| Clinicians depend on the ability to efficiently locate patient-specific information when performing their clinical tasks. Clinical information systems are playing an increasingly important role for supporting the patient care process and seek to facilitate access to needed information. Knowledge of potential users can provide insight on how to effectively organize and present information in these systems. This research explores discovery of clinician information needs from log files, and demonstrates how this knowledge can be used to anticipate the information needs of clinicians in clinical information systems.;In this dissertation, I introduce an automated quantitative method for acquiring knowledge of clinicians' information needs in order to enhance a wireless handheld clinical application. I developed and applied Clinical Information System (CIS) log analysis and mining to obtain statistics and to discover patterns from CIS log files. The results of this analysis were used to develop a knowledge base of patient-specific clinician information needs. I have used this knowledge base to create context-sensitive "shortcuts" in PalmCIS, an application that provides clinicians with access to patient data via a wireless handheld device. These shortcuts seek to anticipate the information needs of users and provide direct links to those data. An evaluation was performed to determine the usability and impact of shortcuts in PalmCIS, as well as to assess the validity of CIS log analysis as a method for identifying patient-specific clinician information needs. Results indicate that shortcuts are usable and facilitate access to information, and that the usefulness of a shortcut may be dependent on the context (e.g., user role, patient problems, and task).;A combination of qualitative methods and CIS log analysis may prove to be valuable for building comprehensive models of clinician information needs in different contexts. These models are particularly useful for development of wireless handheld clinical applications, which aim to efficiently present information to the user. By adapting content to the context, shortcuts in wireless handheld clinical applications have the potential to facilitate the task at hand. |
