In Silico Model for Acute Lung Injury Prediction and Clinical Trial Design

急性肺损伤预测和临床试验设计的计算机模型

• 批准号: 7814744
• 负责人: OGNJEN GAJIC
• 金额: $ 93.14万
• 依托单位: MAYO CLINIC ROCHESTER
• 依托单位国家: 美国
• 财政年份: 2010
• 资助国家: 美国
• 起止时间: 2010-08-15 至 2012-08-14
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7814744
• 关键词: Accounting; Acute; Acute Lung Injury; Address; Admission activity; Adult Respiratory Distress Syndrome; Affect; Alveolar; Area; Biology; Blood capillaries; Cardiopulmonary Physiology; Caregivers; Characteristics; Clinical; Clinical Research; Clinical Trials; Clinical Trials Database; Clinical Trials Design; Complication; Computer Simulation; Computerized Medical Record; Conduct Clinical Trials; Critical Care; Critical Illness; Databases; Development; Disease; Dreams; Early Diagnosis; Empirical Research; Engineering; Enrollment; Environment; Ethics; Event; Floods; Functional disorder; Future; Goals; Hospital Mortality; Hospitals; Hour; Impairment; Individual; Inflammation; Injury; Intensive Care; Knowledge; Length of Stay; Liquid substance; Literature; Measurement; Mechanical Ventilators; Mechanical ventilation; Medical; Medicine; Membrane; Methodology; Modeling; Morbidity - disease rate; Multicenter Studies; Multiple Organ Failure; Operative Surgical Procedures; Organ failure; Outcome; Pathogenesis; Patients; Physiological; Physiology; Prevention; Prevention strategy; Preventive Clinical Trial; Preventive Intervention; Property; Public Health; Publishing; Quality of life; Reaction Time; Rehabilitation therapy; Research; Research Infrastructure; Research Personnel; Respiratory Failure; Risk; Role; Schedule; Sepsis; Shock; Solutions; Source; Stratification; Supportive care; System; Testing; Therapeutic; Time; Translational Research; Trauma; United States; Ventilator-induced lung injury; Vision; base; capillary; care burden; cost; data mining; design; evidence base; heuristics; high risk; improved; injury prevention; intervention effect; lung injury; mathematical model; mortality; multidisciplinary; novel; outcome forecast; public health relevance; statistics; surfactant

DESCRIPTION (provided by applicant): This application addresses the broad Challenge Area (15) Translational Science and the specific Challenge Topic 15-LM-103: In silico hypothesis testing for biology and medicine. The objective of this project is to determine the feasibility of mathematical models in both optimizing clinical trial designs for, and predicting onset of, Acute Lung Injury. Acute Lung Injury (ALI) and its more severe form, acute respiratory distress syndrome (ARDS) are representative examples of important critical care ailments affecting 200,000 patients each year in the US. Subsequent progression to multiple organ failure carries a grave prognosis with high mortality, and long term morbidity affecting both patients and caregivers. Carefully designed clinical trials have confirmed the relationships between mechanical ventilators and fluid management and their effects on ALI, hence leading to improvements in supportive treatment. Unfortunately, feasibility, as well as ethical and financial barriers greatly limit the ability to conduct meaningful clinical trials in the critical care setting. Indeed, the impact of clinical trials to date has been marginal, and certainly not cost-effective. To circumvent the aforementioned difficulties related to ALI in critical care medicine, this unique collaborative proposal between expert critical care researchers and mathematicians will expand existing complementary mathematical models and assess their ability individually and in tandem to predict the onset of ALI, and enable in silico hypothesis testing. This novel modeling approach will assist in design and implementation of efficient clinical trials targeting prevention of this devastated disorder. Specifically, the investigators will combine: a) mechanistic model that elucidates the pathophysiology of ALI, b) a rule based model to capture clinicians' putative knowledge and expertise, and c) a probabilistic/data mining model derived from empirical evidence extracted from electronic medical records and clinical trial databases. The goal is to address the complexity of ALI at the systems' level by leveraging each individual model's particular strengths and characteristics. Hence, such an approach estimates a patient's physiologic status based on both quantitative measurements and qualitative clinical observations enabling the predicting of ALI and evaluating therapeutic options. The benefit of this methodology will be evaluated by comparing our model's prediction against the US Critical Illness and Injury Trials Group-Lung Injury Prediction Study. The aim of the USCIITG-LIPS1 clinical study, slated to start in May of 2009, is to identify patients at high risk of ALI for future enrollment into preventive clinical trials. The results of this project will include an assessment of the feasibility of such an approach to improving clinical trial design, as well as predicting ALI/ARDS onset. Our vision is a validated methodology for predicting acute and devastating ICU illnesses - transitioning from today's evidence-based to tomorrow's model-based medicine. PUBLIC HEALTH RELEVANCE: In the US, Acute Lung Injury (ALI) and its more severe form Acute Respiratory Distress Syndrome (ARDS) are a major public health problem afflicting 200,000 patients per year and accounting for an in-hospital mortality of 40% and 3.5 million hospital-days. Despite advancements in supportive care the burden of ALI remains high not only in terms of mortality and morbidity, but also long term decrease in quality of life, and enormous cost of both intensive care and rehabilitation. Very little has been done on the prevention of this devastating hospital complication, largely due to difficulties in timely identification of patients at high risk as well as ethical and financial barriers. The impact of clinical trials has been marginal, and certainly not cost-effective. Conventional clinical trials are indeed unlikely to provide solution for effective prevention of ALI and ARDS. This research will be conducted by a unique multidisciplinary team of clinical research and engineering experts, aiming to design a mathematical model of ALI development for future in silico hypothesis testing of different ALI prevention strategies, and better design of clinical trials. It will allow for improved patient outcome by early diagnosing ALI as well as by testing the effect of interventions without putting the patient at risk.

