Modeling Attention and Situation Awareness in Acute Care Workplaces

态势感知-了解正在发生的事情-对于安全关键型故障的操作员非常重要。失去态势感知可能危及安全，甚至导致死亡。在本项目中，我们调查的安全关键领域，如急性护理的情况意识，考虑一个基于模型的，连续的，客观的措施来衡量和预测的情况意识。SEEV模型综合了显著性、努力、期望和价值等因素来描述外显视觉注意的分配。注意力-情境感知（A-SA）模型扩展了SEEV模型，以产生一般的情境感知测量。然而，对SEEV模型的研究调查了相当小的误差，只涉及一个操作员，主要目的是改进模型拟合，并且是在模拟环境中进行的; A-SA模型尚未进行实证研究。我们的目标是通过扩展SEEV模型，以更好地适应来自真实的环境的数据，从而了解急诊科工作人员的注意力分布（目标1）并验证扩展模型（目标2）。为此，我们将扩展本模型通过操作的显着性和努力参数，以适应需求的大型和更分散注意力的急性护理工作环境，我们将考虑团队和个人的任务，以捕捉共享和独特的注意力需求的团队情况。我们将通过记录眼动追踪数据并对真实的全身麻醉诱导过程中团队（麻醉师和麻醉护士）的注意力分布进行建模来评估扩展模型。我们的目标是实现和验证A-SA模型，该模型通过考虑各种认知因素（如工作量、操作员经验和认知偏差），以SEEV模型的形式扩展注意力分布（目标3）。A-SA模型可以基于注意力分布数据和考虑认知因素来预测特定时间点的态势感知。为了进行验证，我们将运行一个医疗模拟，收集眼动跟踪数据以及几个所谓的SAGAT探针，这些探针在文献中被认为是一种客观有效的态势感知措施。我们将评估A-SA模型在过程中的各个点上预测SAGAT分数的能力。该项目将深入了解安全关键领域中人类注意力分配的一般机制，并将深入了解态势感知和团队态势感知的构建。从应用的角度来看，这些见解可以用来提高态势感知，通过促进特定技术的设计或使用模型进行培训，并评估技术与态势感知。我们的最终目标是帮助理解人类在社会技术系统中的认知，并为患者安全做出贡献。

Situation awareness – being aware of what is going on – is important for operators in safety-critical workspaces. Losing situation awareness may jeopardize safety or even result in fatalities. In the present project, we investigate situation awareness in safety-critical domains such as acute care by considering a model-based, continuous, and objective measure to measure and predict situation awareness. The SEEV model integrates the factors Salience, Effort, Expectancy, and Value to describe the allocation of overt visual attention. The attention-situation awareness (A-SA) model extends the SEEV model to produce a general situation awareness measure. Studies of the SEEV model, however, have investigated rather small workspaces, have only addressed a single operator, were mainly aimed at improving the model fit, and were conducted in simulated settings; the A-SA model has not been empirically investigated.We aim to understand the attention distribution of staff in acute care workspaces by extending the SEEV model to accommodate data from real environments better (Aim 1) and validate the extended model (Aim 2). To this end, we will extend the present model by operationalizing the salience and effort parameters to suit the demands of the large and more distracting acute care work environment, and we will consider team and individual tasks in order to capture the shared and distinct attention demands of the team situation. We will evaluate the extended model by recoding eye-tracking data and modeling the attention distribution of the team (anesthesiologist and anesthetic nurse) during real inductions of general anesthesia.We aim to implement and validate the A-SA model which extends the attention distribution in form of the SEEV model by considering various cognitive factors such as workload, operators’ experience, and cognitive biases (Aim 3). The A-SA model can make predictions about situation awareness at specific points in time based on attention distribution data and consideration of cognitive factors. For the validation, we will run a medical simulation and collect eye tracking data as well as several so-called SAGAT probes, which are considered to be an objective and valid situation awareness measure in the literature. We will evaluate how well the A-SA model can predict SAGAT scores at various points during the procedure.The project will provide insights into the general mechanisms of human attention allocation in safety-critical domains, and it will also provide insights into the construct of situation awareness and team situation awareness. From an applied perspective, these insights can be used to improve situation awareness by facilitation the designing of specific technology or use of the models for training purposes and evaluating technology in relation to situation awareness. Our ultimate aim is to contribute to the understanding of cognition of humans in socio-technical systems and contribute to patient safety.