Withings Scan Monitor 2.0

The Withings Scan Monitor 2.0 is a smart scale system that provides a 6-lead electrocardiogram by recording two channels (Lead II and III) via finger and feet electrodes, calculating additional leads, and utilizing the ECG-SW1 algorithm to detect atrial fibrillation and classify heart rhythms for adult users. Its methodology involves a 30-second user-interactive recording, with clinical studies demonstrating high accuracy for rhythm classification and ECG parameter equivalence to standard 12-lead ECGs, albeit with considerations for noise and inconclusive readings.

The Withings Scan Monitor 2.0 is a connected smart scale equipped with a retractable handle, designed for adult use by healthcare professionals, individuals with known or suspected heart conditions, and health-conscious users in various settings including hospitals, clinics, long-term care facilities, and homes. Its primary function is to record, store, and transfer a two-channel electrocardiogram (ECG), specifically lead II and lead-III, from which it also calculates and displays lead I, aVR, aVF, and aVL. The device achieves this when a user stands on the scale and holds the handle, initiating a 30-second ECG recording via electrical sensors. This data is then analyzed by the device's software to measure heart rate and classify the ECG rhythm into categories such as Normal Sinus Rhythm, Atrial Fibrillation (when prescribed or used under a physician's care), Inconclusive, or Noise-Inconclusive, with results displayed on the scale and managed through a companion mobile application. The device employs three ECG measuring electrodes (contacting fingers and feet) and two feedback electrodes to mitigate environmental electromagnetic interference, operating with a 500Hz sampling rate. The core analysis relies on the ECG-SW1 algorithm. Clinical validation, involving 274 adult subjects (234 analyzed), demonstrated its ability to detect atrial fibrillation and normal sinus rhythm with a sensitivity of 0.99 and specificity of 0.99 when excluding inconclusive and noise-affected recordings; overall sensitivity for AFib was 73.3% and specificity for Normal Sinus Rhythm was 95.5% when including all classifications from 321 recordings. The study also confirmed the 6-lead ECG's clinical equivalence to a standard 12-lead ECG in terms of wave visibility (P, QRS, T waves accuracy >0.992), polarity accuracy (P-wave 0.986, QRS 0.948, T-wave 0.911), and interval durations (PR/QRS difference std dev <20ms, QT difference std dev 30.3ms), alongside accurate heart rate measurement (mean difference -0.96 bpm, std dev 3.81 bpm).

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. Given the 7.7% "noise" classification rate and the "Inconclusive" category, what are the primary contributors to these uninterpretable recordings, and what strategies could be implemented in future iterations or user guidance to minimize them?; How does the performance of the ECG-SW1 algorithm in the Withings Scan Monitor 2.0 (using finger and feet contacts) compare to its performance in the predicate device (ScanWatch, using wrist/finger contacts), particularly regarding signal quality from different anatomical sites and body positions (standing vs. seated)?; The clinical study excluded 40 out of 274 patients from data analysis; what were the reasons for these exclusions and how might they impact the generalizability of the reported sensitivity and specificity?; While overall AFib/NSR sensitivity and specificity are high when excluding inconclusive/noise, how does the device's performance in real-world, unsupervised home use, potentially with a higher prevalence of suboptimal recording conditions, compare to the clinical study results?; Considering the QT interval standard deviation of differences was 30.3ms, what is the clinical significance of this variability for QT assessment, especially when compared to the <20ms for PR/QRS, and are there specific patient populations for whom this level of variability might be a concern?