Isolating Mechanisms Underlying Hospital Variation in End-of-Life ICU Use

描述（由申请人提供）：达特茅斯卫生保健地图集已经证明了在生命结束时重症监护病房（ICU）使用的医院之间存在相当大的差异。确定这种差异在多大程度上是由于提供者行为的差异造成的努力受到其他因素的阻碍，这些因素在医院之间同时存在差异，例如患者的临床状况、社会心理环境和治疗偏好。我们建议在目前资助的NIH对高强度学术医疗中心和低强度学术医疗中心（R21 NR010265）的医疗重症监护病房（micu）进行观察性研究的基础上，在相同的两个机构进行模拟实验。虽然观察性研究关注的是患者入住ICU后做出的决定，但我们对终末期癌症患者的模拟旨在研究是否首先将患者送入ICU的决定，考虑到这些事件的不可预测的时间，这一决策事件很难进行观察性研究。此外，通过将医生置于具有相同病例的模拟环境中，我们可以在生命结束强度谱的两端隔离两家医院之间的提供者差异来源。通过探索医生解释其行为的基本原理，我们可以进一步将这种差异的部分原因解析为正式规范（医院政策和程序）和非正式社会规范。在推动人们做出临终决定的诸多因素中，非正式的社会规范是人们最不了解的因素之一。社会规范也可能被社会营销干预所改变。通过用这些实验数据补充我们观察性研究的数据，我们将对生命末期ICU的使用情况有一个更清晰、更广泛的了解，我们将利用这些数据为未来的干预研究提供信息，以提高患者和家属对医生决策的满意度。具体目的如下：目的1：使用高保真模拟比较一个高强度和一个低强度学术医疗中心的医院医生的ICU入院、姑息和代码状态文档决策。目的2：使用来自两个学术医疗中心的模拟数据，并辅以先前从中等强度学术医疗中心收集的数据，研究沟通技巧与ICU入院、姑息和代码状态文档决策之间的关系。目的3：确定影响高强度和低强度学术医疗中心ICU入院、姑息和代码状态文件决策的正式和非正式社会规范。公共卫生相关性：该项目的总体目标是增强我们对终末期癌症患者使用临终（EOL）重症监护病房（ICU）的医院水平差异原因的理解。我们将使用高保真仿真，其复杂程度与飞行员使用的飞行模拟器相似，以评估和比较来自高EOL ICU使用率的学术医疗中心的医院医生与来自低EOL ICU使用率的学术医疗中心的医生的沟通和决策过程。通过将医生置于具有相同病例的模拟环境中，我们可以在生命结束强度谱的两端隔离两家医院之间的提供者差异来源。通过探索医生解释其行为的基本原理，我们可以进一步将这种差异的部分原因解析为正式规范（医院政策和程序）和非正式社会规范。我们的研究结果将用于制定医院层面的干预措施，以改善临终病人的沟通和决策以病人为中心。这项研究还将为我们开发的模拟方法提供进一步的支持，该方法可以在未来用作识别医生行为机制的技术，并作为改变医生行为的培训工具。

DESCRIPTION (provided by applicant): The Dartmouth Atlas of Health Care has demonstrated considerable variability among hospitals in intensive care unit (ICU) use at the end of life. Efforts to determine how much of this variability is due to differences in provider behavior are hindered by other factors that vary simultaneously across hospitals, such as patients' clinical condition, psychosocial circumstances, and preferences for treatment. We propose to augment our currently-funded NIH observational study of the medical intensive care units (MICUs) of a high-intensity academic medical center and a low-intensity academic medical center (R21 NR010265) with a simulation experiment at the same two institutions. While the observational study focuses on decisions that are made after patients have been admitted to the ICU, our simulation of a patient with end-stage cancer is designed to study the decision whether to admit a patient to the ICU in the first place, a decision event that is hard to study observationally, given the unpredictable timing of these events. In addition, by placing physicians in a simulated environment with an identical case, we can isolate the provider sources of variation among two hospitals at opposite ends of the end-of-life intensity spectrum. By exploring the rationales physicians offer to explain their behavior, we can further parse part of the causes of the variation into formal norms (hospital policies and procedures) and informal social norms. Of the many factors driving end-of-life decision making, informal social norms are one of the least understood. Social norms are also potentially modifiable with social marketing interventions. By complementing the data from our observational study with these experimental data, we will have a much clearer and broader picture of ICU use at the end of life, which we will use to inform a future intervention study to improve patient and family satisfaction with physician decision making. The specific aims are: Aim 1: To compare ICU admission, palliation, and code status documentation decisions among hospital- based physicians at one high-intensity and one low-intensity academic medical center using high-fidelity simulation. Aim 2: To examine the relationship between communication skills and ICU admission, palliation, and code status documentation decisions using simulation data from the 2 academic medical centers, augmented by previously collected data from a mid-intensity academic medical center. Aim 3: To identify formal and informal social norms that influence ICU admission, palliation, and code status documentation decisions at the high-intensity and the low-intensity academic medical centers. PUBLIC HEALTH RELEVANCE: The overall goal of this project is to enhance our understanding of the reasons for hospital-level variations in end-of-life (EOL) intensive care unit (ICU) use among patients with end-stage cancer. We will use a high- fidelity simulation, similar in sophistication to flight simulators used for pilots, to assess and compare the communication and decision-making processes of hospital-based physicians from an academic medical center with high EOL ICU use to those of physicians from an academic medical center with low EOL ICU use. By placing physicians in a simulated environment with an identical case, we can isolate the provider sources of variation among two hospitals at opposite ends of the end-of-life intensity spectrum. By exploring the rationales physicians offer to explain their behavior, we can further parse part of the causes of the variation into formal norms (hospital policies and procedures) and informal social norms. Our findings will be used to develop hospital-level interventions to improve the patient-centeredness of communication and decision making for dying patients. This study will also provide further support for the simulation method we have developed, which could be used in the future as a technique for identifying mechanisms underlying physician behavior and as a training tool for changing physician behavior.