A consumer smartwatch worn overnight correctly identified adults with moderate-to-severe obstructive sleep apnea 94% of the time, according to a Stanford Medicine study published this month in the Journal of Clinical Sleep Medicine — one of the strongest performance results yet for a wearable device being evaluated as a sleep apnea screening tool.
Researchers at Stanford Health Care's Sleep Medicine Center tested the Samsung Galaxy Watch 6 against gold-standard polysomnography in 147 adults who had either a prior apnea diagnosis or a high pretest likelihood of moderate-to-severe disease. Participants wore the watch for two nights during in-lab sleep studies and for three additional nights at home, allowing researchers to compare the device's performance in both controlled and real-world settings, according to a report on the published findings.
What the Numbers Show
At the watch's default detection threshold, the device achieved:
- 94% sensitivity for moderate-to-severe obstructive sleep apnea, meaning it correctly flagged the large majority of people who actually had the condition
- 67% specificity, meaning roughly a third of people without moderate-to-severe disease were still flagged as positive
- 100% sensitivity and specificity in the subgroup of high-risk participants who also had a high hypoxic burden — a measure of how much and how often blood oxygen drops during sleep
The study builds on a multiyear collaboration between Samsung and Stanford Medicine, led by Dr. Robson Capasso, chief of sleep surgery at Stanford University School of Medicine, and Dr. Clete Kushida, a Stanford sleep medicine specialist who has studied wearable diagnostics for years. The underlying detection feature already holds FDA De Novo authorization and has been cleared in dozens of countries since its initial rollout, but this study is among the first to publish detailed sensitivity and specificity data in a peer-reviewed sleep journal.
Why Specificity Still Matters
The gap between the watch's sensitivity and specificity is the detail clinicians are watching most closely. A highly sensitive test is good at catching real cases, which matters for a condition — obstructive sleep apnea — that an estimated 30 million Americans have but the majority remain undiagnosed. But a specificity of 67% means a meaningful share of watch users without moderate-to-severe apnea would still receive a positive flag, potentially prompting unnecessary anxiety or follow-up testing.
That tradeoff is consistent with a broader pattern in wearable sleep apnea research. A pooled analysis of 12 smartwatch-based detection studies found an average sensitivity of 88.6% and specificity of 73.0% — figures the Stanford-Samsung results roughly track, if skewing toward higher sensitivity at some cost to specificity. Samsung's own validation testing, submitted to regulators, previously reported a somewhat lower sensitivity (82.67%) and higher specificity (87.69%) in a larger cohort of 470 adults, underscoring how performance can shift depending on the population tested and the detection threshold used.
Screening, Not Diagnosis
Both the researchers and Samsung are explicit that the feature is intended to flag risk, not replace a diagnostic sleep study. The Samsung Health Monitor app restricts the feature to adults 22 and older and directs anyone who receives a positive result to follow up with a healthcare provider for formal testing, whether an in-lab polysomnography or a home sleep apnea test.
The appeal is accessibility. A home or lab-based sleep study requires scheduling, cost, and often a referral, while a smartwatch a patient already owns can passively monitor for risk signals over multiple nights without any of those barriers. For a condition where the diagnostic bottleneck — not the treatment options — is often what keeps people untreated for years, a low-friction first flag could meaningfully shorten the path to care.
What This Means for Patients
Patients who already own a compatible Galaxy Watch and receive a positive apnea-risk notification should treat it as a prompt to talk to a doctor, not as a diagnosis in itself — the one-in-three chance of a false positive at this setting means a negative formal sleep study is entirely possible and does not mean the watch malfunctioned. Conversely, a 94% sensitivity means a negative reading is reasonably, though not perfectly, reassuring.
For the much larger population who have never been tested for sleep apnea despite risk factors like loud snoring, witnessed pauses in breathing, or persistent daytime fatigue, this kind of passive, low-cost screening could be the nudge that gets them into a sleep clinic. The device cannot yet substitute for a clinical workup, but as wearable screening tools mature and publish more rigorous peer-reviewed validation data, they may increasingly serve as the first step in a diagnostic pathway that, for most undiagnosed patients today, never begins at all.
