Acoustic biomarkers in asthma monitoring

Introduction: Digital biomarkers are collected by digital devices for long-term, continuous disease monitoring. It is unclear if acoustic biomarkers (cough counting, breath sounds) can be used to reliably monitor asthma control.

Aim: This systematic review aims to evaluate whether sufficient evidence of validity and reliability exists to support the use of acoustic biomarkers in the diagnosis and monitoring of patients with asthma. 

Methods: A search of PubMed, Embase and Scopus was conducted for articles published between 1990-2021 for studies reporting quantitative outcomes of acoustic digital biomarkers in asthma. Studies were included if the audio signal analysis was compared to a clinically validated asthma questionnaire, a lab biomarker or human algorithm count annotation. 

Results: 1215 studies were identified, 30 screened for eligibility and 17 were found eligible. A total of 720 178 signals from 1409 patients (470 asthmatic patients and 939 healthy volunteers) were observed. 12 studies were hospital-based studies, 3 were based on real-world recordings, 2 were outpatient-based. Signals were recorded using custom devices (n=10), smartphones (n=5), microphones (n=2) and smartwatches (n=1). The average sensitivity and specificity to identify asthma control were 89.3% [range 54.0 - 100.0%] and 87.4% [67.0 - 100.0] respectively. The average sensitivity and specificity to identify an acute asthma exacerbation were 87.0% [63.0 - 100.0%] and 85.4% [48.0 - 100.0%] respectively.

Conclusion: Acoustic biomarkers show promise in the monitoring of asthma control. Prospective studies are needed to ascertain if monitoring of acoustic biomarkers leads to reduced exacerbations.