Improving the user experience of blended care tools in clinical practice

Mental health problems cause a major burden on individuals and society, making it crucial to understand these problems as they unfold in daily life. The Experience Sampling Method (ESM) is a valuable research method for capturing people’s thoughts, feelings, and behaviors in real time, offering the potential to improve psychotherapeutic treatment. However, existing ESM tools are mainly research-focused and often not optimized for clinical use. This PhD thesis aimed to bridge the gap between research and practice by developing an ESM tool that is both scientifically grounded and user-friendly for mental health practitioners.

In Chapter 1, we first charted the perspectives of mental health practitioners and ESM researchers using a survey study on the utility of ESM for mental health care. Both recognized ESM's usefulness across treatment phases, such as for diagnosis, intervention, and relapse prevention. In particular, ESM was most useful for practitioners in capturing context-specific symptoms. Personalizing ESM diaries was seen as crucial, yet both groups indicated the need to minimize the effort. This study provided direct input for developing a personalized, user-friendly ESM dashboard.

In Chapter 2, we developed a fully functional proof of concept for a user-friendly clinical dashboard, allowing practitioners to cocreate personalized diaries with their patients. We presented the main dashboard features and visual feedback possibilities, all requiring no coding skills. During the development process, principles of user-centered design approach were used and improvements were tested with end-users.

Next, we evaluated the extent to which practitioners could interpret and make clinical decisions based on visual feedback, using statistically relevant methods to visualize uncertainty such as error-bars and textual descriptions about effect size and confidence intervals. Therefore, Chapter 3 presented an experiment involving 40 practitioners who evaluated these visualizations. The results indicated that providing textual descriptions about effect size and confidence intervals increased practitioners’ confidence levels and accuracies compared to error bars alone. To effectively communicate uncertainty in visualizations to practitioners, a brief textual description is recommended.

Finally, in Chapter 4, we aimed to understand the actual spontaneous use of our newly clinical dashboard, including practitioners’ experiences, workflows, and suggestions for improvements. To this end, we did an interview study with 19 practitioners who had already integrated the ESM dashboard into their clinical practice. Using content analysis, we found that practitioners primarily used ESM during diagnosis to confirm hypotheses, with fewer use for real-time interventions, such as sending exercises or coping tips. Despite recognizing ESM’s potential, 31.6% had not yet adopted it for real-time interventions goals. Practitioners highlighted the need for peer learning opportunities and enhanced support documentation for broader adoption.

Overall, this PhD project has contributed to the development of a personalized ESM tool that meets clinical needs. By addressing the perspectives of practitioners and ESM researchers, designing a functional dashboard, and evaluating its real-world applicability, this thesis brings evidence-based ESM closer to actual clinical practice, and as such clear the way for responsible market adoption.