Using the electronic health record to risk-stratify patients with systemic lupus erythematosus

项目总结 系统性红斑狼疮(SLE)是一种发病率极高的异质性自身免疫性疾病 和死亡率。疾病的异质性对SLE患者的临床护理和 成功的临床试验。对于异质性疾病，识别聚集性会导致分类的改进， 生物标记物和靶向治疗。电子健康记录(EHR)是识别 对系统性红斑狼疮患者进行分组和风险分层。我们已经发布了一种算法，可以准确地识别SLE患者 并收集了2376名SLE患者的队列，平均随访时间为9年。该数据是 与世界上最大的生物库之一BioVU有关，已经有400名SLE患者进行了基因分型。电子病历和 BioVU允许使用新的方法，例如使用帐单代码的表现组范围关联研究(PheWAS) 对整个电子病历进行全面扫描。我们已经在系统性红斑狼疮中进行了第一次治疗。 我们的总体目标是以一种智能的方式使用现成的EHR数据来改善SLE患者的预后。 我们假设系统性红斑狼疮是由具有不同病程和 并存，我们基于电子病历的方法结合了遗传信息，将作为新的工具 对系统性红斑狼疮患者进行风险分层。 在目标1中使用Phewas，我们将发现合并和不合并SLE患者之间的差异 自身抗体和具有和不具有预先指定的SLE易感性的单核苷酸多态(SNPs)。 在目标2中，我们将使用人口统计学、自身抗体、合并症和系统性红斑狼疮进行聚类分析。 SNPs用于对SLE患者进行风险分层。我们将评估肾脏结果、存活率和接受的治疗 星系团。在目标3中，我们将评估诱导治疗对EHR中SLE肾炎的治疗反应。 并与已发表的结果进行比较。 这些目标是对系统性红斑狼疮患者进行风险分层和确定治疗反应的必要的第一步。 然后建立预测治疗反应的模型，并进行基于EHR的靶向临床试验 治疗。在遗传学、生物医学信息学和 生物统计学将推动巴尔纳多博士的事业发展。 范德比尔特是一个特殊的环境，支持巴纳多博士向独立的 内科科学家。值得注意的优势包括合成衍生品(SD)，这是一种未识别的EHR，超过2.7 百万受试者和BioVU，一个与SD相关的遗传生物库。医学部和医学部 风湿科支持巴尔纳多博士的职业生涯。她的导师克罗福德博士和丹尼博士 在风湿学和生物医学信息学领域获得国际认可，并在以下领域取得了成功 辅导。凭借范德比尔特对年轻研究人员的机构承诺和生物医学领域的专业知识 在信息学方面，Barnado博士将成功地利用她的创新提案获得独立的R01资金。

PROJECT SUMMARY Systemic lupus erythematosus (SLE) is a heterogeneous, autoimmune disease with highly variable morbidity and mortality. Disease heterogeneity is a major challenge to both the clinical care of SLE patients and successful clinical trials. With heterogeneous diseases, identifying clusters results in improved classification, biomarkers, and targeted therapies. Electronic health records (EHRs) represent a powerful tool to identify clusters and risk-stratify SLE patients. We have published an algorithm that accurately identifies SLE patients in the EHR and have assembled a cohort of 2,376 SLE patients with a mean follow-up of 9 years. This data is linked to one of the world's largest biobanks, BioVU, with 400 SLE patients already genotyped. The EHR and BioVU allow for novel methods such as phenome-wide association studies (PheWAS) that use billing codes for a comprehensive scan of the entire EHR. We have performed the first PheWAS in SLE. Our overall goal is to use readily available EHR data in an intelligent way to improve outcomes in SLE patients. We hypothesize that SLE is composed of multiple clusters of patients with different disease courses and comorbidities, and our EHR-based methodology that incorporates genetic information will serve as novel tools to risk-stratify SLE patients. Using PheWAS in Aim 1, we will uncover differences in comorbidities between SLE patients with and without autoantibodies and with and without pre-specified SLE susceptibility single nucleotide polymorphisms (SNPs). In Aim 2, we will perform clustering analyses using demographics, autoantibodies, comorbidities, and SLE SNPs to risk-stratify SLE patients. We will assess renal outcomes, survival, and treatments received among the clusters. In Aim 3, we will evaluate treatment response to induction therapy for SLE nephritis in the EHR and compare to published outcomes. These aims are the necessary first steps to risk-stratify SLE patients and define treatment response in the EHR to then build models to predict treatment response and conduct EHR-based pragmatic clinical trials of targeted therapies. Additional mentored training and didactic coursework in genetics, biomedical informatics, and biostatistics will advance Dr. Barnado's career. Vanderbilt serves as an exceptional environment to support Dr. Barnado's transition to an independent physician scientist. Notable strengths include the Synthetic Derivative (SD), a de-identified EHR with over 2.7 million subjects, and BioVU, a genetic biobank linked to the SD. The Department of Medicine and Division of Rheumatology are in support of Dr. Barnado's career. Her mentors, Drs. Crofford and Denny, are internationally recognized in rheumatology and biomedical informatics with successful track records of mentoring. With Vanderbilt's institutional commitment to young investigators and expertise in biomedical informatics, Dr. Barnado's will successfully leverage her innovative proposal to independent R01 funding.