A mobile application to support collection and analytics of real-time critical care data

Data collection, in high intensity environments, poses several challenges including the ability to observe multiple streams of information. These problems are especially evident in critical care, where monitoring of the Advanced Trauma Life Support (ATLS) protocol provides an excellent opportunity to study the efficacy of applications that allow for the rapid capture of event information, providing theoretically-driven feedback using the data. Our goal was, (a) to design and implement a way to capture data on deviation from the standard practice based on the theoretical foundation of error classification from our past research, (b) to provide a means to meaningfully visualize the collected data, and (c) to provide a proof-of-concept for this implementation, using some understanding of user experience in clinical practice.

We present the design and development of a web application designed to be used primarily on mobile devices and a summary data viewer to allow clinicians to, (a) track their activities, (b) provide real-time feedback of deviations from guidelines and protocols, and (c) provide summary feedback highlighting decisions made. We used a framework previously developed to classify activities in trauma as the theoretical foundation of the rules designed to do the same algorithmically, in our application. Attending physicians at a Level 1 trauma center used the application in the clinical setting and provided feedback for iterative development. Informal interviews and surveys were used to gain some deeper understanding of the user experience using this application in-situ.

Activity visualizations were created highlighting decisions made during a trauma code as well as classification of tasks per the theoretical framework. The attendings reviewed the efficacy of the data visualizations as part of their interviews. We also conducted a proof-of-concept evaluation by way of usability questionnaire. Two attendings rated 4 out of the usability 6 categories highly (inter-rater reliability: R = 0.87; weighted kappa = 0.59). This could be attributed to the fact that they were able to fit the use of the application into their regular workflow during a trauma code relatively seamlessly. A deeper evaluation is required to answer explain this further.

Our application can be used to capture and present data to provide an accurate reflection of work activities in real-time in complex critical care environments, without any significant interruptions to workflow.