Activity Theory based evaluation of DMSS–R

The Dementia Management and Support System Research and Evaluation version (DMSSR), developed at the Department of Computing Science at the University of Ume̊a was evaluated using a group of medical students, an intern and an expert. The goal of the study was to, from a non-expert perspective, study how the system may be improved to offer better support for medical reasoning and as a learning tool, in addition to how the user interface may be improved for increased intuitiveness. Data was gathered in the form of observations of sessions with real and fabricated patient cases. These were supplemented with interviews and by asking the test subjects to “think aloud”. The data from the qualitative study was analysed using Activity Theory, supplemented by heuristic evaluation and cognitive load theory. Breakpoints were identified, and suggestions for remedying these are suggested. Due to low number of test subjects, the results of the study should be seen as indicative only. The main result of the study is that DMSS-R must offer explanatory feedback to promote learning and thus improved medical reasoning support for non-experts, in addition to several user interface changes. Additionally, the attitudes toward the system were studied, and the results show that in spite of not being convinced by the system-suggested diagnosis, the test subjects are overall very positive toward using the system and think that it may provide valuable support in the future.