Validation of Software for Assessment of Sleep Apnea From Data Acquired by a Wearable Smartwatch

• Conditions: Sleep Apnea
• Interventions: Diagnostic Test: Simultaneous assessment of sleep apnea by polysomnogram and wearable device

• Registration Number: NCT03775291
• Lead Sponsor: Fitbit LLC

Brief Summary

This clinical validation study aims to evaluate the utility of Fitbit's Sleep Apnea Alert software for minimally invasive monitoring of sleep apnea events to alert users of their risk of sleep apnea. The Sleep Apnea Alert software analyses data from a Fitbit commercially available wrist photoplethysmogram (PPG) device. The Sleep Apnea Alert software is an investigational software as medical device (SaMD) and is designed to retrospectively process data and flag users who have physiological signals consistent with an apnea-hypopnea index (AHI) of 15 or greater. This clinical validation study will be used to validate Fitbit's PPG-based sleep apnea algorithm. The outputs of the Fitbit Sleep Apnea Alert software will not be available to study participants, as the scope of this study is to gather validation data only and does not include testing product usability components.

Detailed Description

Not available

Study Timeline

• Study Start: 2018-11-09
• Study First Submitted: 2018-12-03
• Study First Submitted QC: 2018-12-11
• Study First Posted: 2018-12-13
• Status Verified: 2022-08
• Last Update Submitted: 2022-08-18
• Last Update Posted: 2022-08-22
• Primary Completion: 2022-12-31
• Study Completion: 2023-04-30

Recruitment & Eligibility

• Status: UNKNOWN
• Sex: All
• Target Recruitment: 303

Inclusion Criteria

• Capable of giving informed consent
• Comfortable reading the provided written study instructions in English (and/or local language) and using a smartphone in English (and/or local language).
• Have medical insurance either in the form of private insurance or a national health program.

Exclusion Criteria

• No implanted cardiac devices
• Not pregnant. Pregnancy status will be confirmed per patient report of medical history, i.e., patients will be asked if they are pregnant or may be pregnant, and date of last menstrual cycle.
• No medical conditions which might lead to consistently low levels of perfusion in the vasculature at the wrist

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Arm && Interventions

Group	Intervention	Description
Low pre-test likelihood for sleep apnea	Simultaneous assessment of sleep apnea by polysomnogram and wearable device	Subjects at low risk for having sleep apnea due to lack of known risk factors and no complaints of symptoms related to sleep apnea (e.g., excessive daytime sleepiness)
High pre-test likelihood for sleep apnea	Simultaneous assessment of sleep apnea by polysomnogram and wearable device	Suspected sleep apnea patients who are undergoing sleep tests as part of normal medical care or are also known to be non-compliant with therapy.

Primary Outcome Measures

Name	Time	Method
Comparison of PSG-assessed AHI with multi-night estimate derived from wearable device.	Nightly basis averaged over 7 nights.	Percent agreement between subjects classified as Apnea Hypopnea Index (AHI)≥15 or \<15 as determined by a single night of lab based polysomnogram , versus the corresponding estimate (averaged over 7 nights) from the wearable device. \[This primary outcome data will also be expressed as per-subject sensitivity and specificity of the wearable device to correctly identify subjects with AHI≥15 and AHI\<15 respectively\]

Secondary Outcome Measures

Name	Time	Method
Single night comparison between PSG and wearable device data on a per-subject basis	1 night	Percent agreement between subjects classified as Apnea Hypopnea Index (AHI)≥15 or \<15 as determined by a single night of lab based polysomnogram , versus the corresponding estimate on the same night as PSG from the wearable device. \[This primary outcome data will also be expressed as per-subject sensitivity and specificity of the wearable device to correctly identify subjects with AHI≥15 and AHI\<15 respectively\]
Single night comparison between PSG and wearable device data on a per-epoch basis	1 night	Percent agreement on a per one-minute epoch basis for the presence of sleep apnea between the polysomnogram and the corresponding simultaneous estimate from the wearable device .

Trial Locations

Locations (6)

St Vincent's University Hospital; 🇮🇪 Dublin, Ireland

Neurotrials Inc; 🇺🇸 Atlanta, Georgia, United States

Sleep Center Orange County; 🇺🇸 Irvine, California, United States

SleepMed Inc; 🇺🇸 Columbia, South Carolina, United States

Jasper Summit Research; 🇺🇸 Jasper, Alabama, United States

Peninsula Sleep Center; 🇺🇸 Burlingame, California, United States