文献速递重症综合管理与镇静镇痛

使用词嵌入神经网络模型测量ICU记录中的内隐偏见

Measuring Implicit Bias in ICU Notes Using Word-Embedding Neural Network Models.
PubMed 文献 · Chest · 2024-01-08
一句话结论ICU记录中可识别内隐偏见,不同种族和民族描述词与污名化词语的上下文关系因记录时间和地点而异。

该研究利用词嵌入算法测量ICU临床记录中种族/民族描述词与污名化词语(暴力、被动、不依从等)的上下文相似度。数据来自UCSF(2012-2022年)和BIDMC(2001-2012年)的危重成人记录。在UCSF记录中,黑人描述词与暴力词语的上下文相似度低于白人描述词;而在BIDMC记录中,黑人描述词与暴力词语的上下文相似度高于白人描述词。UCSF数据还显示,黑人描述词与被动和不依从词语的上下文相似度高于拉丁裔描述词。作者认为内隐偏见在ICU记录中可识别,NLP模型可能传递训练数据中的偏见,在临床预测中使用NLP算法可能加剧差异,使用语言模型时可能需要主动去偏策略以实现算法公平。

要点

  • 在UCSF记录中,黑人描述词与暴力词语的上下文相似度低于白人描述词。
  • 在BIDMC记录中,黑人描述词与暴力词语的上下文相似度高于白人描述词。
  • UCSF数据中,黑人描述词与被动和不依从词语的上下文相似度高于拉丁裔描述词。
  • 不同种族和民族描述词与污名化词语的上下文关系因记录时间和地点而异。
  • 作者认为NLP模型可能传递训练数据中的内隐偏见,在临床预测中使用NLP算法可能加剧差异。

适用人群与场景

  • UCSF:2012年至2022年收治的危重成人患者的记录。
  • BIDMC:2001年至2012年收治的危重成人患者的记录。

研究设计与关键结果

  • 研究类型:未明确说明,使用无监督词嵌入算法测量种族或民族描述词与污名化目标词或词组(暴力、被动、不依从、不坚持)之间的余弦相似度。
  • UCSF记录中,黑人描述词与暴力词语的上下文相似度低于白人描述词。
  • BIDMC记录中,黑人描述词与暴力词语的上下文相似度高于白人描述词。
  • UCSF数据集中,黑人描述词与被动和不依从词语的上下文相似度高于拉丁裔描述词。

证据类型、等级与局限

  • 证据类型:未明确说明。
  • 局限:原文未报告局限性。

对床旁实践的改变

  • 作者认为,内隐偏见在ICU记录中可识别。
  • 作者认为,NLP模型似乎能够传递训练数据中的内隐偏见,在临床预测中使用NLP算法可能加剧差异。
  • 作者认为,在临床研究中使用语言模型时,可能需要主动去偏策略以实现算法公平。

提炼说明

  • 本条依据论文摘要整理;摘要未涉及「诊断标准与关键阈值」「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开

BACKGROUND: Language in nonmedical data sets is known to transmit human-like biases when used in natural language processing (NLP) algorithms that can reinforce disparities. It is unclear if NLP algorithms of medical notes could lead to similar transmissions of biases.

RESEARCH QUESTION: Can we identify implicit bias in clinical notes, and are biases stable across time and geography?

STUDY DESIGN AND METHODS: To determine whether different racial and ethnic descriptors are similar contextually to stigmatizing language in ICU notes and whether these relationships are stable across time and geography, we identified notes on critically ill adults admitted to the University of California, San Francisco (UCSF), from 2012 through 2022 and to Beth Israel Deaconess Hospital (BIDMC) from 2001 through 2012. Because word meaning is derived largely from context, we trained unsupervised word-embedding algorithms to measure the similarity (cosine similarity) quantitatively of the context between a racial or ethnic descriptor (eg, African-American) and a stigmatizing target word (eg, nonco-operative) or group of words (violence, passivity, noncompliance, nonadherence).

RESULTS: In UCSF notes, Black descriptors were less likely to be similar contextually to violent words compared with White descriptors. Contrastingly, in BIDMC notes, Black descriptors were more likely to be similar contextually to violent words compared with White descriptors. The UCSF data set also showed that Black descriptors were more similar contextually to passivity and noncompliance words compared with Latinx descriptors.

INTERPRETATION: Implicit bias is identifiable in ICU notes. Racial and ethnic group descriptors carry different contextual relationships to stigmatizing words, depending on when and where notes were written. Because NLP models seem able to transmit implicit bias from training data, use of NLP algorithms in clinical prediction could reinforce disparities. Active debiasing strategies may be necessary to achieve algorithmic fairness when using language models in clinical research.

来源:Chest,查看原文。版权归原作者所有。

参考来源

  1. Cobert J, Mills H, Lee A, et al. Measuring Implicit Bias in ICU Notes Using Word-Embedding Neural Network Models. Chest. 2024. PMID: 38199323. doi:10.1016/j.chest.2023.12.031

供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。