Doctor of Philosophy (PhD)
The University of Texas School of Biomedical Informatics at Houston
Dr. Jiajie Zhang
Electronic health information brokering systems are of interest to public health informatics because they emphasize how data can be effectively shared and utilized across healthcare institutions and among providers so as to improve the quality of care, increase efficiency, and reduce costs (Lumpkin, 2002). In the domain of public health (PH) specifically, where complete and timely reporting of data is critical for all epidemiological and disease surveillance activities (Langmuir, 1976), it is imperative to ensure proper functioning of the electronic information exchange infrastructure. Receiving multiple types of data, in various formats from numerous sources, and triaging them to the appropriate surveillance system is no easy task for a department of health, whether at state, local or federal level (Magnuson, 2005).
The administrators of the electronic message brokering system, and the coordinators of surveillance systems in each public health jurisdiction, are responsible for ensuring that the data is received, archived, validated and triaged appropriately in a timely and complete fashion. This requires continuous monitoring of trends in messaging and system performance and active responses to aberrations. To achieve this, administrators depend heavily on dashboards to provide awareness of exchange system status and its reporting at any point of time. Unfortunately, current dashboards do not offer the context or cognitive support needed for interpreting the information presented. As research has demonstrated in other domains, in order to make sense of the data and react, dashboard users are required to draw upon domain knowledge, higher level association between domains, operational rules, organizational missions, personal objectives, tasks at hand, priorities, past experiences, historic events, recent events, psychosocial and political constructs, and more (Resnick, 2005; Mirhaji, Srinivasan, Casscells, & Arafat, 2004). The burden of ‘interpretation’ always falls on the cognitive system of the human operator, which is prone to error and malfunctioning when risk and emergency overwhelm psychological factors (Parsa, Richesson, Smith, Zhang, & Srinivasan, 2004; Parsa, Zhang, Smith, Majid, Casscells, & Lillibridge, 2003). On the basis of the surveillance literature it can be seen that meaningful and holistic interpretation of data requires the generation of higher-level explanations based on knowledge and expertise from numerous principles (Parsa, Richesson, & Srinivasan, 2004; Parsa, Richesson, Smith, Zhang, & Srinivasan, 2004), while context is essential to illustrate the ‘big picture’ view of dynamic and complex problems (Parsa, Zhang, Smith, Majid, Casscells, & Lillibridge, 2003). These reservations imply that the process for building health information dashboards should consider not only user functions, tasks and goals but also the user’s situational awareness (SA) requirements. This vision adds a new layer to information representation that needs to be accounted for when conceptualizing the implementation of health information dashboards. A review of the literature reveals a lack of methods to design for situational awareness in dashboard systems in complex domains (Resnick, 2005; Li, 2007).
This research introduces a new method to present contextualized information that can improve user SA. I present the design rationale, method, and results of an evaluation study that measures the situational awareness generated by adopting this new context-driven representation model.
Srinivasan, Arunkumar, "A Method for Representing Contextualized Information (MeRCI) to Improve Situational Awareness Among Electronic Message Brokering System Dashboard Users" (2011). UT SBMI Dissertations (Open Access). 29.
MeRCI, situational awareness, medical informatics, electronic health records, health information systems