To the Editor: We thank Drs. Vavilala and Lam for their interest in our recent report and for their insightful comments. However, there appears to be misunderstanding of the methodology used in our study. The authors are correct in stating that reliance upon spontaneously occurring blood pressure changes to assess the efficacy of cerebral pressure autoregulation leaves uncertainty about whether the cerebral autoregulatory system has been adequately tested. This is only true in situations where there is lack of concordance between blood pressure (BP) and cerebral blood flow (CBF). However, in situations when there is significant frequency-specific coherence between changes in BP and those in CBF, it is reasonable to infer that the changes in BP are causally related to the changes in CBF. It is for precisely this reason that we emphasized in this report (and in its title) that we were testing not for the presence of autoregulation but rather its absence, i.e., the pressure-passive state. We agree with Drs. Vavilala and Lam that CBF is influenced by a host of other factors, including cerebral metabolic rate (CMR). It is likely that at any given time a variety of stimuli are exerting an influence on the regulation of CBF. For these reasons, we have used the systems analysis approach together with coherence and transfer function analysis to better discriminate between the effects on CBF of changes in BP and changes in other “input” stimuli. It is well known that under normal conditions, changes in CMR trigger changes in CBF. However, unless CMR and BP were changing at the exact same frequency, changes in CMR would not influence the coherence between BP and CBF.

Finally, Drs. Vavilala and Lam quite correctly state that the output of an end-tidal CO2 (ET-CO2) monitor may not correlate accurately with measured arterial PaCO2. As other techniques for continuous PaCO2 measurement are, unfortunately, not well established, this remains a reliable and useful trend monitor. As stated in the Methods section of our paper, we adjusted the ET-CO2 values to measured arterial PaCO2 levels at the beginning and end of the studies.

Thank you for the opportunity to clarify these issues.