Developing Clinical Decision Support Systems Adaptive to Clinicians' Fatigue (Cessation Fatigue)

摘要：疲劳是医疗保健的一种潜在危险(尤其是在急诊科)，导致 判断力差，医疗差错增多。疲劳引发的不良事件具有负面的财务和 急诊室和其他环境中的患者/职业安全影响。常用的时间和任务管理 策略(例如，多任务处理)在疲惫的临床医生中效果要差得多。潜在的原因， 已对疲劳引起的不良事件和后果进行了研究，然而降低风险的干预措施 是有限的。目前提出的消除疲劳的解决方案(例如，限制工作时间、减少患者负荷)可以 这是有帮助的，但疲劳是一个复杂的结构，这使得开发简单的解决方案变得不太可行。 几十年的适应性自动化文献表明，临床决策支持(CDS)系统可以适应 到临床医生疲劳的瞬间变化，有可能拦截疲劳引起的人为错误和 避免潜在的不良事件。 对CDS的批评是，它只在有潜力的情况下提供去上下文的决策支持 适应其用户(即一线临床医生)。疲劳程度不同的用户有不同的需求。当CDS 支持是去上下文的，它成为实际上导致临床医生疲劳的背景的一部分。 临床医生不欢迎CDS提示，制定策略避免与CDS互动，这可能会导致 消极的结果。自适应CDS将根据临床医生的疲劳程度进行自我配置，以提供正确的 信息级别，在正确的时间提供给正确的用户。 本研究的主要目的是发展在EDS中适应性CDS的基础，即对 用户的疲劳和适应用户的疲劳程度。将使用混合方法设计来实现我们的 目的通过两个目标：(1)检验ED临床医生疲劳对(A)临床决策和 (B)将CDS用于抗生素处方；(2)开发和评估CDS的设计和实施 适应ED临床医生疲劳的CDS指南。 本研究的独特贡献在于(1)为一种新型的健康信息技术奠定了基础 (2)将认知决策理论融入到CDS设计中；(3)开发了一种 CDS适应了普遍存在的负面工作条件，疲劳。 将全面传播三项主要成果。首先，我们将提供详细的描述 疲劳对临床决策的影响。其次，我们将对疲劳的影响进行详细描述 临床决策支持系统在急救系统中的应用。第三，我们将报告自适应的设计指南 CD，从而支持其他学者和设计师的可复制性。最终，这项提议将得到改善 临床医生在具有挑战性的工作条件下的表现，从而影响患者的预后。

Abstract: Fatigue is a latent hazard in health care, (particularly in emergency departments-ED), leading to poor judgement and increased medical errors. Fatigue-induced adverse events have negative financial and patient/occupational safety impact in EDs and other settings. Commonly used time- and task-management strategies (e.g., multitasking), are much less effective in clinicians with fatigue. Potential causes, consequences, and fatigue-induced adverse events, have been studied, however interventions to mitigate risks is limited. Currently proposed solutions to obviate fatigue (e.g., limit working hours, decreased patient load) can be helpful, but fatigue is a complex construct, making it less feasible to develop uncomplicated solutions. Decades of adaptive automation literature suggest that clinical decision support (CDS) systems that can adapt to in-the-moment variations in clinician’s fatigue, have potential to intercept fatigue-induced human errors and preclude potential adverse events. A criticism of CDS, is that it only provides decontextualized decision support when it has potential to be adapted to its users (i.e., frontline clinicians). Users with different fatigue level have different needs. When CDS support is decontextualized, it becomes part of the background that actually contribute to clinician fatigue. Clinicians not welcoming CDS prompts, develop strategies to avoid interacting with the CDS, which can lead to negative outcomes. Adaptive CDS would configure itself based on a clinician’s fatigue level to provide the right level of information, to the right user, at the right time. The primary objective of this study is to develop the foundation for adaptive CDS in EDs, that is sensitive to a user’s fatigue and adapts to the user’s fatigue level. A mixed method design will be used to achieve our objective through two aims: (1) Examine the impact of ED clinician fatigue on (a) clinical decision making and (b) the use of the CDS for antibiotic prescription; (2) Develop and evaluate CDS design and implementation guidelines for a CDS that adapts to ED clinician fatigue. The unique contribution of this study lies in (1) creating a foundation for a novel health information technology (HIT), adaptable CDS; (2) integrating cognitive decision-making theories into the CDS design; (3) developing a CDS to accommodate the prevalent negative work condition, fatigue. Three main deliverables will be disseminated comprehensively. First, we will provide a detailed description of impact of fatigue on clinical decision making. Second, we will provide a detailed description of impact of fatigue on the use of clinical decision support systems in EDs. Third, we will report on design guidelines for adaptive CDS, thereby supporting replicability by other scholars and designers. Eventually, this proposal will improve clinician’s performance under challenging work conditions, hence patient outcomes.