| K-number | K242967 |
| Device name | Loss of Pulse Detection |
| Applicant | Fitbit |
| Product code | SDY |
| Device class | Class II |
| Decision date | Feb 25, 2025 |
| Decision | Substantially Equivalent |
| Regulation | — |
Loss of Pulse Detection is a software-only mobile application that analyzes photoplethysmography (PPG) data from wrist-worn consumer devices to identify loss of pulse events and alert the user via audio, visual, and haptic notifications. If the user remains unresponsive, the app attempts to initiate an emergency call through a connected smartphone or smartwatch. It is intended for over-the-counter use in people 22 years and older and is not intended to replace traditional diagnosis or monitoring methods.
The subject device uses a convolutional neural network algorithm analyzing PPG and accelerometer sensor data to detect loss of pulse. It comprises a software component on the smartwatch and a mobile app component on the user's smartphone. The predicate (Fitbit Irregular Rhythm Notifications) similarly analyzes PPG data using signal processing algorithms but detects atrial fibrillation instead. Both devices operate passively in the background, require user stillness for analysis, and provide opportunistic notifications when conditions are met.
Not stated in this summary.
Both devices are software-only photoplethysmograph analysis tools for over-the-counter use with compatible wrist-worn products, intended for users over 22 years old with pre-existing cardiac condition exclusions. While the subject device detects loss of pulse and the predicate detects atrial fibrillation, both employ similar PPG signal processing architectures, have comparable limiting factors (motion, tattoos, inadequate blood flow), are non-continuous screening tools, and provide user notifications without diagnosing. The algorithmic differences for detecting different cardiac conditions do not raise new safety or effectiveness questions. Clinical testing demonstrated 69.3% sensitivity and 99.965% specificity with no device-related adverse events, supporting equivalence to the predicate's performance profile.
View the full FDA submission: accessdata.fda.gov