Data Driven Approaches to Improving Risk Prediction of Pulmonary Complications After Major Inpatient Surgery

摘要 背景：术后肺部并发症（PPC）是发病率的常见和主要驱动因素 和住院大手术后的死亡率。各种风险预测评分可识别出高风险的患者， 开发PPC和观察性研究将围手术期护理实践与随后的 风险然而，麻醉提供者没有患者特定的基于证据的干预措施来预防 肺部并发症 研究：拟议的研究将利用丰富的围手术期信息来确定 患者、程序和护理过程因素的相互作用使患者面临PPC风险。这将 将数据科学的进步纳入PPC的术前预测（目标1）。然后我们将修改和 根据来自除颤器、监护仪和 患者对在麻醉护理递送期间做出的真实的生活决定的反应（目标2）。这将允许 了解哪些特征最可能导致患者特定风险（目标3）。拟议的研究和 培训将为Colquhoun博士提供数据科学方面的技能，以便他过渡到独立研究人员。 候选人：道格拉斯A Colquhoun博士是纽约州立大学麻醉学终身助理教授。 密歇根大学。他是麻醉学和重症监护医学委员会认证， 积极的临床实践，在围手术期护理的病人接受大手术。在T32研究期间 培训补助金，博士Colquhoun发展的专业知识，在派生的结果和过程的照顾， 电子病历数据。他的长期职业目标是预防术后肺部并发症 通过为麻醉提供者提供数据驱动的策略，在护理点进行管理。 环境：密歇根大学是多中心围手术期的协调中心 结果组（MPOG）是一个由50个麻醉和外科部门组成的国际联盟， 围手术期信息系统。Sachin Kheterpal，医学博士，MBA是Colquhoun博士的主要导师， MPOG的主管Kheterpal博士和麻醉科有着丰富的发展历史， 并部署创新的软件解决方案，以解决围手术期医学和研究中的问题。博士 Colquhoun还将得到来自整个机构的专家共同导师的建议， 预防和处理术后并发症咨询小组专家。

ABSTRACT BACKGROUND: Postoperative pulmonary complications (PPCs) are common and major drivers of morbidity and mortality after major inpatient surgery. Various risk prediction scores identify patients at high risk of developing PPCs and observational research has connected peri-operative care practices with subsequent risk. However, anesthesia providers do not have patient specific evidence based interventions to prevent pulmonary complications. RESEARCH: The proposed research will draw on a wealth of perioperative information available to identify the interactions of patient, procedure and process of care factors which place patients at risk of PPCs. This will incorporate advances in data science to the pre-operative prediction of PPCs (Aim 1). We will then revise and improve this estimate in light of the high fidelity intraoperative data stream from ventilators, monitors and patient response to real life decisions being made during the delivery of anesthesia care (Aim 2). This will allow understanding of what features most contributed to patient specific risk (Aim 3). The proposed research and training will provide Dr Colquhoun with the skills in data science to his transition to an independent researcher. CANDIDATE: Dr Douglas A Colquhoun is a tenure track Assistant Professor of Anesthesiology at the University of Michigan. He is board certified in Anesthesiology and Critical Care Medicine and maintains an active clinical practice in the perioperative care of patients undergoing major surgery. During a T32 Research Training Grant, Dr Colquhoun developed expertise in the derivation of outcomes and processes of care from electronic medical record data. His long term career goal is to prevent postoperative pulmonary complications by offering anesthesia providers data driven strategies for management delivered at the point of care. ENVIRONMENT: The University of Michigan is the coordinating center for the Multicenter Perioperative Outcomes Group (MPOG) an international consortium of 50 anesthesiology and surgical departments with perioperative information systems. Sachin Kheterpal, MD, MBA is the primary mentor for Dr. Colquhoun, and is the Director for MPOG. Dr Kheterpal and the Department of Anesthesiology have a rich history of developing and deploying innovative software solutions to address problems in perioperative medicine and research. Dr Colquhoun will additionally be advised from expert co-mentors drawn from across the institution and a scientific advisory panel expert in the prevention and management of postoperative complications.