326-6949","authenticated-orcid":false,"given":"Anish","family":"Narayan","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Frederick","family":"Mariajoseph","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Malik","family":"Farooq","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Adrian","family":"Praeger","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Justin","family":"Moore","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,8,3]]},"reference":[{"key":"4419_CR1","doi-asserted-by":"publisher","DOI":"10.1002\/14651858.cd013312","author":"FG Pontes","year":"2019","unstructured":"Pontes FG, de Vasconcelos B, Baptista-Silva V, da Silva JC EM (2019) Treatments for unruptured intracranial aneurysms. Cochrane Database Syst Rev. https:\/\/doi.org\/10.1002\/14651858.cd013312","journal-title":"Cochrane Database Syst Rev"},{"issue":"2","key":"4419_CR2","doi-asserted-by":"publisher","first-page":"609","DOI":"10.1038\/s41591-025-04074-y","volume":"32","author":"AM Bean","year":"2026","unstructured":"Bean AM, Payne RE, Parsons G et al (2026) Reliability of LLMs as medical assistants for the general public: a randomized preregistered study. Nat Med 32(2):609\u2013615. https:\/\/doi.org\/10.1038\/s41591-025-04074-y","journal-title":"Nat Med"},{"key":"4419_CR3","doi-asserted-by":"publisher","unstructured":"Hadjiathanasiou A, Goelz L, Muhn F et al (2026) Artificial intelligence in neurovascular decision-making: a comparative analysis of ChatGPT-4 and multidisciplinary expert recommendations for unruptured intracranial aneurysms. Neurosurg Rev 48(1). https:\/\/doi.org\/10.1007\/s10143-025-03341-3","DOI":"10.1007\/s10143-025-03341-3"},{"key":"4419_CR4","doi-asserted-by":"publisher","unstructured":"Perez E, Ringer S, Luko\u0161i\u016bt\u0117 K et al (2022) Discovering language model behaviors with model-written evaluations. https:\/\/doi.org\/10.48550\/arXiv.2212.09251. arXiv","DOI":"10.48550\/arXiv.2212.09251"},{"key":"4419_CR5","doi-asserted-by":"publisher","unstructured":"Sharma M, Tong M, Korbak T et al (2023) Towards understanding sycophancy in language models. arXiv. https:\/\/doi.org\/10.48550\/arXiv.2310.13548","DOI":"10.48550\/arXiv.2310.13548"},{"issue":"39","key":"4419_CR6","doi-asserted-by":"publisher","first-page":"33566","DOI":"10.1609\/aaai.v40i39.40645","volume":"40","author":"K Wang","year":"2026","unstructured":"Wang K, Li J, Yang S, Zhang Z, Wang D (2026) When truth is overridden: uncovering the internal origins of sycophancy in large language models. Proc AAAI Conf Artif Intell 40(39):33566\u201333574. https:\/\/doi.org\/10.1609\/aaai.v40i39.40645","journal-title":"Proc AAAI Conf Artif Intell"},{"key":"4419_CR7","unstructured":"Anthropic (2025) Claude\u2019s constitution. https:\/\/www.anthropic.com\/constitution. Accessed 25 May 2026"},{"key":"4419_CR8","unstructured":"Cui J, Chiang WL, Stoica I, Hsieh CJ (2025) OR-Bench: an over-refusal benchmark for large language models. ICML. https:\/\/icml.cc\/virtual\/2025\/poster\/46052"},{"key":"4419_CR9","unstructured":"Bai Y, Jones A, Ndousse K et al (2022) Training a helpful and harmless assistant with reinforcement learning from human feedback. arXiv. https:\/\/arxiv.org\/abs\/2204.05862"},{"key":"4419_CR10","unstructured":"Eloundou T, Beutel A, Robinson D et al (2024) First-person fairness in chatbots. OpenAI. https:\/\/openai.com\/index\/evaluating-fairness-in-chatgpt"},{"key":"4419_CR11","doi-asserted-by":"crossref","unstructured":"Ouyang L, Wu J, Jiang X et al (2022) Training language models to follow instructions with human feedback. arXiv. https:\/\/arxiv.org\/abs\/2203.02155","DOI":"10.52202\/068431-2011"},{"issue":"1","key":"4419_CR12","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41746-025-02008-z","volume":"8","author":"S Chen","year":"2025","unstructured":"Chen S, Gao M, Sasse K et al (2025) When helpfulness backfires: LLMs and the risk of false medical information due to sycophantic behavior. NPJ Digit Med 8(1):1\u20139. https:\/\/doi.org\/10.1038\/s41746-025-02008-z","journal-title":"NPJ Digit Med"},{"key":"4419_CR13","doi-asserted-by":"publisher","unstructured":"Kim TM, Luo L, Kim SE, Manrai AK, Topol E, Rajpurkar P (2026) The doctor will agree with you now: sycophancy of large language models in multi-turn medical conversations. Proc 1st Workshop Linguist Anal Health (HeaLing) 19\u201334. https:\/\/doi.org\/10.18653\/v1\/2026.healing-1.2","DOI":"10.18653\/v1\/2026.healing-1.2"},{"issue":"3","key":"4419_CR14","doi-asserted-by":"publisher","first-page":"2077","DOI":"10.1007\/s10143-022-01768-6","volume":"45","author":"K Hannah","year":"2022","unstructured":"Hannah K, Silvestre J, Factor JV, Joy K (2022) A meta-analysis on the prevalence of anxiety and depression in patients with unruptured intracranial aneurysms: exposing critical treatment gaps. Neurosurg Rev 45(3):2077\u20132085. https:\/\/doi.org\/10.1007\/s10143-022-01768-6","journal-title":"Neurosurg Rev"},{"key":"4419_CR15","doi-asserted-by":"publisher","unstructured":"Sharma N, Liao QV, Xiao Z (2024) Generative echo chamber? Effect of LLM-powered search systems on diverse information seeking. https:\/\/doi.org\/10.1145\/3613904.3642459","DOI":"10.1145\/3613904.3642459"}],"container-title":["Neurosurgical Review"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10143-026-04419-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10143-026-04419-2","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10143-026-04419-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T08:47:52Z","timestamp":1785746872000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10143-026-04419-2"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,3]]},"references-count":15,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["4419"],"URL":"https:\/\/doi.org\/10.1007\/s10143-026-04419-2","relation":{},"ISSN":["1437-2320"],"issn-type":[{"value":"1437-2320","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,3]]},"assertion":[{"value":"5 June 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 June 2026","order":2,"name":"revised","label":"Revised","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"23 July 2026","order":3,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"3 August 2026","order":4,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"This study was approved by the institution's ethics review board. Approval was granted by the institutional review board (reference: RES-25-0000-885Q).","order":1,"name":"Ethics","label":"Ethics approval","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Individual patient consent was waived by the institutional research ethics committee owing to the retrospective, fully de-identified nature of the data. No identifiable patient information was submitted to any LLM at any point.","order":2,"name":"Ethics","label":"Consent to participate","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable. The manuscript does not contain any individual person\u2019s identifying data, images, or videos.","order":3,"name":"Ethics","label":"Consent to publish","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Large language models (Claude Opus 4.6, ChatGPT-5.4, Gemini 3 Pro Thinking) were the objects of evaluation in this study; they did not contribute to authorship.","order":4,"name":"Ethics","label":"Use of large language models","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":5,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"508"}}