Using Routine Care Electronic Medical Record Data and Artificial Intelligence to Develop a Passive Digital Marker to Predict Postoperative Delirium

使用常规护理电子病历数据和人工智能开发被动数字标记来预测术后谵妄

• 批准号: 10449523
• 负责人: Sanjay Mohanty
• 金额: $ 12.61万
• 依托单位: INDIANA UNIV-PURDUE UNIV AT INDIANAPOLIS
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-01 至 2027-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10449523
• 关键词: Active Learning; Acute; Age; Aging; Alzheimer' s Disease; Alzheimer' s disease model; American; Artificial Intelligence; Attention; Award; Cardiopulmonary; Cardiovascular system; Clinical Sciences; Cognition; Cognitive; Computerized Medical Record; Data; Databases; Delirium; Dementia; Detection; Development; Development Plans; Diagnosis; Disease; Dropout; Elderly; Electronic Health Record; Enrollment; Environment; Funding; Future; Goals; Health; Health system; Healthcare; Hospitals; Impaired cognition; Indiana; Informed Consent; Institutes; International; Intervention; K-Series Research Career Programs; Length of Stay; Machine Learning; Medical Informatics; Mentors; Methods; Modeling; Operative Surgical Procedures; Outcome; Patient Care; Patients; Performance; Perioperative; Physiological; Polypharmacy; Postoperative Period; Preparation; Prevention; Process; Psychiatric Diagnosis; Randomized Clinical Trials; Randomized Controlled Clinical Trials; Randomized Controlled Trials; Recovery of Function; Reproducibility; Research; Research Personnel; Research Project Grants; Risk; Risk Factors; Senior Scientist; Structure; Surgical complication; Survivors; Syndrome; System; Techniques; Testing; Text; Time; Translational Research; United States National Institutes of Health; Universities; acceptability and feasibility; advanced analytics; brain health; career; career development; clinical decision support; cognitive disability; cognitive enhancement; cognitive recovery; cohort; comorbidity; cost; design; digital; digital models; efficacy evaluation; efficacy trial; feasibility testing; frailty; functional decline; functional disability; functional status; health data; high risk; implementation science; improved; machine learning algorithm; mortality; multi-component intervention; multidisciplinary; nutrition; older patient; patient oriented; postoperative delirium; predictive modeling; prevent; prospective; research study; risk prediction; risk stratification; routine care; screening; support tools; usability

PROJECT SUMMARY Postoperative delirium is among the most common complications following elective surgery. Delirium is associated with prolonged lengths of stay, functional decline, cognitive impairment, higher costs, and higher mortality. This proposal describes a career development plan that will transform Dr. Mohanty into a patient- oriented investigator focused on improving the perioperative brain health of every older American undergoing major surgery. Dr. Mohanty’s proposed project will involve merging the data of more than 35000 patients who underwent major surgery in a statewide health system with electronic health record (EHR) data from the Indiana Network for Patient Care (INPC), a Regional Health Information Exchange. Using this merged database, the candidate will develop and test a passive predictive model (“digital marker”) for postoperative delirium risk using routine care EHR data, including unstructured, free text notes, and a machine learning algorithm. This “digital marker” will then be tested in a pilot randomized clinical trial in preparation for a large efficacy trial which will evaluate the impact of this scalable “digital marker” on short and long-term cognitive outcomes. The proposed career development plan integrates: close mentoring from a multidisciplinary team of senior scientists; coursework and structured didactics in medical informatics, including advanced analytics, artificial intelligence, and clinical decision support; implementation science; experiential learning via the conduct of the proposed research project; and a supportive research environment. This environment includes an internationally recognized NIH-funded delirium study group, the distinguished Indiana University Center for Aging Research, the NIA-funded Indiana Alzheimer’s Disease Center, and the NIH-funded Indiana Clinical and Translational Sciences Institute (CTSI). This career development award will guarantee protected time that will be necessary to advance the candidate’s career in perioperative brain health. In addition, the award will provide critical support to collect data for a future R01 efficacy randomized controlled trial.

项目总结 术后精神错乱是择期手术后最常见的并发症之一。精神错乱是 与住院时间延长、功能衰退、认知障碍、更高的费用和更高的 死亡率。这份提案描述了一项职业发展计划，它将把莫汉蒂医生转变为一名患者- 定向研究人员专注于改善每一位正在接受手术的美国老年人的围手术期脑健康 大手术。莫汉蒂博士提出的项目将涉及合并超过35000名患者的数据，这些患者 在全州医疗系统中进行了大手术，电子健康记录(EHR)数据来自 印第安纳州病人护理网络(INPC)，一个地区性健康信息交换机构。使用此合并 数据库，候选人将开发和测试一个被动预测模型(“数字标记”)，用于术后 使用常规护理电子病历数据的精神错乱风险，包括非结构化、免费文本笔记和机器学习 算法。这种“数字标记”随后将在试点的随机临床试验中进行测试，为大规模的 疗效试验，将评估这种可扩展的“数字标记”对短期和长期认知的影响 结果。 拟议的职业发展计划包括：由资深学生组成的多学科团队的密切指导 科学家；医学信息学的课程作业和结构化教学，包括高级分析，人工 智能和临床决策支持；实施科学；通过开展 建议的研究项目；以及支持性的研究环境。此环境包括 由NIH资助的国际公认的精神错乱研究小组，著名的印第安纳大学研究中心 老龄化研究，NIA资助的印第安纳州阿尔茨海默病中心，以及NIH资助的印第安纳州临床和 翻译科学研究所(CTSI)。这一职业发展奖将保证受保护的时间将 对于促进应聘者在围手术期大脑健康方面的职业生涯是必要的。此外，该奖项将 为将来的R01疗效随机对照试验收集数据提供关键支持。