A pilot study from researchers at University College London and Moorfields Eye Hospital suggests that smartphones may offer a new way to monitor not only how patients with inherited retinal diseases (IRDs) function in daily life, but also how they are coping psychologically.
Published in npj Digital Medicine, the study evaluated the feasibility of collecting “digital phenotyping” data from patients with retinitis pigmentosa (RP) and choroideremia (CHM) using a bespoke iPhone application called OverSight. The platform combines passive data collection – including mobility and typing behavior – with patient-reported outcome measures to build a real-world picture of visual function and wellbeing.
The researchers enrolled 25 patients with genetically confirmed IRDs and followed them for 12 months. Participants used the OverSight app, which collected mobility metrics through Apple's HealthKit framework and keyboard-derived behavioral data through SensorKit. The app also administered validated questionnaires assessing quality of life, visual function, anxiety, depression, and disease-specific functional difficulties.
The primary objective was feasibility, and the results were encouraging. All participants successfully completed onboarding, while 92 percent remained in the study after one year. Among retained participants, 74 percent generated sufficient mobility data for analysis, and nearly all of those also provided usable keyboard data. Completion rates for patient-reported outcome measures exceeded 70 percent across all questionnaires.
Although the study was not powered to assess clinical outcomes, several intriguing associations emerged.
Participants generated a median of 6,087 steps per day and demonstrated a median walking speed of 1.18 meters per second. Typing behavior also appeared informative: older participants typed significantly slower than younger participants, a finding consistent with established evidence of age-related psychomotor slowing. Younger individuals were also more likely to use anxiety-related language during smartphone interactions.
Perhaps most unexpectedly, patients who reported greater difficulty with daytime peripheral vision used fewer anxiety-related and low-mood words in their typing. While this might initially suggest lower levels of distress, the study authors caution against such an interpretation, stating: “This pattern may reflect reduced expressive output rather than reduced emotional distress, given that individuals with poorer visual function typed less overall.”
The study findings highlight both the promise and complexity of passive behavioral monitoring. Rather than replacing conventional clinical assessments, digital phenotyping could provide a complementary view of patients' day-to-day experiences between clinic visits.
This may be particularly valuable in inherited retinal diseases, where visual acuity often remains relatively preserved despite substantial declines in peripheral vision and mobility. Traditional clinic-based measures can therefore miss important aspects of functional impairment and quality of life. By contrast, continuous smartphone-derived data may capture subtle behavioral changes as they occur in real-world settings.
The authors envision future systems capable of creating “closed feedback loops,” where behavioral changes detected through smartphone monitoring trigger timely support, self-management advice, or clinical review. They also suggest that digital phenotyping could eventually complement patient-centered endpoints in clinical trials, particularly as gene therapies and other advanced treatments for inherited retinal diseases continue to emerge.
