耳鼻喉科Pubmed文献追踪每日 14:00 同步
← 返回全部文献
Article record

评估人工智能生成的前庭神经鞘瘤患者咨询的准确性

Evaluating the accuracy of artificial intelligence-generated patient counseling for vestibular schwannoma.

AI/ML耳科IF 1.8Q2

文献信息

PMID42805092
原文在 PubMed 查看原文 ↗
发表日期2026
作者John M Sommerfeldt
查看其他 4 位作者

James J Pierre；Jakob L Fischer；Charles A Riley；Anthony M Tolisano

作者单位Department of Otolaryngology-Head and Neck Surgery, Tripler Army Medical Center, Honolulu, HI, USA.
查看其他 4 条单位信息

Department of Otolaryngology-Head and Neck Surgery, Walter Reed National Military Medical Center, Bethesda, MD, USA.；Department of Otolaryngology-Head and Neck Surgery, San Antonio Uniformed Services Health Education Consortium, San Antonio, TX, USA; Department of Surgery, Uniformed Services University of the Health Sciences, Bethesda, MD, USA.；Veteran Affairs (VA) Boston Healthcare System, Boston, MA, USA; Department of Otolaryngology-Head and Neck Surgery, Boston University School of Medicine, Boston, MA, USA.；Department of Otolaryngology-Head and Neck Surgery, Walter Reed National Military Medical Center, Bethesda, MD, USA; Department of Surgery, Uniformed Services University of the Health Sciences, Bethesda, MD, USA. Electronic address: Anthony.Tolisano@usuhs.edu.

期刊American journal of otolaryngology
SCI 分区Q2
IF1.8
研究类型AI/ML · 临床
所属专科耳科

中文摘要

目的： 比较Google、ChatGPT和医疗服务提供者在回答前庭神经鞘瘤常见问题方面的质量。
方法： 使用搜索引擎频率数据生成十个关于前庭神经鞘瘤病因、诊断和管理的常见患者问题。针对这些问题，从汇编的Google搜索结果、ChatGPT-4和提供者回答中生成答案。由三位盲法、独立、委员会认证的耳鼻喉科医生使用DISCERN工具对回答质量进行评分。计算可读性分数。使用Fleiss kappa确定评分者间一致性。
结果： 评分者间一致性高，Fleiss kappa评分为0.73。Google、ChatGPT和专家回答的总DISCERN评分分别为23.7、30.4和28.4，标准误为0.85。ChatGPT回答的平均字数（214 ± 98.2）显著长于Google（68.8 ± 72.3）和提供者回答（108 ± 91.7）（p < 0.05）。ChatGPT的可读性评分最高，Flesch阅读易读性评分为33.1 ± 10.9，其次是Google（30.3 ± 20.7）和提供者回答（15.8 ± 8.70），ChatGPT和Google的表现显著优于提供者（p < 0.05）。
结论： ChatGPT对前庭神经鞘瘤常见问题的回答总体质量与提供者相当。此外，与提供者和Google搜索衍生信息相比，这些回答被计算为具有最高的可读性水平。
证据等级： III。

英文摘要

OBJECTIVE: To compare the quality of Google, ChatGPT, and providers in answering common questions regarding vestibular schwannoma.
METHODS: Search engine frequency data was used to generate ten common patient questions regarding the etiology, diagnosis, and management of vestibular schwannoma. Answers to these questions were generated from compiled Google search, ChatGPT-4, and provider responses. Response quality was graded by three blinded, independent, board-certified otolaryngologists using the DISCERN instrument. Readability scores were calculated. Inter-rater agreement was determined using Fleiss's kappa.
RESULTS: Inter-rater agreement was high with a Fleiss's kappa score of 0.73. Total DISCERN scores among Google, ChatGPT, and expert response were 23.7, 30.4, and 28.4, respectively, with a standard error of 0.85. Average word count for ChatGPT (214 ± 98.2) responses were significantly longer (p < 0.05) than both Google (68.8 ± 72.3) and provider responses (108 ± 91.7) responses. ChatGPT had the highest readability score with a Flesch Reading Ease score of 33.1 ± 10.9, followed by Google (30.3 ± 20.7) and provider response (15.8 ± 8.70), with ChatGPT and Google performing significantly better than providers (p < 0.05).
CONCLUSION: The overall quality of ChatGPT responses to common questions about vestibular schwannoma was comparable to that of providers. Moreover, these responses were calculated to represent the highest readability levels when compared to both providers and Google search derived information.
LEVEL OF EVIDENCE: III.