文献速递前列腺癌

开发并验证自然语言处理系统评估前列腺癌会诊中医生沟通质量

Development and validation of a natural language processing system to assess quality of physician communication in prostate cancer consultations.
PubMed 文献 · Prostate Cancer Prostatic Dis · 2025-08-21

本研究基于50份会诊转录文本开发自然语言处理(NLP)模型,识别前列腺癌共享决策(SDM)五个内容领域的句子,并在20份独立会诊中验证模型能否通过最高主题一致性的句子评估整体沟通质量。随机森林模型在内部验证中识别各主题句子的曲线下面积(AUC)为0.84至0.99,在独立会诊中前10句模型识别句子对沟通质量分级的准确率为80%至100%。

要点

  • NLP模型可准确识别前列腺癌会诊中SDM关键内容领域的句子。
  • 模型最高主题一致性句子可对整体沟通质量进行分级。
  • 为可扩展的风险沟通质量评估提供了基础。

核心结论

  • NLP模型能准确捕获前列腺癌会诊中的关键信息并对医生沟通质量进行分级,为可扩展的风险沟通质量评估奠定基础。

研究设计与样本

  • 使用50份会诊转录文本训练和验证NLP模型以识别与关键概念相关的句子,并在20份独立会诊转录中测试模型预测主题一致性最高的句子能否确定整体沟通质量。

主要发现

  • 开发数据集共28,927个句子,75%用于训练,25%用于内部验证。
  • 随机森林模型在内部验证集中识别主题一致句子的AUC为:肿瘤风险(TR)0.98、病理结果(PR)0.94、预期寿命(LE)0.89、癌症预后(CP)0.92、基线泌尿功能(UF)0.84、勃起功能(EF)0.96、勃起功能障碍(ED)0.98、尿失禁(UI)0.97、刺激性泌尿症状(LUTS)0.99。
  • 在20份独立会诊中,模型识别的前10个句子对整个会诊沟通质量分级的准确率分别为:TR 100%、PR 90%、LE 95%、CP 95%、UF 80%、EF 95%、ED 85%、UI 100%、LUTS 95%。

临床怎么用

  • 该NLP系统可用于向前列腺癌患者提供关键信息,并审计医生沟通质量,促进共享决策。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: AUA guidelines for shared decision making (SDM) in prostate cancer recommend discussion of five content areas in consultations: (1) cancer severity (tumor risk (TR), pathology results (PR)); (2) life expectancy (LE); (3) cancer prognosis (CP); (4) baseline urinary and erectile function (UF and EF); and (5) treatment side effects (erectile dysfunction (ED), urinary incontinence (UI), and irritative urinary symptoms (LUTS)). However, patient retention of information after the visit and inconsistent risk communication by physicians are barriers to informed SDM. We sought to develop natural language processing (NLP) models based on recorded consultations to provide key information to patients and audit quality of physician communication. METHODS: We used 50 consultation transcripts to train and validate NLP models to identify sentences related to key concepts. We then tested whether communication quality across entire consultations could be determined by sentences with the highest model-predicted topic concordance in 20 separate consultation transcripts. RESULTS: Our development dataset included 28,927 total sentences, with 75% reserved for training and 25% for internal validation. The Random Forest model had the highest accuracy for identifying topic-concordant sentences, with area under the curve 0.98, 0.94, 0.89, 0.92, 0.84, 0.96, 0.98, 0.97, and 0.99 for TR, PR, LE, CP, UF, EF, ED, UI, and LUTS compared with manual coding across all concepts in the internal validation dataset. In 20 separate consultations, the top 10 model-identified sentences correctly graded communication quality across entire consultations with accuracies of 100%, 90%, 95%, 95%, 80%, 95%, 85%, 100%, and 95% for TR, PR, LE, CP, UF, EF, ED, UI, and LUTS compared with manual coding, respectively. CONCLUSIONS: NLP models accurately capture key information and grade quality of physician communication in prostate cancer consultations, providing the foundation for scalable quality assessment of risk communication.

来源:Prostate Cancer Prostatic Dis,查看原文。版权归原作者所有。

引用

  1. Zheng R, Friedrich NA, Luu M, et al. Development and validation of a natural language processing system to assess quality of physician communication in prostate cancer consultations. Prostate Cancer Prostatic Dis. 2025. PMID: 40841832. doi:10.1038/s41391-025-01011-5

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。