SNOZAS+: Exploratory Study Presented at the European Sleep Research Congress 2026
78.8% of available sleep polygraphy results showed moderate-to-severe obstructive sleep apnea.
115 of 146 available polygraphy results showed AHI ≥ 15/h, following triage and medical assessment, in an exploratory study including 1,351 adults.
Editorial version: 4 November 2026
The finding in the examined subgroup
115 of the 146 available polygraphy results showed moderate-to-severe obstructive sleep apnea (OSA): an apnea-hypopnea index (AHI) ≥ 15 events/hour.
This observed yield among available examinations does not mean 78.8% of all participants, and does not establish the positive predictive value of the algorithm alone.
Other findings, with their denominators
adult questionnaires included out of 1,375 collected.
110 of 115 confirmed cases identified by the respiratory component of SNOZAS+.
agreement with adapted STOP-Bang across 1,277 evaluable profiles.
Adapted STOP-Bang identified 107 of the 115 confirmed cases (93.0%). These proportions relate to the tested subgroup; they are not validated estimates of diagnostic sensitivity in the general population.
Why multicategory triage?
SNOZAS+ structures self-reported symptoms to guide medical assessment. Respiratory risk assessment focuses on possible sleep apnea; multicategory triage also organizes other complaints and overlapping presentations.
- Identify respiratory features consistent with possible OSA.
- Consider insomnia symptoms and mixed presentations such as COMISA (comorbid insomnia and sleep apnea).
- Explore other sleep and alertness complaints and guide assessment by an appropriate clinician, with testing where indicated.
This study concerns the respiratory component of SNOZAS+. It does not validate non-respiratory categories and is not a validation of SleepDetective.
Methods: an exploratory study in a real-world pathway
After the online questionnaire and initial triage, all participants were offered a consultation with a sleep physician. An offer of consultation does not mean an appointment was booked or attended. The analysis concerns initial classifications and the examinations available within this pathway.
Cohort counts and results come from the published abstract. The clarification that consultation was offered to everyone, without assuming attendance, is operational information supplied by the SNOZAS team.
Comparison with adapted STOP-Bang
The strict respiratory classification of SNOZAS+ was compared with a 7-item STOP-Bang score in 1,277 evaluable records. Neck circumference was omitted and a threshold ≥ 3 was used.
| Tool | Identified / confirmed cases | Proportion in this subgroup |
|---|---|---|
| SNOZAS+ respiratory component | 110 / 115 | 95.7% |
| Adapted STOP-Bang (≥ 3) | 107 / 115 | 93.0% |
SNOZAS+ identified four confirmed cases with an adapted STOP-Bang score below 3. Conversely, one case identified by adapted STOP-Bang was not identified by SNOZAS+.
Across the 1,277 evaluable profiles: 82.0% agreement, Cohen’s kappa 0.60 (95% CI: 0.55–0.65). A difference in classification does not demonstrate clinical superiority.
COMISA: further assessment of overlapping conditions
In the group initially triaged towards another sleep condition, 8 patients had an available polygraphy result. Six had a confirmed respiratory disorder; 5 of those 6 had already been flagged by the algorithm as having COMISA features.
The abstract does not state that all 6 respiratory disorders had AHI ≥ 15/h. This very small subgroup supports further investigation of overlapping conditions, without extrapolation to the cohort or other patients.
Limitations and future validation
- An exploratory study, rather than definitive prospective validation.
- Polygraphy was available for a subgroup only, with potential selection and verification bias.
- The data cannot establish diagnostic performance generalisable to all 1,351 adults.
- Future prospective studies need systematic objective testing across the different triage categories.
- Performance of non-respiratory categories requires separate validation.
Triage is not a diagnosis. A clinician must interpret symptoms, decide on testing and make a diagnosis. Presentation at ESRS does not mean ESRS approval, certification or validation of the device.
Scientific publication and sources
- Author of the scientific study
- Dr Guillaume Marchand
The Sleep Initiative, Paris, France. - Original title
- Exploratory Concordance Study of the Respiratory Orientation Component of a Multicategory Sleep Triage Algorithm
28th Congress of the European Sleep Research Society — Sleep Europe 2026. Maastricht, the Netherlands, 20–23 October 2026.
Abstract published in the Journal of Sleep Research collection “Late Breaking Abstracts—Clinical Science”, volume 35, supplement S1 (2026), e70459.
Collection DOI: 10.1111/jsr.70459. This identifies the abstract collection, not an individual SNOZAS+ research article.
- Read the scientific abstract in the Journal of Sleep Research ↗
- Visit the congress on the ESRS website ↗
The original poster PDF is still to be supplied; no poster download is currently available on this page.
Edited by SNOZAS — Kai Medical Ltd. Dr Guillaume Marchand is the author of the cited study; he is not credited as the author of this editorial summary.
Scientific FAQ
What is SNOZAS+?
A questionnaire that structures self-reported sleep complaints to support triage for medical assessment. It does not make a diagnosis.
Which sleep conditions does it explore?
Respiratory symptoms, insomnia, overlapping presentations and other sleep and alertness complaints. This study does not validate non-respiratory categories.
What does the 78.8% result mean?
115 of the 146 available polygraphy results showed AHI ≥ 15/h. The 146 examinations came from both initial triage groups: 138 from the presumed OSA group and 8 from the other group. This concerns the examined subgroup following triage and medical assessment, not all participants. It does not establish the positive predictive value of the algorithm alone.
How does it differ from STOP-Bang?
STOP-Bang assesses respiratory risk, whereas SNOZAS+ supports multicategory triage. The study used an adapted 7-item STOP-Bang, without neck circumference, at a threshold ≥ 3.
Has SNOZAS+ been clinically validated?
This study provides exploratory findings on its respiratory component. Prospective validation with systematic objective testing remains necessary. It does not validate all categories or SleepDetective.
Is a clinician still necessary?
Yes. A clinician interprets the answers, considers alternative explanations and decides on examinations and treatment. In the studied pathway, all participants were offered a consultation; this does not mean everyone booked or attended an appointment.
Where can I read the study results?
The original abstract is available in the Journal of Sleep Research collection linked above. The original poster PDF is still to be supplied.