Assessing the feasibility of using smartphone data to identify risk of idiopathic pulmonary arterial hypertension

TL;DR

This study explores using smartphone and wearable data to detect idiopathic pulmonary arterial hypertension earlier, showing promising results with digital health tools.

Contribution

The study introduces a novel approach using real-world physical activity and heart rate data for early detection of IPAH.

Findings

A classifier using activity and heart rate data achieved an ROC AUC of 0.87 for distinguishing IPAH patients from controls.

Adding in-app questionnaire data improved the classifier's performance to an ROC AUC of 0.94.

Wearable metrics correlated with clinical 6MWD, supporting their potential in traditional risk assessment.

Abstract

Idiopathic pulmonary arterial hypertension (IPAH) is a progressive, life-limiting condition often diagnosed late due to non-specific symptoms and requirement of invasive right heart catheterisation. This pilot study explores the feasibility of using real-world physical activity data from wearable devices and a smartphone app (My Heart Counts) to aid earlier detection. We analysed up to eight years of retrospective data from 109 UK participants, including patients with IPAH, disease controls, and healthy individuals. A classifier trained on pre-diagnostic activity and heart rate, distinguished individuals with IPAH from healthy and disease controls with an ROC AUC of 0.87, improving to 0.94 with in-app questionnaire input. Validation in a matched US cohort yielded an ROC AUC of 0.74. Wearable-derived metrics correlated with clinical 6MWD supporting their potential to complement traditional risk assessment. These pilot findings suggest that digital health tools may support earlier detection and remote monitoring of IPAH warranting larger scale studies.

Figures