Ambulatory and Perioperative Blood Pressure: Reply

Dr. Alexander Avidan asserts that hypotension is only associated with adverse outcomes. In fact, a randomized trial identified a 27% reduction in the relative risk of a composite of serious complications in patients assigned to tight individualized versus routine blood pressure management.¹  Available evidence thus suggests a causal relationship.

Dr. Avidan asks whether various monitoring methods might explain differences between ambulatory and perioperative pressures in our recent report.²  We recorded ambulatory pressures with upper-arm cuff oscillometry, a well-validated technique.³  Oscillometric ambulatory pressures better estimate individual blood pressure than single office measurements⁴  and are considered the reference method for assessing out-of-office pressures.⁵  For example, an international consensus group recently defined automated ambulatory blood pressure monitoring as “the optimal method to establish baseline values” before surgery.⁶ 

We similarly used upper-arm oscillometry to measure blood pressure before induction of general anesthesia and in 98% of patients during surgery. The relationship between oscillometrically and directly measured pressures is thus irrelevant to our assertion that preinduction pressure poorly represents normal ambulatory pressure. Upper-arm cuff oscillometry is by far the most common intraoperative blood pressure monitoring approach and is routinely used by clinicians to guide care. As with any monitor, oscillometric pressures are imprecise, but systematic bias is small.⁷ 

Dr. Avidan notes that arm-to-arm blood pressure differences may have contributed to the poor relationship we observed between preinduction and ambulatory pressures. Population average systolic and diastolic pressures are similar in the right and left arms.⁸  There can be inter-arm differences in individual patients,⁸  but substantive differences are relatively rare. In patients like ours, for example, only about 5% of patients have a systolic inter-arm difference of 10 mmHg or more.⁹  Presumably, inter-arm differences in mean arterial pressure would be even less.

Only 2% of our patients had an arterial catheter. Our clinical routine is to correctly level and zero pressure transducers, and repeatedly evaluate signals for incorrect damping. Importantly, we report mean arterial pressures, which are hardly influenced by damping. And again, errors would presumably be random and unlikely to influence our conclusions.

Dr. Avidan notes that we recorded ambulatory pressures at 30-min intervals and, of course, monitored pressures far more often intraoperatively. The issue, though, is not the frequency per se, but whether the frequency was sufficient to identify clinically important variation under each condition. Ambulatory pressures presumably change slowly compared to intraoperative pressures—which is exactly why intraoperative pressures are recorded frequently. Our monitoring intervals were therefore appropriate for each period.

Our definitions of daytime and nighttime pressure and our approach to identification and exclusion of artefactual blood pressure readings followed international guidelines.^10,11  The number of excluded artifacts during ambulatory blood pressure monitoring was well below accepted standards used to define “complete recording.”¹² 

In summary, our study methodology was sound and fully justifies our conclusion that preinduction mean arterial pressure cannot be used as a surrogate for the normal daytime mean arterial pressure, and that intraoperative mean arterial pressures are lower than the lowest nighttime mean arterial pressure in most patients.

The authors declare no competing interests.