PsycheMERGE: Leveraging electronic health records and genomics for mental health research

PsycheMERGE：利用电子健康记录和基因组学进行心理健康研究

• 批准号: 10066366
• 负责人: Lea K Davis
• 金额: $ 75.04万
• 依托单位: MASSACHUSETTS GENERAL HOSPITAL
• 依托单位国家: 美国
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-02-15 至 2022-11-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10066366
• 关键词: Address; Adult; Age; Algorithms; Attention deficit hyperactivity disorder; Biological; Bipolar Disorder; Childhood; Chronic Disease; Clinical; Clinical Research; Communities; Complex; Computerized Medical Record; Data; Diagnosis; Diagnostic; Disease; Early Diagnosis; Early treatment; Eating Disorders; Electronic Health Record; Electronic Medical Records and Genomics Network; Ethnic Origin; Genetic; Genetic Research; Genomic medicine; Genomics; Gilles de la Tourette syndrome; Goals; Health; Health system; Heritability; Individual; Infrastructure; Insurance Coverage; Link; Machine Learning; Major Depressive Disorder; Mental Health; Mental disorders; Methods; Modeling; Mood Disorders; National Human Genome Research Institute; Obsessive-Compulsive Disorder; Outcome; Patients; Performance; Phenotype; Probability; Process; Psychiatry; Psychopathology; Public Health; Race; Research; Research Personnel; Resources; Risk; Risk Factors; Schizophrenia; Site; Suicide; Surveillance Modeling; Testing; Training; Validation; Variant; autism spectrum disorder; base; biobank; case control; clinical decision support; clinical risk; clinically relevant; comorbidity; cost; data resource; delay sex; disability; disorder risk; effective therapy; genome wide association study; genome-wide; genomic predictors; health care service utilization; high risk; improved; indexing; mortality; neuropsychiatric disorder; phenotyping algorithm; polygenic risk score; precision medicine; prediction algorithm; predictive modeling; psychiatric genomics; random forest; risk prediction; sex; support tools; support vector machine

Neuropsychiatric disorders are the leading causes of disability in the US and are associated with increased mortality (e.g. through suicide and associations with chronic diseases and their risk factors). Evidence suggests that early detection and treatment of psychiatric illness is essential to improving long-term outcomes and may even modify illness trajectories at a biological level. Unfortunately, a substantial proportion of patients undergo a long diagnostic odyssey before receiving an appropriate diagnosis and initiating effective treatment. Efforts to improve surveillance for emerging or occult psychopathology are often complex, costly, and have limited yield. Thus, there is an urgent public health need to improve clinical decision support for the early detection of psychiatric disorders in clinical settings. The growing availability of large-scale biobanks linking EHRs to biospecimens has created a powerful, but still relatively untapped, opportunity for psychiatric research. In 2007, the NHGRI organized the Electronic Medical Records and Genomics (eMERGE) network which has brought together investigators around the U.S. to facilitate EHR-based genomic research and the implementation of genomic medicine. To date, however, EHR-based risk prediction and genomics have not been widely leveraged for psychiatric research. To address this gap, we have created a new, large-scale collaborative consortium—PsycheMERGE—which leverages the resources and existing infrastructure of the eMERGE network, the Psychiatric Genomics Consortium (PGC), and local EHR and biobank resources. In this proposal, we aim to: (1) phenotypically and genomically validate and harmonize case and control phenotypes across multiple disorders (2) build clinically-useful risk surveillance models for mood disorders that also leverage cross-institutional genomewide data, and (3) examine whether EHR- and genomic-based risk profiles are associated with clinically-relevant health outcomes. We will further use these risk profiles to examine disparities in diagnostic delay by age, sex and race/ethnicity. The resulting diagnostic and risk prediction algorithms will be made available to the scientific community through the eMERGE network. Successful completion of these aims would represent a major advance in demonstrating the utility of EHR resources for precision medicine approaches to psychiatry, provide the first step toward clinical decision support tools that can be implemented within health systems, and create an invaluable resource for the scientific community.

神经精神疾病是美国残疾的主要原因，并且与增加 死亡率（例如自杀以及与慢性疾病及其危险因素的关联）。 表明精神疾病的早期发现和治疗对于改善长期结果至关重要 不幸的是，很大一部分患者甚至可能在生物学水平上改变疾病轨迹。 在接受适当的诊断并开始有效的治疗之前，要进行长时间的诊断诊断。 改善对新出现或隐匿的精神病理学的监测的努力往往是复杂的、成本高昂的，并且 因此，公共卫生部门迫切需要改善早期临床决策支持。 临床环境中精神疾病的检测越来越多的大型生物库连接起来。 生物样本的电子病历为精神病学创造了强大但相对尚未开发的机会 2007 年，NHGRI 组织了电子病历和基因组学 (eMERGE) 网络。 它汇集了美国各地的研究人员，以促进基于电子病历的基因组研究和 然而，迄今为止，基于 EHR 的风险预测和基因组学尚未得到实施。 为了解决这一差距，我们创建了一个新的、大规模的。 合作联盟——PsycheMERGE——利用了资源和现有基础设施 eMERGE 网络、精神病基因组学联盟 (PGC) 以及当地 EHR 和生物库资源。 根据提案，我们的目标是：（1）从表型和基因组角度验证和协调病例和对照表型 跨越多种疾病 (2) 建立临床上有用的情绪障碍风险监测模型 利用跨机构全基因组数据，以及 (3) 检查是否基于 EHR 和基因组的风险概况 与临床相关的健康结果相关。我们将进一步使用这些风险概况来检查。 诊断延迟因年龄、性别和种族/民族而存在差异 由此产生的诊断和风险预测。 算法将通过 eMERGE Success 网络提供给科学界。 这些目标的完成将代表着在展示电子病历资源的效用方面取得了重大进步 精神病学的精准医学方法，为临床决策支持工具迈出了第一步 可以在卫生系统内实施，并为科学界创造宝贵的资源。