Search Filters

Search Results

Found 1 results

510(k) Data Aggregation

    K Number
    K221698
    Manufacturer
    Date Cleared
    2022-12-21

    (191 days)

    Product Code
    Regulation Number
    874.3325
    Reference & Predicate Devices
    Predicate For
    N/A
    AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
    Intended Use

    The Eargo Self-Fitting Hearing Aids are intended to amplify and transmit sound to the ear and thereby compensate for perceived mild to moderate hearing impairment in individuals 18 years of age or older. They are adjusted by the user to meet the user's hearing needs. No pre-programming or hearing test is necessary. The product is intended to be used without the assistance of a hearing care professional.

    Device Description

    The Eargo Self-Fitting Hearing Aid is a self-fitting air-conduction hearing aid system that incorporates wireless technology in its programming and use. The hearing aid system consists of a pair of earbud-style hearing aids (left and right), a charging case, and a companion mobile application (app) available for iOS (version 12 or later) and Android (version 7 or later) mobile devices. The hearing aids are designed to be virtually invisible, inserted completely and discreetly within the ear canal. Each hearing aid contains a microphone to allow for audio input, which is amplified by the hearing aid. The mobile app facilitates Eargo's proprietary self-fitting process using a combination of proprietary ultrasonic (for fitting) and Bluetooth Low Energy (BLE; for programming fitting settings) wireless communication. The mobile app also allows the user to control the hearing aids using proprietary ultrasonic wireless communication and enables firmware updates to the hearing aid system via BLE. App-based user controls include program and settings changes. In addition, each hearing aid contains an accelerometer sensor that allows for ondevice user control of the hearing aids. On-device user controls allow the user to make program changes without the mobile app. Each hearing aid contains a rechargeable Li-ion battery and is charged by the charging case that also functions as a carrying case. The charging case contains a single-cell Li-ion rechargeable battery, which charges the hearing aids via wireless (near-field inductive) charging when the hearing aids are correctly placed into the charging case.

    AI/ML Overview

    Here's a breakdown of the acceptance criteria and the study details for the Eargo Self-Fitting Hearing Aids, based on the provided text:

    Acceptance Criteria and Device Performance

    The document doesn't explicitly present a single "Table of Acceptance Criteria" with numerical targets and direct "Reported Device Performance" for each criterion in the same way one might find for a pass/fail mechanical test. Instead, the acceptance criteria are embedded within descriptions of various tests, and the "reported device performance" is the successful conclusion (e.g., "Pass," "Comparable to predicate," "Acceptable usability and satisfaction").

    However, I can extract the implicit acceptance criteria and the demonstrated performance where quantitative or qualitative measures are reported:

