Using Behavioral Economics and Implementation Science to Advance the Use of Genomic Medicine Utilizing an EHR Infrastructure across a Diverse Health System

基因检测结果改变医疗管理的医疗条件的数量 患者数量呈指数级增长。然而，少数符合条件的患者接受了基因检测，尽管 对下游护理的影响。系统-(识别符合条件的患者并返回结果的方法)、临床医生( 例如，知识、有限的劳动力)和患者水平(例如，对成本和不利影响的担忧) 壁垒助长了不确定性和依赖现状的倾向--未能使用基因组信息来指导 医疗护理。实施科学方法和框架是解决这一实践差距的理想选择， 特别是那些考虑多层次障碍和人类决策在以下背景下的作用的国家 不确定性。我们的团队已经建立了基础设施，以解决提供基因检测的系统障碍 整个医疗系统-电子健康记录(EHR)内的集成系统，可直接 基因测试的排序和结果作为结构化数据--现在有多个传播请求。我们的 团队还使用行为经济学作为实施科学框架，通过以下方式改善医疗保健 使用Nudge(电子病历默认设置、患者启动)同时克服临床医生和患者的障碍 解决健康差距(例如，少数族裔之间存在更大的实践差距)。合并这些领域，我们 提出一个高度创新的项目，该项目将首次对临床医生(EHR)使用轻推进行评估 默认情况：1)参考遗传学诊所或2)订购基因检测)和/或轻推患者 (在预约前就基因检测的好处向主要患者进行沟通)。在目标1中，我们将 开发针对10种临床疾病的电子表型算法，这将推动特定诊断的遗传学 推荐和测试；我们将与利益相关者咨询委员会合作完善我们的微调。在目标2中，我们将 对228名临床医生进行混合类型3实施研究，采用整群随机设计 (医生，高级执业医师)作为随机单位(N=120个组)和16,500名患者 用10种条件中的一种来检查对基因检测速度的影响：患者启动轻推， 两个临床医生轻推，结合患者和每个临床医生轻推，而不是通用的最佳实践 警报(BPA)(无临床医生或患者轻推)。我们将检查患者(例如种族)、临床医生(例如专业)和 系统(例如，社区与学术中心)调节对基因检测率和评估的微调影响 有效结果(临床医生在确定致病变种后采取行动的比率)。在目标3中，我们 将采用系统的方法传播我们的eHR与基因检测的集成，以eHR为基础 算法，以及通过Epic，PheKB，NHGRl的铁锤为临床试验构建的其他材料和系统， 和GitHub。我们的研究将产生巨大的影响，因为它将产生一个新颖、有效和可转让的基于 能够持续提供基因组药物的基础设施，极大地推动了该领域的发展。

The number of medical conditions for which the results of genetic testing change the medical management of patients is exponentially increasing. However, a minority of eligible patients receive genetic testing, despite the implications for downstream care. System- (methods to identify eligible patients and return results), clinician( e.g., knowledge, limited workforce), and patient- (e.g. , concerns about costs and adverse effects) level barriers foster uncertainty and a tendency to rely on the status quo - failing to use genomic information to guide medical care. Implementation science methods and frameworks are ideal for addressing this practice gap, especially those that consider multi-level barriers and the role of human decision-making in contexts with uncertainty. Our team has built the infrastructure to address system-barriers to delivering genetic testing across our health system - an integrated system within the electronic health record (EHR) that enables direct ordering and resulting of genetic tests as structured data - now with multiple requests for dissemination. Our team also is using behavioral economics as an implementation science framework to improve healthcare by using nudges (EHR defaults, patient priming) to overcome clinician and patient barriers, concurrently addressing health disparities (e.g., higher practice gaps among racial minorities). Merging these areas, we propose a highly innovative project that will evaluate, for the first time, the use of nudges to clinicians (EHR defaults for either: 1) referring to genetics clinic or 2) ordering for genetic testing) and/or nudges to patients (communication to prime patients about the benefits of genetic testing prior to appointment). In Aim 1, we will develop electronic phenotyping algorithms for 10 clinical conditions, which will drive diagnosis-specific genetics referral and testing; we will refine our nudges working with a Stakeholder Advisory Council. In Aim 2, we will conduct a hybrid type 3 implementation study, using a cluster randomized design with 228 clinicians (physician, Advanced Practice Practitioners) as the unit of randomization (N= 120 clusters) and 16,500 patients with one of the 10 conditions to examine the impact on the rate of genetic testing of: the patient priming nudge, the two clinician nudges, combining the patient and each of the clinician nudges, vs. a generic best practice alert (BPA) (no clinician or patient nudge). We will examine patient (e.g., race), clinician (e.g., specialty), and system (e.g. , community vs. academic center) moderators of nudge effects on genetic testing rate and assess an effectiveness outcome (rate of clinician action following identification of a pathogenic variant). In Aim 3, we will engage in systematic methods to disseminate our EHR integration of genetic testing, EHR-based algorithms, and other materials and systems built for the clinical trial through Epic, PheKB, NHGRl's AnVIL, and GitHub. Our study will be immensely impactful, as it will yield a novel, effective, and transferrable EHRbased infrastructure that enables the sustainable delivery of genomic medicine, greatly advancing the field.