Effective Cerebral Perfusion Pressure: Does the Estimation Method Make a Difference?

Introduction: The effective cerebral perfusion pressure (CPPe), zero-flow pressure (ZFP), and resistance area product (RAP) are important determinants of cerebral blood flow. ZFP and RAP are usually estimated by linear regression analysis of pressure-velocity relationships of the middle cerebral artery. The aim of this study was to validate 4 other estimation methods against the standard linear regression method.

Methods: In a previous study, electroencephalography, arterial blood pressure, and middle cerebral artery flow velocity were measured in patients during internal cardioverter defibrillator implantation procedures to determine the electroencephalography frequency ranges that represent ischemic changes during periods of circulatory arrest. In this secondary analysis, arterial blood pressure and middle cerebral artery flow velocity were used to estimate CPPe, ZFP, and RAP by 4 different methods-the 3-point intercept calculation (LR3, systolic/mean/diastolic) and methods described by Czosnyka (systolic/diastolic), Belford (mean/diastolic), and Schmidt (systolic/diastolic)-and compare them with the reference linear regression method. CPPe was calculated as the difference between mean arterial pressure and ZFP. The primary endpoint was the difference, correlation, and agreement of these differently estimated CPPe measurements.

Results: In total, 174 measurements in 35 patients were collected under steady-state conditions before the first circulatory arrest phase during internal cardioverter defibrillator testing. CPPe, ZFP, and RAP measurements based on the 3-point intercept and Czosnyka calculation methods showed small mean differences, good agreement, low percentage errors, and excellent correlation when compared with the reference method. Agreement and correlation were moderate for the Belford method and unsatisfactory for the Schmidt method.

Conclusions: CPPe, ZFP, and RAP measurements based on 2 alternative calculation methods are comparable to the linear regression reference method.