Clinical Data in Context
Clinical Data in Context
复制标题
上下文中的临床数据
DOI：
10.1145/3314409
复制
发表时间：
2019
影响因子：
--
通讯作者：
Mark W. Newman
中科院分区：
文献类型：
--
作者：
Shriti Raj;Joyce M. Lee;Ashley Garrity;Mark W. Newman
用上下文数据增强的临床数据可以帮助慢性病患者理解他们的疾病。然而，现有的工具不支持多个数据流的解释。为了更好地了解个人如何理解临床和背景数据，我们采访了1型糖尿病患者及其护理人员，使用患者数据的背景增强可视化作为探针，以促进解释活动。我们观察到，我们的参与者在解释他们的数据时进行了四项分析活动--发现基于背景的趋势并解释它们，对多个因素进行三角测量，建议特定于背景的行动，并假设影响结果的其他背景因素。我们还观察到在分析过程中遇到的两个挑战-无法识别明确的趋势挑战行动计划和违反直觉的见解损害了对数据的信任。将我们的发现置于现有的意义构建框架中，我们证明意义构建不仅可以为行动提供信息，而且可以指导探索信息需求的发现。我们进一步认为，意义建构是一个有价值的方法，探索上下文数据。根据我们的发现和对现有意义构建框架的反思，我们为意义构建工具提供了设计指南，以提高对影响患者的背景因素的认识，并支持患者机构理解健康数据。
Clinical data augmented with contextual data can help patients with chronic conditions make sense of their disease. However, existing tools do not support interpretation of multiple data streams. To better understand how individuals make sense of clinical and contextual data, we interviewed patients with Type 1 diabetes and their caregivers using context-enhanced visualizations of patients' data as probes to facilitate interpretation activities. We observed that our participants performed four analytical activities when interpreting their data -- finding context-based trends and explaining them, triangulating multiple factors, suggesting context-specific actions, and hypothesizing about alternate contextual factors affecting outcomes. We also observed two challenges encountered during analysis -- the inability to identify clear trends challenged action planning and counterintuitive insights compromised trust in data. Situating our findings within the existing sensemaking frameworks, we demonstrate that sensemaking can not only inform action but can guide the discovery of information needs for exploration. We further argue that sensemaking is a valuable approach for exploring contextual data. Informed by our findings and our reflection on existing sensemaking frameworks, we provide design guidelines for sensemaking tools to improve awareness of contextual factors affecting patients and to support patients' agency in making sense of health data.