描述（由申请人提供）： 该应用程序解决了广泛的挑战领域（15）转化科学和特定的挑战主题15-LM-103：生物学和医学的计算机假设检验。本项目的目的是确定数学模型在优化急性肺损伤临床试验设计和预测急性肺损伤发作方面的可行性。急性肺损伤（ALI）及其更严重的形式，急性呼吸窘迫综合征（ARDS）是美国每年影响20万患者的重要重症监护疾病的代表性例子。随后进展为多器官衰竭具有严重的预后，死亡率高，长期发病率影响患者和护理人员。精心设计的临床试验已经证实了机械通气和液体管理之间的关系及其对ALI的影响，从而改善了支持性治疗。不幸的是，可行性以及伦理和财务障碍极大地限制了在重症监护环境中进行有意义的临床试验的能力。事实上，迄今为止，临床试验的影响是微不足道的，当然也不符合成本效益。为了规避上述与重症监护医学中的ALI相关的困难，专家重症监护研究人员和数学家之间的这种独特的合作提案将扩展现有的互补数学模型，并单独和串联评估其预测ALI发作的能力，并实现计算机假设检验。这种新的建模方法将有助于设计和实施有效的临床试验，以预防这种破坏性疾病。具体而言，研究者将结合联合收割机：a）阐明ALI病理生理学的机制模型，B）基于规则的模型，以获取临床医生的推定知识和专业知识，以及c）从电子病历和临床试验数据库中提取的经验证据导出的概率/数据挖掘模型。我们的目标是通过利用每个模型的特定优势和特性，在系统级别上解决ALI的复杂性。因此，这种方法基于定量测量和定性临床观察来估计患者的生理状态，从而能够预测ALI并评估治疗选择。将通过比较我们的模型预测与美国危重疾病和损伤试验组肺损伤预测研究来评估这种方法的益处。USCIITG-LIPS 1临床研究计划于2009年5月开始，其目的是确定未来纳入预防性临床试验的ALI高风险患者。该项目的结果将包括评估这种方法改善临床试验设计的可行性，以及预测ALI/ARDS发作。我们的愿景是一种经过验证的方法，用于预测急性和毁灭性的ICU疾病-从今天的循证医学过渡到明天的基于模型的医学。 公共卫生相关性：在美国，急性肺损伤（ALI）及其更严重的形式急性呼吸窘迫综合征（ARDS）是一个主要的公共卫生问题，每年折磨200，000名患者，并造成40%的住院死亡率和350万个住院日。尽管在支持性治疗方面取得了进步，但ALI的负担仍然很高，不仅在死亡率和发病率方面，而且在生活质量方面长期下降，以及重症监护和康复的巨大成本。在预防这种毁灭性的医院并发症方面做得很少，主要是由于难以及时识别高风险患者以及道德和经济障碍。临床试验的影响是微不足道的，当然也不符合成本效益。传统的临床试验确实不太可能提供有效预防ALI和ARDS的解决方案。这项研究将由一个独特的多学科临床研究和工程专家团队进行，旨在设计一个ALI发展的数学模型，用于未来不同ALI预防策略的计算机假设检验，以及更好的临床试验设计。它将允许通过早期诊断ALI以及通过测试干预的效果而不使患者处于风险中来改善患者的预后。