Using clinician text notes in electronic medical record data to validate transgender-related diagnosis codes
Using clinician text notes in electronic medical record data to validate transgender-related diagnosis codes
复制标题
DOI：
10.1093/jamia/ocy022
复制
发表时间：
2018-07-01
影响因子：
6.4
通讯作者：
Fine, Michael J.
中科院分区：
文献类型：
--
作者：
Blosnich, John R.;Cashy, John;Fine, Michael J.
目标：跨性别者容易受到负面健康风险和结果的影响，但研究仍然有限，因为电子病历 (EMR) 等数据源缺乏性别认同信息的标准化收集。大多数 EMR 不包括自我认定性别认同的黄金标准，但国际疾病分类 (ICD) 包括指示跨性别相关临床服务的诊断代码。然而，尚不清楚这些代码是否可以表明跨性别身份。本研究的目的是确定 EMR 中患者的临床医生笔记包含可证实 ICD 编码的跨性别身份的跨性别相关术语的程度。方法：数据来自美国退伍军人事务部企业数据仓库。跨性别患者的定义是存在与跨性别相关的临床服务相关的 ICD9 和 ICD10 代码，并按 3:1 比例划分了非跨性别患者作为对照组。提取患者的临床医生文本笔记并搜索与跨性别相关的单词和短语。结果：在根据 ICD 代码定义为跨性别的 7560 名患者中，搜索算法识别出 6753 名（89.3%）具有跨性别相关术语。在 22072 名没有 ICD 代码的非跨性别患者中，246 名（1.1%）有跨性别相关术语；经过审查，11 名患者被确定为跨性别者，表明假阴性率为 0.05%。结论：当 EMR 中无法提供自我识别的性别认同数据时，使用 ICD 定义的跨性别身份可以促进卫生服务研究。
Objective: Transgender individuals are vulnerable to negative health risks and outcomes, but research remains limited because data sources, such as electronic medical records (EMRs), lack standardized collection of gender identity information. Most EMR do not include the gold standard of self-identified gender identity, but International Classification of Diseases (ICDs) includes diagnostic codes indicating transgender-related clinical services. However, it is unclear if these codes can indicate transgender status. The objective of this study was to determine the extent to which patients' clinician notes in EMR contained transgender-related terms that could corroborate ICD-coded transgender identity.Methods: Data are from the US Department of Veterans Affairs Corporate Data Warehouse. Transgender patients were defined by the presence of ICD9 and ICD10 codes associated with transgender-related clinical services, and a 3:1 comparison group of nontransgender patients was drawn. Patients' clinician text notes were extracted and searched for transgender-related words and phrases.Results: Among 7560 patients defined as transgender based on ICD codes, the search algorithm identified 6753 (89.3%) with transgender-related terms. Among 22 072 patients defined as nontransgender without ICD codes, 246 (1.1%) had transgender-related terms; after review, 11 patients were identified as transgender, suggesting a 0.05% false negative rate.Conclusions: Using ICD-defined transgender status can facilitate health services research when self-identified gender identity data are not available in EMR.