POWER: Predicting Obesity with Enhanced EHR Resources

摘要 肥胖是一个普遍存在的公共卫生问题，起源于儿童时期，尽管在治疗方面取得了进展， 肥胖，初级预防是预防发病和早死的关键。有一个迫切的、未得到满足的需要， 预测哪些婴儿和幼儿肥胖的风险最高。生命的最初几年提供了一个充满希望的 这是一个敏感时期，以解决婴儿体重过度、快速增加以及随后的肥胖和心脏代谢风险。 以前的研究表明，风险因素与婴儿体重快速增加之间存在关联，但新的增长 婴儿期BMI峰值的大小和时间等特征可能会更好地了解肥胖风险 与已知的风险因素相结合。生命最初几年的初级保健就诊，时间短，次数多 与渴望得到指导的父母接触，为风险预测和干预提供了理想的环境。的 总体假设是，在临床环境中的儿童肥胖预防策略可以得到改善 通过预测模型和临床决策支持，旨在纳入可改变的风险因素， 婴儿生长模式该提案的总体目标是改善以下临床相关预测： 儿童肥胖并设计一个包含实时风险预测的临床决策支持工具。要求1 确定了婴儿期已知的危险因素与新的婴儿期生长模式之间的关联。目的2 开发、比较和验证使用新的婴儿生长特征的预测模型，包括 个人和群体模式。目标3测试风险预测和临床决策的实施 电子健康记录中的支持工具。概述研究目标和职业发展计划 提供查尔斯伍德，医学博士，公共卫生硕士的技能，以实现他成为一个独立的整体职业目标 研究人员专注于在初级保健环境中预防肥胖，重点是生命的最初几年。博士 伍德的培训计划包括体验式学习和教学课程，以实现以下短期目标 培训目标：1）掌握重复测量分析和预测建模的方法， 验证; 2）使用电子健康记录（EHR）源执行数据链接和协调; 3）学习 并实践EHR数据的最佳使用研究; 4）将最先进的方法纳入设计 临床决策支持工具; 5）继续发展职业和专业技能。伍德博士将收到 来自儿科成果研究专家团队（史密斯博士）的重点指导和咨询， 初级保健肥胖干预（佩林博士），肥胖和人口健康（斯金纳博士），流行病学的 儿童成长（吴博士），重复测量分析（Kuchibhatla博士）和EHR用于研究（Dr. Goldstein）。杜克大学丰富的研究环境将使伍德博士能够完成他的研究， 职业发展计划，并开始解决他的长期目标，比较风险预测策略， 儿童肥胖和进行前瞻性观察和干预性试验，在初级保健设置。

ABSTRACT Obesity is a pervasive public health problem with origins in childhood, and despite advances in treatment for obesity, primary prevention is essential to prevent morbidity and early death. There is an urgent, unmet need to predict which infants and young children are at highest risk of obesity. The first years of life provide a promising sensitive period to address excess, rapid infant weight gain and subsequent obesity and cardiometabolic risk. Previous work demonstrates associations between risk factors and rapid infant weight gain, but novel growth characteristics, such as magnitude and timing of infancy BMI peak, may provide better insight into obesity risk when combined with known risk factors. Primary care visits in the first years of life, with brief and frequent contact with parents eager for guidance, provide an ideal setting for both risk prediction and interventions. The overarching hypothesis is that childhood obesity prevention strategies in the clinical settings can be improved through predictive models and clinical decision support designed to incorporate modifiable risk factors and infancy growth patterns. The overall objectives of this proposal are to improve clinically-relevant prediction of childhood obesity and design a clinical decision support tool that incorporates real-time risk prediction. Aim 1 identifies associations between known risk factors during infancy and novel infancy growth patterns. Aim 2 develops, compares, and validates predictive models using novel infant growth characteristics, including individual and group-based patterns. Aim 3 tests implementation of a risk prediction and clinical decision support tool within the electronic health record. The outlined research aims and career development plan provides Charles Wood, MD, MPH the skills to achieve his overall career goal of becoming an independent investigator focused on obesity prevention in the primary care setting focused on the first years of life. Dr. Wood’s training plan includes experiential learning and didactic coursework to achieve the following short-term training goals: 1) master approaches to repeated measures analysis and predictive modelling construction and validation; 2) execute data linkage and harmonization using electronic health record (EHR) sources; 3) learn and practice optimal use of EHR data for research; 4) incorporate state-of-the-art approaches to designing clinical decision support tools; and 5) continue to develop career and professional skills. Dr. Wood will receive focused mentorship and consultation from a team of experts of pediatric outcomes research (Dr. Smith), primary care obesity interventions (Dr. Perrin), obesity and population health (Dr. Skinner), epidemiology of childhood growth (Dr. Woo), repeated measures analysis (Dr. Kuchibhatla), and EHR use for research (Dr. Goldstein). The rich research environment at Duke University will allow Dr. Wood to fulfill his research and career development plans and begin to address his long-term goals of comparing risk prediction strategies for childhood obesity and conducting prospective observational and interventional trials in the primary care setting.