Validation of snoring detection using a smartphone app

Area of Science:

• Sleep Science
• Digital Health Technology

Background:

• Snoring is a common symptom linked to obstructive sleep apnea in adults.
• Smartphone technology enables convenient at-home snoring monitoring via dedicated apps.
• The diagnostic accuracy of these snoring detection applications remains largely unverified.

Purpose of the Study:

• To evaluate the snoring detection accuracy of Snore Clock, a commercially available smartphone application.
• To compare the app's performance against manual snoring rate assessments by a specialist.

Main Methods:

• Snoring data were collected using the Snore Clock app on smartphones.
• The app recorded sleep duration, snoring duration, loudness (dB), and snoring rate (%).
• An ENT specialist manually reviewed and validated the app's snoring detection results.

Main Results:

• Analysis of 201 snoring records from 11 patients showed a strong correlation (r=0.907) between app and manual detection.
• Snore Clock demonstrated a mean snoring detection accuracy of 95%.
• Positive predictive value was 65%±35%, negative predictive value 97%±4%, sensitivity 78%±25%, and specificity 97%±4%.

Conclusions:

• Snore Clock exhibits high compatibility across smartphone brands and strong predictive capabilities for snoring.
• The app's validated accuracy supports its utility for convenient at-home snoring detection.
• Findings suggest Snore Clock can be a reliable tool for initial snoring assessment.