(90 days)
The Lunoa System is indicated for prescription use for the treatments with positional obstructive sleep apnea with a non-supine apnea-hypopnea index
The Lunoa System is a rechargeable battery-operated medical device, worn around the chest in an elasticized chest strap (Figure 1), intended to keep patients with positional obstructive sleep apnea (POSA) from sleeping in the supine position. The System consists of a sensor device, chest strap, docking station, power adapter, travel case, and portal.
The Lunoa System is a medical device for treating positional obstructive sleep apnea (POSA). The acceptance criteria for the device were implicitly established through the clinical studies conducted to demonstrate its safety and effectiveness. The summarized results from these studies serve as the reported device performance.
Here's a breakdown of the requested information based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are inferred from the demonstrated improvements in clinical endpoints through various studies. The reported device performance is taken directly from the "Results Summary" (Table 5-2) and the descriptions within the text.
| Acceptance Criteria (Inferred from Study Goals) | Reported Device Performance (Lunoa System / SPT v1.1) |
|---|---|
| Reduction in percentage of supine sleep time (STS) | Median % STS decreased from 49.9% to 0.0% (Van Maanen et al. 2013) |
| Median % STS decreased from 21% to 2.0 % (Van Maanen & De Vries 2014) | |
| Median % STS decreased from 40.1% to 7.4% (Benoist et al. 2016) | |
| Median % supine sleep time decreased from 31.1% to 0 % (Eijsvogel et al. 2015) | |
| Median % STS decreased from 31.9% to 0% (Dieltjens et al. 2015) | |
| Median % STS decreased from 43.0% to 11% (Benoist & de Ruiter et al. 2016) | |
| Mean % STS decreased from 47% to 17% (Laub et al. 2016) | |
| Mean % STS decreased from 41.6% to 12.7% (De Ruiter et al. 2017) | |
| Reduction in Apnea-Hypopnea Index (AHI) | Median AHI decreased from 16.4 to 5.2 (Van Maanen et al. 2013) |
| NR (Van Maanen & De Vries 2014) | |
| Median AHI decreased from 18.3 to 12.5 (Benoist et al. 2016) | |
| Median AHI decreased from 13.1 to 5.8 (Eijsvogel et al. 2015) | |
| Median AHI decreased from 20.8 to 11.1 (SPT). SPT + MAD reduced to 5.7 (Dieltjens et al. 2015) | |
| Median AHI decreased from 13.0 to 7.0 (Benoist & de Ruiter et al. 2016) | |
| Mean AHI decreased from 18 to 10 (Laub et al. 2016) | |
| Mean AHI decreased from 13.2 to 7.1 (De Ruiter et al. 2017) | |
| Improvement in Epworth Sleepiness Scale (ESS) | Decreased significantly (Van Maanen et al. 2013, Van Maanen & De Vries 2014, Benoist et al. 2016) |
| No change between groups (Eijsvogel et al. 2015, Laub et al. 2016) | |
| NR (Dieltjens et al. 2015) | |
| No significant change (Benoist & de Ruiter et al. 2016) | |
| No significant change (De Ruiter et al. 2017) | |
| Improvement in Functional Outcomes of Sleep Questionnaire (FOSQ) | Increased significantly (Van Maanen et al. 2013, Van Maanen & De Vries 2014) |
| NR (Benoist et al. 2016, Eijsvogel et al. 2015, Dieltjens et al. 2015, Laub et al. 2016, De Ruiter et al. 2017) | |
| No change between groups (Benoist & de Ruiter et al. 2016) | |
| Compliance with therapy | 92.7% (Van Maanen et al. 2013) |
| 71.2% (Van Maanen & De Vries 2014) | |
| 89% (Benoist et al. 2016) | |
| 75.9% (Eijsvogel et al. 2015) | |
| 89.3% (Benoist & de Ruiter et al. 2016) | |
| 75.5% (Laub et al. 2016) | |
| 100% (De Ruiter et al. 2017) |
2. Sample Size Used for the Test Set and Data Provenance
The text describes eight investigator-initiated clinical studies. Each study contributed to the overall clinical evidence. There isn't a single "test set" described in the conventional sense of a distinct dataset used solely for final validation after development. Instead, the performance is demonstrated across these multiple clinical trials.
