Bioethics of evidence

Parallel case for modeling only: Should a doctor rely on the heart-rate reading from a patient's fitness watch to decide on treatment?\n\nClaim: A fitness watch reading should not be used to make a medical decision on its own, because a consumer device is only as reliable as the conditions it was measured under.\nEvidence: Wrist-worn optical heart-rate sensors are known to lose accuracy during vigorous exercise, on darker skin tones, and when the band is loose, sometimes reporting a rate that is off by twenty beats per minute or more compared with a clinical electrocardiogram.\nReasoning: A number on a bright screen feels precise, but the watch estimates heart rate indirectly by shining light through the skin, and that estimate depends on fit, motion, and skin. A clinician who treats the reading as a verified fact skips the question of how it was produced. The device is a starting point, not a diagnosis, so a responsible decision confirms the reading with a validated clinical measurement before acting on it. That is why medical devices are cleared through testing that consumer wearables are not, and why the person reading the number, not just the number, carries responsibility for how it is used.