How accurate is my wearable for detecting sleep apnea?

Not accurate enough to diagnose or exclude it. Consumer wearables can estimate sleep duration, rough stage classifications (light, deep, REM), and detect certain respiratory disturbance patterns with moderate correlation to polysomnography, but they cannot diagnose obstructive sleep apnea with sufficient reliability for clinical decision-making. An Oura ring or Whoop band saying your sleep quality was "good" does not exclude clinically significant OSA.

This gap is clinically important because OSA is one of the four primary reversible drivers of elevated hs-CRP and one of the most underdiagnosed cardiovascular risk factors in the men most likely to use wearables. The man who wears his Oura ring faithfully, sees decent sleep scores, and concludes he does not have sleep apnea has received false reassurance from a tool that was not designed for that diagnosis. The STOP-BANG questionnaire, a validated clinical screener for OSA risk, takes three minutes and has a sensitivity of 87–90% for moderate-to-severe OSA. A Whoop sleep score has no validated diagnostic sensitivity for sleep apnea at all. (Perez-Pozuelo et al., npj Digital Medicine, 2020)

Cardiologist's calibrated position, Early (3) for wearable sleep apnea detection. The accuracy is insufficient for clinical use.

What to do: If your STOP-BANG score is 3 or above, pursue a formal home sleep test regardless of what your wearable's sleep reports show. The two tools are not measuring the same thing.

For the full picture, read What Your Apple Watch Is Trying to Tell You.