How accurate is the sleep apnea detection feature on the Apple Watch Ultra 2

Accuracy Metrics

Apple reports that the sensitivity of the sleep apnea detection algorithm is 66.6%, meaning it correctly identifies about two-thirds of individuals who have moderate to severe sleep apnea. In contrast, the specificity is 95.9%, indicating that it accurately identifies those without the condition most of the time[1][2]. This suggests that while false positives are rare, many cases of sleep apnea may go undetected.

Performance Breakdown

This discrepancy highlights that if a user does not receive an alert, it does not necessarily mean they are free from sleep apnea.

Methodology and Limitations

The detection relies primarily on data from the watch's accelerometer to monitor breathing disturbances during sleep. This approach has been criticized for potentially missing many cases, particularly moderate ones, due to its reliance on motion data alone[1][2]. Experts suggest that incorporating additional metrics, such as oxygen saturation levels or respiratory rates, could enhance accuracy[1].

Expert Opinions

Dr. Marschall Runge, a prominent figure in sleep medicine, acknowledges the feature's role in raising awareness about sleep apnea but cautions against relying solely on it for diagnosis. He emphasizes the importance of follow-up testing for those who receive notifications from the watch[2][4].

Overall, while the Apple Watch Ultra 2's sleep apnea detection feature is a step forward in consumer health technology, its current accuracy suggests it should be used as a preliminary tool rather than a definitive diagnostic solution.