Integrating EHR and patient-generated health data for breast cancer risk assessment and decision support in a diverse multiethnic population

虽然乳腺癌(BC)的死亡率已经下降，但这种下降已经开始停滞不前，特别是在 少数族裔。被确定为BC高危女性可能受益于化学预防、BC检测 易感基因、筛查和其他个性化的风险降低策略；然而，存在的障碍包括 对人群中的每个妇女进行风险评估所需的时间。电子健康记录(EHR)， 填充风险评估模型的常见来源存在挑战，包括丢失数据和数据 与EHR相比，由患者提供的类型更准确。我们之前提取了EHR的年龄数据， 种族/民族、不列颠哥伦比亚省家族史、良性乳腺疾病和乳房密度，以根据 对9,514名女性进行了乳腺癌监测联盟(BCSC)模式的调查。比较自我报告和 EHR数据显示，有BC一级家族史(14.6%比4.4%)和良性乳腺活检的女性更多 (21.3%比11.3%)与患者报告的数据相一致，但EHR数据发现更多的女性患有非典型性 或小叶原位癌(1.1%vs.2.3%)。EHR有26.8%的女性缺少种族/民族数据 不列颠哥伦比亚省一级家族史占87.2%。恰逢其时的快速医疗互操作性资源 (FHIR)、应用程序编程接口(API)和新的法规提供了完美的自动化解决方案 BC风险评估，集成患者生成的健康数据和EHR数据，以利用优势 每一种方法。在之前的工作中，我们使用迭代设计过程开发了RealRisks决策辅助工具 公平地最大化可接受性和可用性。RealRisks促进对BC风险的了解并收集 患者输入数据以根据Gail模型、BCSC和BRCAPRO计算BC风险。当冷杉时 之后，我们更新了RealRisks以自动填充BC风险计算信息 EHR，并设计了一个原型界面，将这些数据显示给患者，并请求查看和修改 在运行风险评估之前的数据。我们最近进行了一项可行性研究，以证明电子病历 来自fhir的数据可以合并到自动的BC风险计算中。增加……的可能性 制定可传播和公平的战略，将EHR和PGDH数据整合起来，用于风险评估和 个性化的BC风险降低，R21的重点是在不同的多民族中改进和测试我们的方法 女人。我们的目标是：1)进行用户评估，以完善FHIR增强的RealRisks；2)评估 在一项多种族先导研究中，FHIR增强的RealRisks对患者的激活度、风险感知和可用性的影响 高危妇女；以及3)确定将FHIR增强的RealRisks应用于临床护理的多级障碍。 鉴于与不列颠哥伦比亚省的致命关联，需要集中努力提供准确的风险评估和 关于降低风险战略的共同决策，特别是在少数族裔妇女中，她们更有可能 诊断为BC晚期。如果成功，在此应用程序中测试的方法可能会提供路线图 广泛改善对健康数据的数字化获取，并以公平的方式降低不孕症死亡率。

While breast cancer (BC) mortality has declined, this decline has begun to plateau, particularly among racial/ethnic minorities. Women identified as high-risk for BC may benefit from chemoprevention, testing for BC susceptibility genes, screening, and other personalized risk-reducing strategies; however, barriers exist including the time required to conduct risk assessment of each woman in a population. Electronic health records (EHRs), a common source for populating risk assessment models present challenges, including missing data, and data type more accurate when provided by patients compared to EHRs. We previously extracted EHR data on age, race/ethnicity, family history of BC, benign breast disease, and breast density to calculate BC risk according to the Breast Cancer Surveillance Consortium (BCSC) model among 9,514 women. Comparing self-reported and EHR data, more women with a first-degree family history of BC (14.6% vs. 4.4%) and benign breast biopsies (21.3% vs. 11.3%) were identified with patient-reported data, but EHR data identified more women with atypia or lobular carcinoma in situ (1.1% vs. 2.3%). The EHR had missing data on race/ethnicity for 26.8% of women and on first-degree family history of BC for 87.2%. Opportunely, Fast Healthcare Interoperability Resources (FHIR), application programming interfaces (APIs), and new legislation offer an elegant solution for automated BC risk assessment that integrates both patient-generated health data and EHR data to harness the strengths of each approach. In prior work, we developed the RealRisks decision aid using an iterative design process to equitably maximize acceptability, and usability. RealRisks promotes understanding of BC risk and collects patient-entered data to calculate BC risk according to the Gail model, BCSC, and BRCAPRO. When FHIR became available, we updated RealRisks to automatically populate information for BC risk calculation from the EHR, and designed a prototype interface that shows this data to patients with a request to review and modify data before running the risk assessments. We recently conducted a feasibility study to demonstrate that EHR data from FHIR could be incorporated into automated BC risk calculation. To increase the likelihood of developing disseminatable and equitable strategies that integrate EHR and PGDH data for risk assessment and personalized BC risk-reduction, the focus of this R21 is to refine and test our approach among diverse multiethnic women. Our aims are: 1) conduct user evaluations to refine FHIR-enhanced RealRisks; 2) assess the effect of the FHIR-enhanced RealRisks on patient activation, risk perception, and usability in a pilot study of multiethnic high-risk women; and 3) identify multilevel barriers to implementing FHIR-enhanced RealRisks into clinical care. Given the mortally associated with BC, focused efforts are needed to provide accurate risk assessment and shared decision-making about risk-reducing strategies, especially in minority women who are more likely to be diagnosed with advanced stage BC. If successful, the approach tested in this application may provide a roadmap for broadly improving digital access to health data and reducing BC mortality in an equitable manner.