Search Results

The Withings ECG App is a software-only device intended for use with the ScanWatch to create, record, store, transfer and display a single channel electrocardiogram (ECG) similar to a Lead I ECG. The Withings ECG App determines the presence of atrial fibrillation (AFib), sinus rhythm and high heart rate (no signs of AFib with heart rate 100-150 bpm) on a classifiable waveform. The Withings ECG App is not recommended for users with other known arrhythmias.

The Withings ECG App is intended for over-the-counter (OTC) use. The ECG data displayed by the Withings ECG App is intended for informational use only. The user is not intended to interpret or take clinical action based on the device output without consultation of a qualified healthcare professional. The ECG waveform is meant to supplement rhythm classification for the purposes of discriminating AFib from normal sinus rhythm and not intended to replace traditional methods of diagnosis or treatment.

The Withings ECG App is intended to supplement rhythm classification for the purposes of discriminating AFib from normal rhythms. The device is not intended to replace traditional methods or diagnosis.

The ECG acquired by ScanWatch is not intended for manual and/or automated measurement of QT-interval.

The Withings ECG app is not intended for use by people under 22 years old.

The Withings ECG App is a software only mobile medical application that has two components:

• Withings ECG Watch App
• Withings ECG Phone App

The Withings ECG Watch App is integrated on the Withings ScanWatch, model number hwa10. The Withings ECG Watch App analyzes the data collected by electrodes on the Withings ScanWatch to generate an ECG waveform similar to a Lead I, calculate the average heart rate and provide rhythm classification to the user for a given 30 second session.

Withings ECG Watch App consists of a software library called ECG-SW2 library. The ECG-SW2, is a software library that includes an algorithm that processes the raw ECG signals and a tracing filter that filters the ECG signal to provide the user an output on the user interface (watch and smartphone).

The Withings ECG Phone App contains the installation steps, tutorial and the instructions for use that the user must review prior to taking an ECG reading. The Withings ECG Phone App is included in the Withings App, which displays the ECG results and also allows the user to store, manage and share health data.

Here's a breakdown of the acceptance criteria and the study proving the device meets those criteria, based on the provided FDA 510(k) clearance letter for the Withings ECG App:

Overall Summary of the Study:

The Withings ECG App's performance was evaluated through a pivotal, prospective, multi-center clinical trial. This trial assessed the device's ability to accurately classify ECG recordings into Atrial Fibrillation (AFib), sinus rhythm, and high heart rate (no AFib) compared to a cardiologist's rhythm classification from a simultaneously collected 12-lead ECG (the ground truth). Non-clinical testing, including database testing of the algorithm and human factors studies, also supported the device's safety and effectiveness.

1. Table of Acceptance Criteria and Reported Device Performance

The clearance letter does not explicitly state pre-defined "acceptance criteria" as distinct numerical thresholds to be met. Instead, it describes the results achieved by the device, which implicitly serve as the demonstration of acceptable performance to the FDA. The performance metrics presented are sensitivity and specificity for AFib and sinus rhythm classification.

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size for Clinical Test Set: Approximately 626 subjects.
  • 219 in the Atrial Fibrillation cohort
  • 369 in the normal sinus rhythm cohort
  • 33 had other arrhythmias
  • 5 were uninterpretable
• Data Provenance:
  • Clinical Trial: Pivotal, prospective, multi-center clinical trial. The specific countries are not mentioned for the clinical trial itself, but the training data (Deep Train) is described as being from the European Union, and Heartbeats from the United States. Given the FDA clearance, it's highly probable the pivotal clinical trial included data from the U.S. or other regions acceptable to the FDA.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications

• Number of Experts: At least one cardiologist was used for ground truth establishment. The use of "a cardiologist" in the singular suggests a primary determination, though it's common practice for such a determination to be peer-reviewed or confirmed by a panel in high-stakes clinical trials. The document explicitly states: "Rhythm classification of a 12-lead ECG by a cardiologist was compared to the rhythm classification of a simultaneously collected ECG from the Withings ECG App."
• Qualifications of Experts: Cardiologist. Further details on experience (e.g., years of experience, board certification) are not specified in the provided text.

4. Adjudication Method for the Test Set

The adjudication method is implied as comparison to a cardiologist's 12-lead ECG interpretation. While it states "Rhythm classification of a 12-lead ECG by a cardiologist was compared...", it doesn't describe a multi-reader adjudication process (e.g., 2+1 or 3+1). It points to a direct comparison with a single cardiologist's interpretation of the 12-lead ECG as the reference standard.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done

No, an MRMC comparative effectiveness study involving human readers improving with AI vs. without AI assistance was not described in this document. The study focused on the device's standalone performance in classifying ECG rhythms against a clinical gold standard (cardiologist's 12-lead interpretation).

6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done

Yes, a standalone performance study was done. The core of the clinical trial directly assessed the "Withings ECG App's ability to accurately classify an ECG recording" which is its algorithmic performance in classifying AFib, sinus rhythm, and high heart rate. The sensitivity and specificity numbers (99.7% and 99.8% respectively) are metrics of the algorithm's performance in this standalone capacity relative to the ground truth.

7. The Type of Ground Truth Used

The primary ground truth used was expert consensus / clinical diagnosis from a 12-lead ECG. Specifically, "Rhythm classification of a 12-lead ECG by a cardiologist" served as the reference standard.

8. The Sample Size for the Training Set

The document mentions two "user datasets" used for training and testing (via 4-Fold Cross Validation for hyperparameter tuning):

• "Deep train": n = 11,701
• "Heartbeats": n = 5,089

It's important to note that these datasets were used for initial training and hyperparameter tuning (4-Fold cross validations and act as train and test sets). The "clinical study datasets" (HWA08 test, HWA08 CE, WEFA HWA09 part 1) were used as a "first layer of validation sets" to check generalization, and "WEFA HWA09 part 2" was a "second layer of validation after the software freeze". The pivotal clinical trial data (626 subjects) described under "Clinical Testing" was the independent, locked algorithm validation set for regulatory submission.

9. How the Ground Truth for the Training Set Was Established

The document provides the following for the "ML algorithm training and testing" datasets ("Deep train" and "Heartbeats"):

• It states these datasets were "user datasets".
• It does not explicitly detail how the ground truth for these training datasets was established. While the clinical trial's ground truth was a cardiologist's 12-lead ECG interpretation, the method for labeling the much larger training datasets is not provided in this excerpt. This is a common omission in 510(k) summaries which focus on the final validation. It's plausible they were labeled by a similar expert review process, potentially leveraging a larger pool of retrospectively acquired data.

Ask a specific question about this device