We thank Drs. Xiong et al.1 and Carella and Bonhomme2 for their comments regarding the shaping anesthetic techniques to reduce post-operative delirium (SHARP) study.3 We agree that evaluating intraoperative hypotension as a potentially moderating factor in the SHARP trial would be interesting. We found no difference in the number of minutes at a mean arterial pressure less than 55 mmHg between the spinal anesthesia with targeted sedation group (median, 0 min; interquartile range, 0 to 5) and the general anesthesia group (median, 0 min; interquartile range, 0 to 5; P = 0.51). Further, when the number of minutes at a mean arterial pressure less than 55 mmHg (considered as a categorical variable) was added to the main regression model as an interaction term, the interaction term was not significant, indicating that mean arterial pressure less than 55 mmHg did not modify the effect of anesthetic choice on postoperative delirium. Finally, the number of minutes at a mean arterial pressure less than 55 mmHg was not associated with delirium when added to the adjusted model described in the article.3 We did not prospectively record hypotension exposure below a 60- or 65-mmHg threshold but will consider this for potential future studies.
A further question was whether transient neurologic symptoms or urinary retention after spinal anesthesia could have caused increased pain or discomfort and thus influenced the development of postoperative delirium. Since the pain scores were similar on postoperative day 1 between groups, we do not think these complications were strong factors that could have biased the study results. Finally, we agree that Bispectral Index may not be the optimal tool to guide anesthesia depth, and further work using the electroencephalogram is needed. However, we designed this study based on previous studies that suggested a beneficial role for Bispectral Index in guiding anesthetic depth, and we utilized an anesthetic regimen that allowed for sedation to a level lighter than general anesthesia, irrespective of BIS levels.