    Acceptance Criteria (Implicit)Reported Device Performance
    Safety and General Performance (Compliance with Standards)
    FCC Title 47 CFR, subpart 15 subpart C (Intentional Radiators)Pass
    IEC 60601-1:2014 (Basic Safety & Essential Performance)Pass
    IEC 60601-2-66:2019 (Basic Safety & Essential Performance for Hearing Instruments)Pass
    IEC 60601-1-2:2014 (Electromagnetic Compatibility - EMC)Pass
    IEC 60118-13:2019 (Hearing Aid EMC)Pass
    ANSI ASA S3.22:2014 (Hearing Aid Characteristics - Electroacoustics)Pass
    IEC 62304:2006 (Software Lifecycle Processes) & FDA Guidance (Software Content of Premarket Submissions)Pass (Software Verification and Validation)
    ISO 10993-1, -5, -10, -12 (Biocompatibility)Pass
    IEC 62366:2007 + A1:2014 & IEC 60601-1-6:2010 + A1:2013 & FDA Guidance (Summative Usability / Human Factors Validation)Pass
    FDA Guidance (Cybersecurity Compliance)Pass
    ASTM D4169-16 & D7386-16 (Package & Transit Testing)Pass
    Electroacoustic Performance (Table 3)
    MAX OSPL 90 (less than 117 dB SPL per OTC HA requirements)106 dB SPL (less than 117 dB SPL)
    HFA OSPL 90 (adequate for fitting mild-moderate HL as prescribed by NAL-NL2)104 dBSPL (Comparable to predicate, adequate for NAL-NL2)
    HFA FOG (adequate for fitting mild-moderate HL as prescribed by NAL-NL2)26 dB (Comparable to predicate, adequate for NAL-NL2)
    Reference Test Gain (RTG) (adequate for fitting mild-moderate HL as prescribed by NAL-NL2)26 dB (Comparable to predicate, adequate for NAL-NL2)
    Frequency Range (suitable for intended use per OTC HA requirements)<200 - 7500 Hz (Comparable to predicate and suitable)
    Harmonic Distortion (%) (meets OTC HA requirements)< 1% (Same as predicate, meets OTC HA requirements)
    EIN (less than or equal to 32 dBSPL, comparable to predicate, meets OTC HA requirements)< 30 dBSPL (Comparable to predicate, meets OTC HA requirements)
    Latency (less than or equal to 15ms per OTC HA requirements)5.7ms (Comparable to predicate, meets OTC HA requirements)
    Clinical Performance (Self-Fitting Accuracy and Effectiveness)
    Accuracy of Sound Match Hearing Thresholds (Statistical non-difference to Audiology Best Practice)Pairwise t-test showed no statistically significant differences between EargoA (Sound Match in sound-treated booth) thresholds and clinical thresholds (ABP) at all frequencies. Test-retest reliability between EargoA and EargoB (Sound Match in quiet room) also showed no statistically significant differences. Mean SUS score of 71 (above industry benchmark of 68).
    Effectiveness of Self-Fitting (Non-inferiority to Audiologist-Fit) in key outcome measures
    Real Ear Aided Response (REAR) (within 5 dB RMSE of NAL-NL2 targets at 250 Hz, 1 kHz, 2 kHz, 4 kHz, and no significant difference between self-fit and AUD-fit)Average RMSE from NAL-NL2 targets were below 5 dB for both fitting conditions. No significant differences noted between real-ear gain measured between self-fit and AUD-fit settings (Figure 4).
    Abbreviated Profile of Hearing Aid Benefit (APHAB) (self-fit not inferior to AUD-fit; improved over unaided)Very similar aided mean and standard deviation values for self-fit and AUD-fit across APHAB subscales and overall APHAB-global score. Mean scores significantly improved over the unaided condition for both self-fit and AUD-fit on EC, BN, RV subscales and global score (Figure 5).
    Speech Intelligibility in Noise (self-fit similar to AUD-fit)Very similar mean and standard deviation values when comparing aided speech in noise scores for self-fit and AUD-fit (Figure 6).
    Subjective Sound Quality Ratings (self-fit similar to AUD-fit; acceptable in real-world usage)Very similar mean and standard deviation scores when comparing overall sound quality ratings for self-fit and AUD-fit (Figure 7). Data from 255 survey respondents (real-world users) indicate that the device's subjective sound quality was acceptable.
    Usability and Human Factors
    Real-World Fit and Comfort (acceptable ratings)88% rated "Good" or "Excellent" for fit; 94% rated "Never" or "1-2 times daily" for migration; 94% rated "Mild" or "No Discomfort" for discomfort (Table 4).
    Real-World Safety (no observed device-related adverse events)No health-related issues due to device fit or discomfort were observed, and no device failures leading to potential safety issues were observed during unsupervised real-world device usage (ranging from 41 to 210 days).
    Real-World Satisfaction and Usability (acceptable across assessed areas, including Sound Match, adjustments, sound quality, perceived benefit)Responses from 255 subjects demonstrated acceptable usability and satisfaction in all assessed areas (Sound Match completion, device maintenance/cleaning, subjective sound quality, app-based adjustments, perceived benefit).
    Human Factors Validation of Self-Fit Strategy (critical tasks completed successfully by at least 80% with no use errors)All four use case scenarios (Completing Sound Match, Using mobile app to change programs/settings, Making in-situ program changes without mobile app, Reverting settings back to factory defaults) met the acceptance criteria of being completed successfully by at least 80% of participants. No use errors (failure or inability to complete) were observed for any tasks.
    Human Factors Validation of Device Labeling and Handling/Maintenance (critical tasks completed successfully and independently by at least 80%)All six use case scenarios (Understanding outside/inside package labeling, Device charging, Eartip self-selection/replacement/insertion/removal, Mic cap replacement, Device cleaning) met the acceptance criteria of being completed successfully and independently by at least 80% of participants. A task was considered a failure if assistance was needed.

    Study Details:

    2. Sample sizes used for the test set and the data provenance

    • Clinical Validation of Eargo's Sound Match Hearing Thresholds:

      • Test Set Sample Size: 100 subjects (76 with hearing loss, 24 normal hearing).
      • Data Provenance: Three geographically disparate U.S. sites (University of the Pacific - San Francisco and Stockton campuses). Data appears to be prospective, collected specifically for this validation.
    • Clinical Verification of Eargo's Self-Fitting Approach:

      • Test Set Sample Size: 33 subjects.
      • Data Provenance: Center for Applied and Translational Sensory Science at the University of Minnesota, U.S. Data appears to be prospective, collected specifically for this clinical trial.
    • Real-World Fit and Comfort / Real-World Safety:

      • Test Set Sample Size: 33 participants for Fit/Comfort; 31 participants for Safety.
      • Data Provenance: Real-world usage in the U.S. This appears to be prospective data collection.
    • Real-World Evaluation of Satisfaction and Usability:

      • Test Set Sample Size: 255 subjects.
      • Data Provenance: Web-based survey of adults in the U.S. who have used Eargo devices for at least two months. This appears to be retrospective (surveying existing users).
    • Human Factors Validation of Self-Fit Strategy:

      • Test Set Sample Size: 16 participants.
      • Data Provenance: U.S. setting for a summative usability test. This is prospective.
    • Human Factors Validation of Device Labeling and Device Handling/Maintenance:

      • Test Set Sample Size: 24 participants.
      • Data Provenance: U.S. setting for human factors validation testing. This is prospective.

