Integrating Multiple Electronic Health Records Systems to Improve Lung Cancer Outcomes

整合多个电子健康记录系统以改善肺癌结果

• 批准号: 10718073
• 负责人: Summer S Han
• 金额: $ 65.89万
• 依托单位: STANFORD UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-07-12 至 2028-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10718073
• 关键词: Address; Artificial Intelligence; Asian; Behavioral; Biometry; California; Cancer Etiology; Cancer Patient; Cancer Survivor; Cessation of life; Clinical; Clinical Informatics; Clinical Trials; Communities; Consensus; Cost Savings; Data; Data Sources; Databases; Diagnosis; Electronic Health Record; Eligibility Determination; Epidemiology; Ethnic Origin; General Population; Goals; Guidelines; Health; Healthcare; Healthcare Systems; Immunologic Markers; Incidence; Individual; Link; Longterm Follow-up; Lung; Malignant neoplasm of lung; Measures; Methods; Modeling; Morbidity - disease rate; Oncology; Outcome; Patients; Pattern; Population; Population Heterogeneity; Public Health; Race; Radiology Specialty; Recording of previous events; Records; Registries; Research; Research Personnel; Risk; Risk Factors; Smoke; Smoker; Smoking; Smoking Status; Statistical Methods; Survivors; Thoracic Surgical Procedures; Time; Translational Research; United States Preventative Services Task Force; Universities; cancer diagnosis; cancer risk; cancer survival; computed tomography screening; electronic health record system; epidemiologic data; ethnic diversity; high risk; improved; interest; lung cancer screening; mortality; neoplasm registry; neutrophil; never smoker; never smoking; novel; population based; predictive modeling; randomized trial; risk prediction; screening; screening guidelines; sex; shared database; smoking cessation; standard of care; targeted treatment; tool; translational impact

Recent advances in screening and treatment have increased the number of lung cancer (LC) survivors (~571,340 LC survivors as of 2019). However, studies have shown that these survivors have a high risk for developing second primary lung cancer (SPLC), with the median 10-year SPLC risk of 8.36% after surviving 5 years from the initial diagnosis. Further, survivors with SPLC have significantly higher mortality vs. those who remain with single primary LC. Many unaddressed challenges exist: (1) While prior studies identified several risk factors for SPLC, these are mostly measured and fixed at the time of initial diagnosis, with findings focused on survivors who have ever smoked. However, SPLC risk is likely to be influenced by dynamic changes of various factors (e.g., smoking cessation), and our preliminary data show that SPLC risk remains just as high among survivors who never smoked. (2) Nevertheless, current epidemiologic data mainly used for SPLC do not offer detailed data measured after initial diagnosis, (3) nor have risk factors or predictions been evaluated for never-smoking survivors. (4) Further, limited trial evidence exists to address the important clinical question of whether and how to continue CT screening after IPLC diagnosis among LC survivors, which requires a long-term follow-up that is often not feasible in clinical trials. (5) Importantly, data on detailed screening for LC survivors are typically lacking in most population-based data. We plan to address these multiple challenges by leveraging electronic health records (EHRs) and novel analytical approaches to generate evidence to inform clinical decisions. Our long-term goal is to improve LC outcomes by focusing on SPLC utilizing large EHR data combined with novel statistical methods that integrate patient data measured after initial diagnosis. Our Specific Aims are: (AIM 1) to use a novel 3-way linkage to establish an integrated shared database for LC (i.e., Oncoshare-Lung) using EHRs from community-based and academic healthcare systems (with an ethnically diverse population with a high proportion of Asian never smokers) linked to the California Cancer Registry (CCR) ; (AIM 2) to provide a set of clinical decision tools for efficiently managing LC survivors by developing a novel statistical framework for predicting dynamic SPLC risks by capturing data measured after IPLC diagnosis; and (AIM 3) to evaluate the feasibility and utility of a novel causal inference method to assess efficient screening strategies for SPLC in LC survivors using EHRs. We will apply a new causal inference method that explicitly emulates the target trials (hypothetical randomized trials to answer the question of interest) in estimating the effects of continuing CT screening in long- term LC survivors under varying eligibility criteria. We expect that the completion of this research will fill the critical gaps in SPLC by providing: (1) clinical decision tools to assess individuals’ dynamic SPLC risks to identify high-risk survivors for tailored surveillance, (2) new analytic pipelines to evaluate efficient screening criteria for SPLC, and (3) a well-curated database for high-impact translational research for LC outcomes and surveillance in an ethnically diverse population that provides a unique opportunity to examine critical questions in SPLC.

筛查和治疗方面的最新进展增加了肺癌 (LC) 幸存者的数量（约 571,340 人） 截至 2019 年的 LC 幸存者）然而，研究表明这些幸存者罹患癌症的风险很高。 第二原发性肺癌 (SPLC)，生存 5 年后，中位 10 年 SPLC 风险为 8.36% 此外，SPLC 幸存者的死亡率明显高于那些仍患有 SPLC 的患者。 单一原发 LC 存在许多未解决的挑战：(1) 虽然先前的研究确定了一些危险因素。 SPLC，这些大多在初次诊断时进行测量和固定，结果集中于幸存者 然而，SPLC 风险很可能受到各种因素动态变化的影响。 （例如，戒烟），我们的初步数据显示幸存者中 SPLC 风险仍然很高 (2) 然而，目前主要用于 SPLC 的流行病学数据并没有提供详细数据。 在初步诊断后进行测量，(3) 也没有评估从不吸烟的危险因素或预测 (4) 此外，有限的试验证据无法解决是否以及如何解决这一重要的临床问题。 在 LC 幸存者中进行 IPLC 诊断后继续进行 CT 筛查，这需要长期随访 (5) 重要的是，通常缺乏对 LC 幸存者进行详细筛查的数据 我们计划利用电子健康来应对这些多重挑战。 记录（EHR）和新颖的分析方法来生成证据来为我们的长期决策提供信息。 目标是通过专注于 SPLC、利用大量 EHR 数据与新颖的统计数据相结合来改善 LC 结果 整合初步诊断后测量的患者数据的方法我们的具体目标是：（目标 1）使用 新颖的三向链接，使用 EHR 建立 LC 集成共享数据库（即 Oncoshare-Lung） 以社区和学术为基础的医疗保健系统（人口种族多样化，比例很高） 亚洲从不吸烟者）与加州癌症登记处 (CCR) (AIM 2) 关联，提供一套临床数据； 通过开发新的预测统计框架来有效管理 LC 幸存者的决策工具 通过捕获 IPLC 诊断后测量的数据来评估动态 SPLC 风险；以及（AIM 3）评估可行性 以及使用一种新颖的因果推理方法来评估 LC 幸存者 SPLC 的有效筛查策略 我们将应用一种新的因果推理方法来明确模拟目标试验（假设） 随机试验来回答感兴趣的问题）估计长期 CT 筛查的效果 我们预计这项研究的完成将填补不同资格标准下的长期 LC 幸存者。 通过提供以下方法来弥补 SPLC 的关键差距：(1) 临床决策工具，用于评估个人的动态 SPLC 风险，以识别 高风险幸存者进行量身定制的监测，（2）新的分析管道来评估有效的筛查标准 SPLC，以及 (3) 一个精心策划的数据库，用于 LC 结果和监测的高影响力转化研究 在一个种族多元化的人群中，这提供了一个独特的机会来研究 SPLC 中的关键问题。