(453 days)
The Cefaly device is indicated for the prophylactic treatment of episodic migraine in patients 18 years of age or older.
The Cefaly device is a transcutaneous electrical nerve stimulator (TENS) that is applied to the forehead (Fig. 1A) using a self-adhesive electrode positioned over the upper branches of the trigeminal nerve bilaterally (Fig. 1B). It is intended to stimulate the upper branches of the trigeminal nerve in order to reduce the frequency of migraine attacks.
The Cefaly device consists of two distinct components: an electrical pulse generator (EPG) and a self-adhesive electrode (Fig. 1). The Cefaly EPG (Fig. 1A) is made of ABS plastic and consists of electrical circuits controlled by firmware and powered by two 1.5V batteries. The front of the Cefaly EPG has a single button that is used to turn the device on/off and also to adjust the intensity of the electrical stimulus during a treatment session. Visual and auditory indicators inform the user when the device is on vs. off and help them troubleshoot if it is not working properly (e.g., device indicates if batteries need replacing and if electrical connection between device and skin is unacceptable). The back of the Cefaly EPG has two metal blades that serve to electrically connect it to the Cefaly electrode.
The Cefaly electrode (Fig. 1B) consists of a patient-contacting layer of gel, a layer of silvercoated carbon, and a layer of cotton. An ABS plastic pin in the middle of the electrode fits securely inside a receptacle in the back of the Cefaly EPG in order to maintain a secure electrical connection during a treatment session. The Cefaly electrode is meant to be used only with the Cefaly EPG, and vice versa. It can be reused a maximum of 20 times.
A treatment session begins by attaching the Cefaly electrode to the middle of the forehead and attaching the Cefaly EPG to the electrode. When the on/off button is depressed, a pulsatile electrical stimulus is applied for 20 minutes. During the first 14 minutes, the intensity of the stimulus gradually increases until it reaches a maximum. At any time while the stimulus intensity is increasing, the user can press the button on the front of the device to select an intensity that is lower than the maximum, and it will remain constant at this lower value for the remainder of the treatment session. The device turns the stimulus off automatically after 20 minutes, or alternatively, the user can stop a treatment session by pressing the button twice or simply removing the device from their forehead.
The Cefaly device is indicated for the prophylactic treatment of episodic migraine in patients 18 years of age or older. The acceptance criteria and the study that proves the device meets them are described below.
1. Table of Acceptance Criteria and Reported Device Performance
The Cefaly device's acceptance criteria are derived from the clinical performance data demonstrating its safety and effectiveness. The study primarily evaluated two co-primary endpoints and several secondary endpoints.
| Acceptance Criteria (Derived from Study 1 Results) | Reported Device Performance (Study 1) |
|---|---|
| Primary Endpoints: | |
| 1. Statistically significant responder rate (≥50% reduction in monthly migraine days) vs. sham. | 38.2% (Verum) vs. 12.1% (Sham) (p=0.023). Meets Criteria. |
| 2. Statistically significant reduction in monthly migraine days vs. sham. | Verum: -2.06 days; Sham: -0.32 days. Comparison between groups was not statistically significant (p=0.08 for ITT population). Does Not Fully Meet Criteria (for direct comparison). |
| Secondary Endpoints (demonstrating probable benefit): | |
| 1. Statistically significant reduction in monthly acute anti-migraine drug intake vs. sham. | Verum: 36.6% reduction; Sham: 0.5% reduction. Statistically significant (p=0.0072). Meets Criteria. |
| 2. Improvement in monthly migraine attack frequency. | Verum: 18.8% reduction; Sham: 3.5% reduction. (p=0.044 for ITT comparison). Meets Criteria. |
| 3. Improvement in total headache days. | Verum: 32.7% reduction; Sham: 4.1% reduction. (p=0.041 for ITT comparison). Meets Criteria. |
| 4. User satisfaction (at least moderately satisfied). | Verum: 70.6% (very or moderately satisfied); Sham: 39% (very or moderately satisfied). Meets Criteria. |
| Safety Endpoints: | |
| 1. No serious adverse events. | No serious adverse events reported. Meets Criteria. |
| 2. Minor, reversible adverse events. | One subject reported headache causing withdrawal. No adverse device events or unanticipated device reports. Meets Criteria. |
2. Sample Size and Data Provenance for the Test Set
Study Name: PREvention of MIgraine using the STS CEfaly (PREMICE study)
- Sample Size:
- Intent-to-treat (ITT) population: 67 subjects (34 in Verum group, 33 in Sham group).
- Per protocol population: 59 subjects (30 in Verum group, 29 in Sham group).
- Data Provenance: Prospective, multicenter, double-blinded, randomized, and sham-controlled trial conducted in 5 Belgian tertiary headache clinics.
3. Number of Experts and Qualifications for Ground Truth
The study relied on patient-reported outcomes collected through diaries, which served as the primary data for evaluating the efficacy of the device.
- Ground Truth Establishment: Clinical diagnosis of migraine and headache characteristics (occurrence, severity, aura, associated symptoms, and acute drug intake) were based on International Classification of Headache Disorders (ICHD-II) criteria. While not explicitly stated as "experts establishing ground truth for a test set" in the traditional sense of image interpretation or pathology, the ICHD-II criteria are established by a consensus of headache experts and applied by trained medical professionals (neurologists/headache specialists) within the participating tertiary headache clinics for patient inclusion, exclusion, and outcome assessment. The study was run by members of the Belgian Headache Society.
