Following cardiopulmonary bypass surgery, preoperative diastolic pressure was more closely associated with failed extubation at 24 hours than systolic or mean blood pressure, according to a study of a large sample of cardiac surgery patients.
“This relationship is statistically significant when interpreted as a linear regression model,” said lead investigator Christos Koutentis, MB ChB, an anesthesiologist in private practice in New York City, who presented the findings at the International Congress of the Israeli Society of Anesthesiologists (abstract 53). “Preinduction diastolic blood pressure becomes even more significantly associated with failed extubation by 24 hours when interpreted as a quadratic polynomial or spline” (Figures 1 and 2).
Dr. Koutentis and his colleagues did not detect this phenomenon with any of the other systemic or pulmonary hemodynamic parameters. “However, a weaker but consistent relationship was observed with systemic diastolic pressures recorded immediately prior to going on bypass,” he said.
Furthermore, restricted cubic spline modeling, when adjusted, suggested that there may be an inflection point. “But the value of this finding depends on the degree of adjustment and exclusion criteria,” Dr. Koutentis said.
Dr. Koutentis was inspired to conduct this study because of a lack of reliable, evidence-based recommendations for interpreting preoperative or preinduction blood pressure thresholds in cardiac surgery. “Mean arterial pressures have been considered to be of more value because systolic blood pressure and diastolic blood pressure are subject to instability and distortion for a variety of reasons.”
Previously, the investigators reported that preoperative diastolic blood pressure was more significantly associated with vasoplegia, and that this relationship also was not linear.
Role for Preoperative Hemodynamics?
The current study analyzed samples from 1,992 cardiac surgery patients at Yale New Haven Hospital, in Connecticut, who underwent procedures between January 2009 and December 2011.
“Because delayed extubation is considered to be a harbinger for other more serious outcomes, we tested the hypothesis that preoperative diastolic blood pressure had an association with extubation at 24 hours post-surgery,” Dr. Koutentis said.
Dr. Koutentis initially expected mean arterial pressure to be more explanatory. “Intriguingly, a recently published study in the British Journal of Anaesthesia [2017;119(1):65-77] suggested an inflection point of 63 mm Hg, which was quite similar to ours, despite that study using a sample of general surgical patients, with blood pressures being measured in the community setting,” he said.
Dr. Koutentis is also fascinated by some individual outcomes of the current study. An example is that 30-day (and one-year) mortality showed no linear association with any preinduction systemic pressures. “Pointedly, these became closely linked with preinduction diastolic blood pressure, but only when this was transformed into a quadratic or other polynomial relationship,” he said. “Additionally, even prior to transformation, we saw the same effects on hazard ratios for length of hospital stay, and logistic models for returning to the operating room for cardiac-related complications and 30-day readmission rates.”
Still, results of the study should be interpreted with caution, “as you cannot make a simple recommendation based on blood pressure alone,” Dr. Koutentis said. “Because these pressures were taken immediately prior to anesthesia induction, they may not necessarily reflect a valid baseline of blood pressure control in the community setting. We do not know if there was an in-house optimization of blood pressure control prior to entering the operating room or whether these patients had been administered anxiolytics.”
Nonetheless, there may be a potential role for preoperative hemodynamic indices as part of a model for preoperative risk stratification, according to Dr. Koutentis. “Intraoperative or surgical variables are more likely to influence the final outcome, although it is possible that these may be connected to preoperative pressures within a causal pathway.”
Although it may appear contradictory, “I am not necessarily convinced that useful, fully adjusted models should simultaneously be constructed from preoperative blood pressures and other intraoperative or immediately postoperative variables,” Dr. Koutentis said.
Another cautionary finding of the study, when considering inclusion and exclusion criteria for future studies, is that the effect sizes (whether statistically significant or not) of other preexisting conditions, such as end-stage renal disease, cardiomyopathy or whether the procedure is a reoperation, “far outweigh the observed effects of the variables we tested,” Dr. Koutentis said.
Despite the lack of reliable, evidence-based recommendations on preoperative blood pressure thresholds in cardiac surgery, “perhaps we can look more carefully at ‘time to extubation’ in a semiparametric model, such as Cox proportional hazards, or as a nonproportional ordinal logistic regression, focusing on extubation within prespecified time intervals, such as six hourly blocks of time leading up to 24 hours, compared to extubation after 24 hours,” Dr. Koutentis said.
In addition, future studies need to consider how to standardize the conditions under which static preoperative hemodynamic values are recorded, according to Dr. Koutentis. “Dynamic and time-weighted average measures interpreted from both intraoperative and postoperative electronic medical records are also expected to hold more granularity, and thus potential for estimating vectors of instability or variability, provided the adjustments are thoughtfully carried out,” he said.
Longitudinal studies using repeated measures of the same variables might be promising, too, “although adjustment for loading conditions, vasomotor status and contractility might not be straightforward,” Dr. Koutentis said.