Understanding the nature of information seeking behavior in critical care: implications for the design of health information technology

Objective: Information in critical care environments is distributed across multiple sources, such as paper charts, electronic records, and support personnel. For decision-making tasks, physicians have to seek, gather, filter and organize information from various sources in a timely manner. The objective of this research is to characterize the nature of physicians' information seeking process, and the content and structure of clinical information retrieved during this process.

Method: Eight medical intensive care unit physicians provided a verbal think-aloud as they performed a clinical diagnosis task. Verbal descriptions of physicians' activities, sources of information they used, time spent on each information source, and interactions with other clinicians were captured for analysis. The data were analyzed using qualitative and quantitative approaches.

Results: We found that the information seeking process was exploratory and iterative and driven by the contextual organization of information. While there was no significant differences between the overall time spent paper or electronic records, there was marginally greater relative information gain (i.e., more unique information retrieved per unit time) from electronic records (t(6)=1.89, p=0.1). Additionally, information retrieved from electronic records was at a higher level (i.e., observations and findings) in the knowledge structure than paper records, reflecting differences in the nature of knowledge utilization across resources.

Conclusion: A process of local optimization drove the information seeking process: physicians utilized information that maximized their information gain even though it required significantly more cognitive effort. Implications for the design of health information technology solutions that seamlessly integrate information seeking activities within the workflow, such as enriching the clinical information space and supporting efficient clinical reasoning and decision-making, are discussed.