4. Adjudication Method for the Test Set
The study design itself acted as an adjudication method. It was a double-blinded, randomized, sham-controlled trial.
- Blinding: Both subjects and investigators were blinded to whether the device was delivering 'verum' (active) or 'sham' stimulation.
- Data Collection: Subjects filled in diaries to record headache occurrences and severity, and these diary entries formed the basis of the outcome measures.
- Statistical Analysis: Statistical analysis was carried out on an intent-to-treat basis, with missing data handled using the "last value carried forward" method.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not done. The study evaluated the device's effectiveness in patients directly, not by comparing human reader performance with and without AI assistance, as the Cefaly device is a therapeutic stimulation device, not a diagnostic imaging AI tool.
6. Standalone Performance Study
Yes, the PREMICE study (Study 1) can be considered a standalone performance study in the context of device efficacy. It directly measured the effectiveness of the Cefaly algorithm (the specific electrical stimulation parameters) acting on the patient's trigeminal nerve, without active human intervention during the treatment session. The "algorithm" here refers to the pre-programmed electrical output characteristics of the Cefaly device (pulse width, frequency, maximum intensity).
- Algorithm Only: The device delivers an electrical stimulus based on its pre-set parameters. The study compared the effects of the active device's electrical output (verum) against a "sham" device with different, less effective electrical output parameters. Therefore, the performance demonstrated is the standalone effect of the device's programmed electrical stimulation.
7. Type of Ground Truth Used
The ground truth used was expert diagnosis and patient-reported outcomes, adhering to established clinical criteria:
- Clinical Diagnosis: Migraine diagnosis (with or without aura) for subject inclusion and exclusion was based on International Classification of Headache Disorders (ICHD-II) criteria, applied by neurologists/headache specialists.
- Patient Diaries: Subjects' daily entries in diaries regarding headache occurrence, severity (4-point scale), presence of aura, associated symptoms, and acute antimigraine drug intake served as the direct "ground truth" for the outcome measures.
8. Sample Size for the Training Set
This question is not applicable in the typical sense for this device. The Cefaly device is a transcutaneous electrical nerve stimulator, not a machine learning algorithm that is "trained" on a dataset in the way an AI diagnostic tool would be.
- The electrical stimulation parameters (e.g., waveform, frequency, intensity) are pre-programmed based on physiological principles of nerve stimulation and prior research/development, not derived from a dedicated "training set" of patient data in an iterative learning process.
- The "training" data for the device's design would effectively be the physiological and neurological understanding that informed its engineering, as well as potentially earlier pilot studies or device iterations, but not a formally collected "training set" in the context of AI/ML.
9. How the Ground Truth for the Training Set Was Established
As noted above, this question is not directly applicable to a pre-programmed TENS device.
- The "ground truth" for the device's design and mechanism of action is based on neurophysiological understanding of trigeminal nerve stimulation and its role in migraine pathophysiology.
- The selection of specific electrical parameters (pulse width, frequency, intensity) would be informed by:
- Literature review on effective neuromodulation techniques.
- Pre-clinical studies (e.g., in vitro, animal models).
- Pilot human studies to determine tolerable and potentially efficacious stimulation parameters.
- Engineering principles to ensure safe and effective energy delivery.
However, these are part of the device's development process, not a "training set" with established ground truth as understood in AI/ML validation.
§ 882.5891 Transcutaneous electrical nerve stimulator to treat headache.
(a)
Identification. A transcutaneous electrical nerve stimulator to treat headache is a device used to apply an electrical current to a patient's cranium through electrodes placed on the skin.(b)
Classification. Class II (special controls). The special controls for this device are:(1) The patient-contacting components of the device must be demonstrated to be biocompatible.
(2) Appropriate analysis/testing must validate electromagnetic compatibility and electrical, mechanical, and thermal safety.
(3) The technical parameters of the device, including waveform, output modes, maximum output voltage and current (with 500, 2,000, and 10,000 ohm loads), pulse duration, frequency, net charge (µC) per pulse, maximum phase charge at 500 ohms, maximum current density (mA/cm
2 , r.m.s.), maximum average current (mA), maximum average power density (W/cm2 ), and the type of impedance monitoring system must be fully characterized.(4) Electrical performance, adhesive integrity, shelf life, reusability, and current distribution testing of the electrodes must be conducted.
(5) Appropriate software verification, validation, and hazard analysis must be performed.
(6) Clinical performance data must demonstrate that the device is safe and effective as a treatment for headache in the indicated patient population.
(7) Labeling must include the following:
(i) Appropriate contraindications such as not for use in subjects with an implanted metallic or electronic device in the head, a cardiac pacemaker, or an implanted or wearable defibrillator.
(ii) Appropriate warnings such as not to apply the device on the neck or chest, not to use the device in the presence of electronic monitoring equipment, not to use in the bath or shower, not to use while sleeping, not to use while driving, not to use while operating machinery.
(iii) Appropriate precautions such as the long-term effects of chronic use of the device are unknown.
(iv) A summary of the expected risks and benefits of using the device.
(v) A summary of the clinical performance data, including information on the patient population for which the device has and has not been demonstrated to be effective, and any adverse events and complications.
(vi) Information on how the device operates and the typical sensations experienced during treatment.
(vii) A detailed summary of the device technical parameters.
(viii) An expiration date/shelf life for the electrodes and the number of times they can be reused.
(ix) Disposal instructions.