Confirmation of discrepancies between exercise oximetry and American Heart Association post-exercise criteria to diagnose peripheral artery disease in patients with normal ankle-brachial index at rest.