We thank Dr. Chen for his interest in our pilot study determining the feasibility of real-time noninvasive urine oxygen measurements in cardiac surgery patients. The device we investigated was a brand-new technology measuring urine oxygen outside of the body at the distal end of the urinary catheter. We chose to describe a measure of central tendency (mean urine oxygen values) because there were no previous studies to suggest an optimal threshold of urine oxygen measurements using this novel device. The threshold of 25 mmHg reported in our study was exploratory and should be validated in future work. One excellent way to do this, as suggested by Dr. Chen, would be to test the hypothesis that cumulative exposure to urinary hypoxia below this threshold is associated with increased risk of postoperative acute kidney injury.
For comments regarding variables that might affect the accuracy of urine oxygen monitoring, we agree that loop diuretics may affect urine oxygen measurements. We stated in our Methods that no diuretics other than mannitol were administered during surgery. We did investigate the effect of mannitol on urine oxygen measurements and found that there was no significant difference in urine oxygen between patients who did receive mannitol and those who did not. This may be because mannitol is an osmotic diuretic and therefore does not directly affect energy-consuming ion transport in the renal tubule. Alternatively, it might be because the doses of mannitol we used were relatively low (12.5 to 25 g). Larger studies and animal models may help elucidate the effects of various types of diuretics as well as vasoactive agents on real-time urine oxygen monitoring.