- Sample Sizes: The number of patients in each study varied:
- Van Maanen et al. 2013: 31 patients
- Van Maanen & De Vries 2014: 106 patients
- Benoist et al. 2016: 33 patients
- Eijsvogel et al. 2015: 21 TBT, 27 SPT
- Dieltjens et al. 2015: 20 patients
- Benoist & de Ruiter et al. 2016: 81 patients
- Laub et al. 2016: 52 SPT, 49 non-treatment control
- De Ruiter et al. 2017: 29 SPT, 29 MAD
- Data Provenance (Country of Origin):
- Amsterdam, The Netherlands (multiple studies)
- Enschede, The Netherlands
- Edegem, Belgium
- Glostrup, Denmark
- Retrospective or Prospective: All studies are described as prospective, with some being randomized, parallel, cohort studies, and others single-arm cohort studies.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The text does not explicitly state the number of experts used to establish ground truth or their specific qualifications for individual patient diagnoses or data interpretation within the clinical studies. However, the studies were conducted by named authors (researchers/clinicians) and published in peer-reviewed sleep and breathing journals, implying that clinical diagnoses and assessments (e.g., AHI determined by Polysomnography (PSG)) were performed by qualified medical professionals in sleep medicine.
4. Adjudication Method for the Test Set
The text does not explicitly describe an adjudication method (like 2+1, 3+1) for the data collected in the clinical studies. Clinical studies typically involve standard diagnostic procedures (like PSG) which are interpreted by trained staff, but specific adjudication processes for individual case interpretations are not detailed in this summary.
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, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study of human readers with and without AI assistance was not conducted. The Lunoa System itself is a therapeutic device, not an AI diagnostic tool primarily interpreted by human readers. The studies compared the Lunoa System (or its previous version, SPT v1.1) against other treatments or control groups.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
The Lunoa System is a standalone device in the context of its therapeutic function. It independently detects sleep position and provides vibro-tactile feedback. The "algorithm" in this case refers to the device's inherent logic for sensing position and initiating vibration. Its performance was tested as a standalone therapy, without a human actively intervening based on its real-time output (beyond initially setting it up and monitoring compliance). The clinical studies evaluated its effectiveness directly as a therapeutic intervention.
7. The Type of Ground Truth Used
The primary ground truth for evaluating the effectiveness of the Lunoa System was Polysomnography (PSG) data. PSG is considered the gold standard for diagnosing sleep disorders, including obstructive sleep apnea. The key metrics derived from PSG, such as Apnea-Hypopnea Index (AHI) and percentage of supine sleeping time, were used as objective measures of treatment efficacy. Subjective ground truths, such as Epworth Sleepiness Scale (ESS) and Functional Outcomes of Sleep Questionnaire (FOSQ), were also used to assess patient-reported outcomes.
8. The Sample Size for the Training Set
The text does not mention a distinct "training set" for an algorithm in the machine learning sense. The Lunoa System appears to rely on established principles of accelerometry for position detection and vibro-tactile feedback to deter supine sleep. Its development would likely involve engineering validation and perhaps iterative testing, but not necessarily a "training set" as understood for complex AI models in diagnostic imaging. The clinical studies evaluated the final device's performance, not the training of an underlying AI model.
9. How the Ground Truth for the Training Set Was Established
As no explicit "training set" for an AI algorithm is mentioned, the method for establishing its ground truth is not applicable or described in this document. The device's fundamental function (position detection) relies on accelerometer technology, which is verified through engineering principles rather than a labeled training dataset in the AI context.
§ 872.5570 Intraoral devices for snoring and intraoral devices for snoring and obstructive sleep apnea.
(a)
Identification. Intraoral devices for snoring and intraoral devices for snoring and obstructive sleep apnea are devices that are worn during sleep to reduce the incidence of snoring and to treat obstructive sleep apnea. The devices are designed to increase the patency of the airway and to decrease air turbulence and airway obstruction. The classification includes palatal lifting devices, tongue retaining devices, and mandibular repositioning devices.(b)
Classification. Class II (special controls). The special control for these devices is the FDA guidance document entitled “Class II Special Controls Guidance Document: Intraoral Devices for Snoring and/or Obstructive Sleep Apnea; Guidance for Industry and FDA.”