VIDDE: Visualizations for Helping People with COPD Interpret Dyspnea During Exercise
Claudia Chen, Robert Wu, Hashim Khan, Khai Truong, Fanny Chevalier · 2021 · ASSETS '21: The 23rd International ACM SIGACCESS Conference on Computers and Accessibility · doi:10.1145/3441852.3471204
Summary
Chronic Obstructive Pulmonary Disease (COPD) affects approximately 10% of the adult population and is a leading cause of physical disability worldwide. People with COPD experience dyspnea (breathing discomfort) during physical exertion, which often triggers dyspnea-related distress and anxiety (DDA) — a fear that breathlessness signals something dangerously wrong. This creates a vicious cycle: anxiety leads to exercise avoidance, which reduces fitness, which increases dyspnea during activity. While pulmonary rehabilitation (PR) programs effectively address this through supervised exercise with biometric monitoring, access is severely limited — only 831 PR centers in the United States serve an estimated 24 million COPD patients. Chen et al. investigated whether presenting live physiological data (heart rate and blood oxygen saturation) to people with COPD during unsupervised exercise could provide similar reassurance. The research involved three phases: formative interviews with 15 COPD patients that identified five key insights about exercise anxiety and monitoring needs; the design and evaluation of VIDDE (Visualization for Interpreting Dyspnea During Exercise), a tablet application displaying real-time pulse oximeter data during pedal exercise through three case studies; and design probe interviews with 6 additional COPD patients exploring requirements for an exercise companion tool. VIDDE presented biometric data in two visualization modes — a simple current-value display with color-coded safety ranges, and a temporal display showing values over time — alongside a Modified Borg Scale for self-reporting dyspnea intensity and distress.
Key findings
The formative interviews revealed that COPD patients avoid exercise due to fear of dyspnea, seek ways to feel safer during exercise, need to learn pacing and when to rest using oximetry, and want real-time awareness of whether their oxygen saturation is within safe ranges. The three case studies demonstrated varied but largely positive effects of live biometric displays. Angela (69, recently diagnosed, high anxiety) showed reduced DDA ratings with the temporal data display (mean 1.5) compared to the baseline without biometrics (mean 3.4), finding the O2sat readings reassuring and using them as biofeedback to correct her breathing and posture. Brian (74, physically active, low anxiety) used the heart rate visualization as a target to maintain moderate exercise intensity, pushing himself harder each session to keep his heart rate in the "green zone." Charles (68, sedentary, severe dyspnea) discovered through the O2sat display that his oxygen was dropping to dangerous levels during exercise — something he had not previously been aware of — prompting him to stop, rest, and practice pursed lip breathing until levels recovered. The design probe interviews revealed important nuances: some participants worried that showing biometric data to newly diagnosed patients could increase anxiety, and the notification system for out-of-range values needed careful design — participants wanted calming sounds rather than alarming alerts. The researchers proposed a gradual approach to introducing physiological monitoring based on the patient's stage of disease acceptance and experience level.
Relevance
This paper makes an important contribution by expanding the accessibility field's focus beyond sensory and motor disabilities to include chronic respiratory disease — a condition that affects hundreds of millions worldwide and is the leading cause of disability globally. For accessibility practitioners, the key takeaway is that real-time biofeedback can help people with chronic conditions overcome psychological barriers to physical activity, not just physical ones. The finding that the same data visualization can be reassuring for one person and anxiety-inducing for another underscores the need for personalized, adaptive interfaces in health technology. The proposed gradual approach — starting with simple O2sat feedback for newly diagnosed patients and introducing more complex temporal visualizations as users gain experience and confidence — offers a design pattern applicable to many chronic condition management tools. The work also highlights how COVID-19 disproportionately affected research with medically vulnerable populations, with 70% of interested participants excluded due to safety protocols, demonstrating the ongoing challenge of inclusive research during health crises.
Tags: COPD · chronic illness · dyspnea · biofeedback · physiological data · visualization · exercise · pulse oximetry · anxiety · self-management