MyWeekInSight: Designing and Evaluating the Use of Visualization in Self-Management of Chronic Pain by Youth

TL;DR

This design study investigates using visualization of self-collected health data to support self-management of chronic pain in adolescents. Through MWIS, a mobile-friendly visualization dashboard co-developed with clinicians and a health-tech partner, the authors enable cross-facet reflection on sleep, symptoms, emotions, worries, school, and peers over a one-week EMA period. They document a seven-stage process, data abstractions, and design principles, and report two formal evaluations—design evaluation and deployed utility—across 6–11 participants, highlighting both supportive insights and the practical hurdles of real-world deployment. The work offers a structured deployment-and-reflection framework for digital health interventions in intersectionally vulnerable youth populations and proposes design principles and open questions for future adolescent-focused visualization tools.

Abstract

A teenager's experience of chronic pain reverberates through multiple interacting aspects of their lives. To self-manage their symptoms, they need to understand how factors such as their sleep, social interactions, emotions and pain intersect; supporting this capability must underlie an effective personalized healthcare solution. While adult use of personal informatics for self-management of various health factors has been studied, solutions intended for adults are rarely workable for teens, who face this complex and confusing situation with unique perspectives, skills and contexts. In this design study, we explore a means of facilitating self-reflection by youth living with chronic pain, through visualization of their personal health data. In collaboration with pediatric chronic pain clinicians and a health-tech industry partner, we designed and deployed MyWeekInSight, a visualization-based self-reflection tool for youth with chronic pain. We discuss our staged design approach with this intersectionally vulnerable population, in which we balanced reliance on proxy users and data with feedback from youth viewing their own data. We report on extensive formative and in-situ evaluation, including a three-week clinical deployment, and present a framework of challenges and barriers faced in clinical deployment with mitigations that can aid fellow researchers. Our reflections on the design process yield principles, surprises, and open questions.

Figures (5)

• Figure 1: MWIS's seven-stage design, deployment and evaluation process. *Clinical study design boerner_data_2022 and results boerner_making_2023 reported elsewhere.
• Figure 2: Visualization dashboard for deployed version (Deploy). Charts are vertically aligned, ordered and colour-coded by facet. The full scrollable dashboard has a very tall (1:5) aspect ratio; hence, the figure is split into two columns. This example combines data from multiple real participants, showing almost a full week of data. The black outlined frame is an annotation indicating the screen extent of a smartphone in landscape orientation.
• Figure 3: The final redesigned (HandOff) dashboard prototype, with changes from the Deploy version annotated. This example shows real participant data collaged into a mockup using Adobe Illustrator.
• Figure 4: Evolution of three visualization chart designs, across LoFi (sketched), MedFi (Balsamiq), HiFi (React.js), HiFiMod (also React, after the Usability Pilot Study), and Deploy (after the Preliminary Utility Study) versions. Rows show selected charts from the (a) Sleep, (b) Symptoms and (c) Worries facets. We initially prototyped visualizations for two-week EMA periods, but the deployed version shows only one week of data to support the finalized format of the clinical study.
• Figure 5: Challenges we faced, Barriers that might have caused them, and Mitigations that were or could have been used to overcome them, with their interconnections highlighted. Some Barriers and Mitigations relate to at least two Clinical or Youth population or Platform factors.