Developing a Childhood Asthma Risk Passive Digital Marker

项目摘要/摘要 Arthur Owora博士是一位生物统计学家和定量流行病学家，他的长期职业目标是将 通过设计和测试直观临床决策的有效性对临床实践的预测研究 支持工具。这一目标是基于这样一个概念：应用新的生物统计学和机器学习(ML) 日益可用的电子健康记录(EHR)预后数据的方法可以生成预测性数据 关于疾病风险的分析和见解。然后，临床医生可以利用这些洞察力做出有效的临床决策- 在护理点进行，包括更主动和个性化的护理，以改善以患者为中心的情况 结果。这直接响应了NIH国家心肺血液研究所的战略目标 “优化临床和实施研究，以改善健康和减少疾病。” 为了实现他的长期目标，小原博士将利用他在生物统计学和流行病学方面的研究生培训， 在疾病发育性起源的建模以及先前的预后方面的博士后研究 作为一名翻译科学家过渡到研究独立性的研究经验。为此，他要求 关于如何应用新的生物统计学和ML方法来制定数字临床决策的额外培训- 支持工具，以及2)在临床环境中实施和评估这些工具的有效性。 本提案描述了一个为期4年的项目，以开发和确定可用性、可接受性、可行性和 儿童哮喘被动数字标记物早期发现的初步效果。在这里，一个被动语态 数字标记(PDM)指的是可用于检索和合成预先存在的ML算法 在“数字”电子病历中，“被动地”收集0-3岁的母亲/儿童二元预后数据(即病史)。 为6-10岁儿童哮喘风险和表型提供一个客观和可量化的“标记物”。 建议的具体目标建立在Owora博士正在进行的预后研究的基础上：(1)开发和评估 与儿童哮喘风险评分(作为儿童哮喘风险评分的替代指标)相比，儿童哮喘疾病管理的预测性能 标准实践)，以及(2)确定可用性、可接受性、可行性和初步效果 儿科医生中儿童哮喘相关疾病的研究。 为了达到这些目标，Owora博士提出了包括说教式和体验式的培训活动 学会在开发、实施和评估儿童哮喘产品数据管理方面积累专业知识 临床环境。这些培训活动将得到强大的多学科导师团队的支持： 理查德·霍尔登(卫生信息技术翻译科学家)，Eneida Mendonca(儿科医生和 医疗信息学家)、Robert Tepper(内科医生-科学家和肺病学家)、Malaz Boustani(内科医生和 实施科学家)和道格拉斯·兰西特尔(生物统计学家和生物信息学家)。 有了通过K01奖项获得的初步数据、新技能和专业知识，Owora博士计划 提交一项R01拨款，以评估产品数据管理在改善儿童哮喘早期发现方面的效果。

PROJECT SUMMARY/ABSTRACT Dr. Arthur Owora is a biostatistician and quantitative epidemiologist whose long-term career goal is to translate prognostic research into clinical practice by designing and testing the effectiveness of intuitive clinical decision support tools. This goal is predicated on the notion that applying novel biostatistical and machine learning (ML) methodologies to increasingly available electronic health record (EHR) prognostic data can generate predictive analytics and insights regarding disease risk. Clinicians can then use such insights for effective clinical decision- making at point-of-care, including more proactive and personalized care, for improved patient-centered outcomes. This is directly responsive to NIH National Heart, Lung, and Blood Institute’s strategic objective to “Optimize clinical and implementation research to improve health and reduce disease.” To achieve his long-term goal, Dr. Owora will leverage his graduate training in biostatistics and epidemiology, post-doctoral fellowship in the modeling of developmental origins of disease, as well as previous prognostic research experience to transition to research independence as a translational scientist. To this end, he requires additional training in how to apply novel biostatistical and ML methodologies to develop digital clinical decision- support tools, and 2) implement and evaluate the efficacy of such tools in clinical settings. This proposal describes a 4-year project to develop and determine the usability, acceptability, feasibility, and preliminary efficacy of a childhood asthma Passive Digital Marker for early disease detection. Here, a Passive Digital Marker (PDM) refers to a ML algorithm that can be used to retrieve and synthesize pre-existing ‘Passively’ collected mother/child dyad prognostic data (i.e., medical history) at ages 0-3 years in ‘Digital’ EHR to provide an objective and quantifiable ‘Marker’ of a child’s asthma risk and phenotype at ages 6-10 years. Proposed specific aims build on Dr. Owora’s ongoing prognostic research to: (1) develop and evaluate the predictive performance of a childhood asthma PDM, compared to a Pediatric Asthma Risk Score (as a proxy for standard practice), and (2) determine the usability, acceptability, feasibility, and preliminary efficacy of the childhood asthma PDM among pediatricians. To address these objectives, Dr. Owora proposes training activities that include didactic and experiential learning to build expertise in the development, implementation, and evaluation of the childhood asthma PDM in clinical settings. These training activities will be supported by a strong multidisciplinary team of mentors: Richard Holden (Translational Scientist in Health Information Technology), Eneida Mendonca (Pediatrician and Medical Informatician), Robert Tepper (Physician-Scientist and Pulmonologist), Malaz Boustani (Physician and Implementation Scientist), and Douglas Landsittel (Biostatistician and Bioinformatician). With the preliminary data generated, new skills, and expertise gained through this K01 award, Dr. Owora plans to submit a R01 grant to evaluate the efficacy of the PDM for improved early detection of childhood asthma.