    3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

    • Clinical Validation of Sound Match:

      • Ground Truth: "Audiometric thresholds obtained by an Audiologist in a sound-treated booth, following Audiology best practice methods (ABP)."
      • Experts: While the exact number of audiologists is not specified, it was conducted by "an Audiologist" (implying one per test administration) at the "University of the Pacific" which implies qualified personnel. The setup suggests multiple audiologists involved across the two campuses. Qualifications are implied by "Audiology best practice methods."
    • Clinical Verification of Self-Fitting Approach:

      • Ground Truth: "Real Ear Aided Response (REAR) to NAL-NL2 targets... and other measures (APHAB, speech in noise, sound quality ratings) were compared against results from the same hearing aid fit by an audiologist following clinical best practice methods for fitting hearing aids."
      • Experts: "a research audiologist" at the "University of Minnesota" for the AUD-fit condition. This implies a single, qualified audiologist for all AUD-fit programming in this trial. Further expertise is implicitly built into the NAL-NL2 prescription targets and clinical best practices.
    • Other studies (Usability, Safety, Satisfaction, Human Factors): Ground truth was established through participant feedback, observation of task completion, and adherence to established human factors methodologies rather than expert clinical interpretation of results. No external "experts" were explicitly mentioned for ground truth establishment beyond the study design and evaluation team.

    4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

    • The document does not describe a formal adjudication method (like 2+1 or 3+1) for the clinical or human factors studies.
    • Clinical Validation of Sound Match: Comparison was directly between two measurement methods (Eargo Sound Match vs. Audiologist Best Practice), with statistical analysis determining equivalence.
    • Clinical Verification of Self-Fitting: Within-subject crossover design, comparing self-fit outcomes directly against audiologist-fit outcomes, with statistical analysis for non-inferiority/equivalence.
    • Human Factors/Usability: Task success was observed and rated as successful, with difficulty, assistance, or failed/unable to complete. "No use errors" or "completed successfully by at least 80% of participants" were reported. The experimenter assisted only if asked, implying that "assistance" counted against independence.

    5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

    • No Multi-Reader Multi-Case (MRMC) study was done for AI assistance. This device is a self-fitting hearing aid, not an AI-assisted diagnostic tool for human readers. The studies compare a self-fitting method to a professionally-fitted method, both involving a human user (the patient) directly.

    6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done

    • Yes, in the context of the device's self-fitting algorithm. The "Sound Match" feature is described as Eargo's "proprietary, app-based method for self-determining audiometric thresholds," where the "hearing aids act as the transducer, emitting tonal stimuli... The measured hearing thresholds are then used as the basis for fitting the appropriate gain profile(s)." This is the algorithm's direct measurement output.
      • The "Clinical Validation of Eargo's Sound Match Hearing Thresholds" directly assesses the accuracy of this algorithm's output (EargoA/EargoB thresholds) against audiologist-derived thresholds (ABP). This is essentially a standalone evaluation of the algorithm's ability to measure hearing thresholds.

    7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

    • Clinical Validation of Sound Match: Expert Consensus / Standard of Care (audiometric thresholds measured by an audiologist following audiology best practice methods).
    • Clinical Verification of Self-Fitting: Mixed ground truth:
      • Objective: NAL-NL2 prescriptive targets for REAR (an established audiological fitting formula).
      • Comparative Clinical Outcome: Audiologist-fit condition using clinical best practices for APHAB, speech in noise, and subjective sound quality.
    • Usability/Human Factors: Observation of task completion based on pre-defined criteria, and user self-reported satisfaction/experience, aligned with established human factors validation methodologies.

    8. The sample size for the training set

    • The document does not provide specific information on the sample size for the training set used to develop the Eargo self-fitting algorithm or the Sound Match feature. It only describes the validation studies. The algorithm is described as "proprietary," indicating its development was internal to Eargo.

    9. How the ground truth for the training set was established

    • The document does not explicitly describe how the ground truth for the training set was established. Given the proprietary nature of the self-fitting algorithm and Sound Match, this information would likely be confidential commercial information not typically disclosed in a 510(k) summary. However, it can be inferred that its development would have been based on established audiology principles and data, similar to the validation methods, but applied to a development dataset.
    Ask a Question

    Ask a specific question about this device

    Page 1 of 1