Assessing effects of adverse Social Determinants of Health (SDOH) in TTR V122l carriers via Structured data and Natural Language Processing (NLP) extraction, a comparison

通过结构化数据和自然语言处理 (NLP) 提取评估 TTR V122l 携带者健康不良社会决定因素 (SDOH) 的影响，比较

• 批准号: 10830156
• 负责人: Ron Do
• 金额: $ 12.72万
• 依托单位: ICAHN SCHOOL OF MEDICINE AT MOUNT SINAI
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-02-15 至 2025-01-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10830156
• 关键词: Address; Clinical; Code; Complex; DNA Sequence Alteration; Data; Development; Disease Progression; Disease model; Exhibits; Future; Genes; Goals; Health; Medical; Methodology; Methods; Mutation; Natural Language Processing; Outcome; Patients; Penetrance; Process; Research Personnel; Research Project Grants; improved; innovation; patient health information; poor health outcome; social factors; social health determinants; structured data; tool

The research project has two goals. The first is to use natural language processing (NLP) to improve the identification of adverse social determinants of health (SDOH) in patients with a specific genetic mutation. This is important because many social factors are only captured in unstructured medical narratives, and NLP can help identify these factors more accurately than the current method of using specific codes. The second goal is to investigate whether adverse SDOH are associated with poor health outcomes in patients with this mutation. The researchers will look at patients' medical records to identify adverse SDOH and compare their health outcomes to those without adverse SDOH. The researchers hypothesize that adverse SDOH will be associated with worse outcomes, and that more adverse SDOH will be associated with even worse outcomes. However, the study has limitations, including relying on only one NLP tool and using a binary definition of adverse SDOH. Future studies may address these limitations by using different methodologies and more granular data.

该研究项目有两个目标。第一种是使用自然语言处理（NLP）， 改善对健康不良社会决定因素（SDOH）的识别， 特定的基因突变。这一点很重要，因为许多社会因素只在 非结构化的医疗叙述，NLP可以帮助更准确地识别这些因素， 使用特定代码的当前方法。第二个目标是调查是否不利 SDOH与携带该突变的患者的不良健康结局相关。研究人员 我将查看病人的医疗记录，以确定不利的SDOH和比较他们的健康 对那些没有不利的SDOH的结果。研究人员假设，不利的SDOH将 与更糟糕的结果相关，并且更不利的SDOH将与甚至 更糟糕的结果。然而，该研究存在局限性，包括仅依赖一种NLP工具， 使用不利SDOH的二元定义。未来的研究可能会解决这些限制， 使用不同的方法和更精细的数据。