Fitbit ECG App Doctor Guide

Understanding the Fitbit ECG App

United States:

Fitbit ECG is a software-only mobile medical application for Fitbit wearable products that creates, stores, transmits, and displays a single-lead electrocardiogram (ECG) similar to a single-lead ECG. The Fitbit ECG app determines if atrial fibrillation (AFib) or a normal sinus rhythm is present based on the classifiable waveform. It is not recommended for individuals with other known arrhythmias to use the AFib detection feature.

The Fitbit ECG app is intended for over-the-counter (OTC) use. The ECG data displayed by the app is for informational purposes only. Users should not interpret or implement any clinical action based solely on this product without consulting a qualified healthcare professional. The ECG waveform is intended to supplement heart rhythm classification to distinguish between AFib and normal sinus rhythm. It is not a substitute for traditional diagnostic or therapeutic measures. The Fitbit ECG app is not intended for use by individuals under 22 years of age.

Non-U.S. Regions* (EU/EEA/UK):

The Fitbit ECG app is intended to support users in recording waveforms similar to a single-lead ECG and then classifying them as either a sinus rhythm or atrial fibrillation (AFib) based on the classifiable waveform. It is not recommended for individuals with other known arrhythmias to use the ECG and heart rhythm classification. The Fitbit ECG app is intended for home use by individuals aged 22 years and older.

Usage Instructions for Non-U.S. Regions Apply to the Following Countries:

The Fitbit ECG app will be available in the following EU countries: Germany, Ireland, Spain, France, Netherlands, Sweden, Italy, Belgium, Poland, Austria, Czech Republic, Portugal, Romania, and Luxembourg. This Doctor Guide has local language versions for the countries listed above on help.fitbit.com.

The Fitbit ECG app will be available in the UK and other non-EU countries, including Switzerland. This Doctor Guide has local language versions for the countries listed above on help.fitbit.com.

Additional countries will be added as regulatory approvals are obtained.

How Data is Collected

The electronic sensors are located on the metal frame around the back and surface of the Fitbit wristband product. When these sensors are in contact with the user's skin and the Fitbit ECG app is open, they record 30 seconds of the user's heart's electrical signals. A proprietary algorithm determines if the reading indicates a normal sinus rhythm or an irregular rhythm suggestive of atrial fibrillation (AFib). Users can take multiple readings as needed.

Interpreting Results

After completing an ECG reading, users can download a PDF summary report of their ECG results, including the ECG waveform, from the mobile app. Users must use the latest version of Adobe Acrobat Reader to open this PDF report. The results displayed to the user may be:

Clinical Study

A clinical study was conducted to validate the ability of the Fitbit ECG app software algorithm to detect AFib using data from Fitbit wearable products. The primary objective was to validate the ability of the Fitbit ECG app software algorithm to classify heart rhythms as AFib or normal sinus rhythm (NSR) using single-lead ECG equivalent waveforms. The performance target for AFib detection was a lower bound of the confidence interval of 90% for sensitivity with 97.5% confidence, and 92% for specificity with 97.5% confidence. The secondary objective was to confirm the software's ability to generate single-lead clinical equivalent waveforms to 12-lead ECGs from at least 80% (95% confidence) of paired tracings, assessed both qualitatively and quantitatively.

From November 11, 2019, to December 20, 2019, subjects with and without a history of known AFib were recruited at 9 sites across the United States to participate in the study. Upon enrollment, a 10-second 12-lead screening ECG was recorded to confirm subject cohort assignment (AFib or NSR). Subjects were deemed screen failures if they had a history of known AFib but their ECG did not show AFib, or if they had no history of known atrial arrhythmias but showed atrial arrhythmias. Screening ECGs were interpreted by a qualified physician at each site. Subsequently, subjects underwent simultaneous 30-second 12-lead ECG and Fitbit ECG testing. Physicians at a central core laboratory adjudicated the Fitbit ECG and 12-lead ECG tracings. Data classified as SR or AF by the Fitbit software algorithm and physician interpretation of the 12-lead ECG were used for primary endpoint sensitivity and specificity calculations. For secondary endpoint analysis, a subset of data (70 cases each of AF and SR) was randomly selected by core laboratory physicians for qualitative and quantitative similarity assessment. The proportion of tracings that were similar between the Fitbit ECG and 12-lead ECG, and the proportion with R-wave amplitudes within 2mm (0.2mV) were calculated.

A total of 472 subjects were recruited for the study, of which 440 met the eligibility criteria. Thirty subjects failed the screening ECG and were excluded from the study, and two subjects withdrew their consent to participate. Of the 440 eligible subjects, 265 were in the SR cohort and 175 were in the AF cohort. Due to protocol deviations, two subjects were excluded from each cohort, so the analysis included 263 subjects from the SR cohort and 173 subjects from the AF cohort. No adverse events were reported. The Fitbit ECG spot-check software algorithm's ability to detect AF was 98.7% (LCB 95.4%) for sensitivity and 100% (LCB 98.5%) for specificity. The Fitbit ECG single-lead waveforms were deemed qualitatively similar to the single-lead waveforms of the 12-lead ECGs overall by 95.0% (LCB 90.0%). The Fitbit ECG R-wave amplitude was on average larger than the reference 12-lead ECG R-wave peaks by 2.4mm (SD 1.4mm) / 0.24mV (SD 0.14mV). Although this data was slightly larger than the proposed acceptance criteria, the clinical conclusions drawn by clinicians from the Fitbit app ECG waveforms were consistent with those from the 12-lead ECG waveforms, indicating that the Fitbit app ECG waveforms can meet the intended use for physician review.

The primary endpoint, sensitivity and specificity, exceeded the target performance. The qualitative assessment for the secondary endpoint also exceeded the target performance. While the R-wave amplitude secondary endpoint did not meet the target performance, this may be related to the lead placement of the ECG reference device.

User Help Information

This document applies to SKU: FB 603, Version 2.0 or higher.