Inform Shared Decision-making with Advanced Bayesian Causal Inference to Improve Quality of Pediatric Rheumatology Care

项目摘要/摘要 当面对治疗青少年特发性关节炎(JIA)的儿童时，一种异质的慢性疾病 目前还没有已知的病因或治疗方法，也没有万能的解决方案。临床医生依靠他们过去的经验和 向同行和科学界学习，为JIA患者做出最佳决定。然而，尽管 FDA批准的多种治疗方法，只有40%的多关节形式的JIA患者达到了 在可控制的疾病中，它很难治疗。虽然临床专业知识是无价的，但依赖于临床医生的经验 在治疗不常见的疾病(1000名儿童中有1名)时，在护理点(POC)单独使用 不同的表现和不同的治疗反应可能导致错过 早期缓解的机会和不太理想的结果。想象一下数字医疗技术的力量 (DHT)，它使用真实世界的数据来告知每个儿童不同治疗策略的有效性。通过 从提供者和患者网络的过去经验中集体学习，这些网络利用共享的 临床数据登记，分布式哈希表合成和更新以POC患者为中心的知识，同时 考虑患者对共享决策(SDM)的价值观和偏好。目前，没有 这样的分布式哈希德疗法可在POC提供以患者为中心的比较有效性证据。 这项研究旨在通过新颖的以用户为中心的设计来弥合这一差距，以适应PoC和 现有研究人员面向应用包应用基于改进贝叶斯因果学习方法 到真实世界的电子健康记录(EHR)数据。这种分布式哈希表被称为PCATS.JIA(以患者为中心的自适应 将在一个由23个中心组成的学习保健网络的背景下实施JIA护理的治疗战略。 在R21阶段，研究的目标是：1)通过开发PCATS.JIA分布式哈希表，将PCATS平台引入POC 2)共同设计了PCATS.JIA的图形用户界面，作为以患者为中心的SDM 与关键利益相关者(患者、家长和临床医生)合作的工具；以及3)试行测试PCATS.JIA作为SDM 在一次风湿病诊所使用POC的工具。在R33阶段，我们将细化、测试和评估PCATS.JIA 在三个儿科风湿病诊所通过1)使用质量改进科学和实施科学 将PCATS.JIA高可靠性地实施到临床工作流程中的原则；以及2)演示使用 PCATS的实施将导致医疗服务的改善，体现在更多以患者为中心的护理上，改善 健康结果和健康公平。 这项研究承诺提供一种新的以患者为中心的分布式哈希表，它健壮、灵活并提高了质量 以及在儿科风湿病实践中检验的护理和治疗结果的公平性。成为EHR 系统激励性，它可以扩展到不同的临床中心。这种分布式哈希表有可能推广到 更广泛的疾病状况和患者群体。

PROJECT SUMMARY/ABSTRACT When faced with treating a child with Juvenile Idiopathic Arthritis (JIA), a heterogeneous chronic condition with no known etiology or cure, there is no one-size fits all solution. Clinicians rely on their past experiences and learnings from peers and the scientific community to make the best decision for JIA patients. Yet despite multiple FDA approved treatments, only 40 percent of patients with a polyarticular form of JIA achieve a state of controlled disease, it is hard to treat. While clinical expertise is invaluable, relying on clinician experience alone at point of care (POC) when treating an uncommon (1 in 1000 children) and disabling disease with heterogeneous presentation and variable treatment response may result in missing a critical window of opportunity for early remission and suboptimal outcome. Imagine the power of a digital health technology (DHT) that uses real world data to inform effectiveness of different treatment strategies for every child. By learning collectively from past experiences of a network of providers and patients that leverage a shared clinical data registry, the DHT synthesizes, and updates knowledge centered for the patient at the POC, while considering the values and preferences of patients for shared decision-making (SDM). Currently, there is no such DHT available to deliver the patient-centered comparative effectiveness evidence at the POC. This study aims to close such a gap with novel user-centered design to adapt for use at POC an existing researcher facing application package based on advanced Bayesian causal learning methods applied to real world electronic health record (EHR) data. This DHT, called PCATS.JIA (Patient centered adaptive treatment strategies for JIA care) will be implemented within the setting of a 23-center learning health network. In R21 phase, the study aims to: 1) bring the PCATS platform to the POC by developing a PCATS.JIA DHT that is EHR system agnostic; 2) co-design a graphic user interface for PCATS.JIA as a patient-centered SDM tool together with key stakeholders (patients, parents, and clinicians); and 3) pilot test PCATS.JIA as a SDM tool at POC in a single rheumatology practice. In R33 phase, we will refine, test, and evaluate the PCATS.JIA in three pediatric rheumatology clinics by 1) using quality improvement science and implementation science principles to implement PCATS.JIA into the clinic workflow with high reliability; and 2) demonstrating that use of PCATS will result in improved health care service as reflected in more patient-centered care, improved health outcomes, and health equity. The study promises to deliver a novel patient centered DHT that is robust, flexible and improves quality and equity of care and treatment outcomes as tested within the practice of pediatric rheumatology. Being EHR system agonistic, it is scalable to different clinical centers. This DHT has the potential to be generalizable to broader disease conditions and patient populations.