Abstract Sat204: Prediction of Common Carotid Flow using Pressure Values in a Swine Model of CPR

Abstract

Introduction: The primary goals of cardiopulmonary resuscitation (CPR) are to generate adequate myocardial blood flow to enable restoration of mechanical function and adequate brain blood flow to minimize ischemic injury. While the correlation between myocardial blood flow and coronary perfusion pressure during CPR is well established, the correlation between cerebral blood flow and cerebral perfusion pressure is less clear. Common carotid artery (CCA) blood flow is an accessible metric of brain blood flow during CPR in large animal studies. Here, we evaluate the relationships between CCA and several pressure values during CPR. Methods: Analysis was performed on 28 swine. Animals were instrumented to monitor arterial blood pressure (ABP), central venous blood pressure (CVP), intracranial pressure (ICP) and common carotid artery flow. Cerebral perfusion pressure (CePP) was calculated as mean arterial pressure (MAP) minus ICP. Following baseline measurements, ventricular fibrillation cardiac arrest was initiated and CPR started after 8 minutes. Pressure waveforms from the first 8 minutes of CPR (prior to epinephrine administration) were separated into 5 second segments used to calculate all parameters ( A ). Datapoints from the same individual were averaged together and relationships between variables were evaluated using ordinary least-squares regression. Models were compared using R 2 and Root Mean Squared Error (RMSE). Results: Figure panel A illustrates pressures and flow during the chest compression cycle. Regression between ABP value at the systolic peak and % baseline CCA flow ( B ) was weakest (R 2 = 0.31, p=0.003, RMSE=7.1%). Regression between MAP and % baseline CCA flow ( C ) resulted in a moderate positive relationship (R 2 = 0.46, p<0.001, RMSE = 6.2%), while regression between CePP and % baseline CCA flow ( D ) was slightly weaker (R 2 = 0.34, p=0.001, RMSE = 6.9%). Conclusion: While all three parameters showed positive, significant relationships with percent prearrest CCA flow, they may not be adequately reliable surrogates for achieving specific brain blood flow goals during CPR. While all three parameters showed positive, significant relationships with percent prearrest CCA flow, they may not be adequately reliable surrogates for achieving specific brain blood flow goals during CPR. These results emphasize the importance of developing new techniques to quantitatively monitor brain blood flow during physiology-guided CPR.