Paul M Heerdt1, Scott Korfhagen2, Hesham Ezz3, Clara Oromendia4. 1. Department of Anesthesiology, Division of Applied Hemodynamics, Yale School of Medicine, New Haven, CT. Electronic address: paul.heerdt@yale.edu. 2. Department of Anesthesiology, Yale School of Medicine, New Haven, CT. 3. Department of Anesthesiology, Division of Applied Hemodynamics, Yale School of Medicine, New Haven, CT. 4. Department of Healthcare Policy and Research, Division of Biostatistics and Epidemiology, Weill Cornell Medical College, New York, NY.
Abstract
OBJECTIVE: To test the hypothesis that a simulation algorithm populated with data readily available from hemodynamic monitors and echocardiography can accurately model cardiac contractility, diastolic capacitance, and energetics. DESIGN: Bland-Altman analysis of paired data sets. SETTING: University laboratory. PARTICIPANTS: Archived data previously recorded from 7 anesthetized swine. MEASUREMENTS AND MAIN RESULTS: Left ventricular pressure and volume (LVV) data that had been continuously recorded over a range of inotropic conditions were used as reference data. One investigator performed conventional analysis of measured pressure/volume loops during preload reduction to derive reference values for end-systolic elastance (Ees-a measure of contractility), the predicted LVV at an end-diastolic pressure of 30 mmHg (V30-an index of diastolic capacitance and chamber dilation), and pressure-volume area (PVA-a correlate of myocardial oxygen consumption). Other investigators blinded to these results entered pressure, cardiac output, and ejection fraction measurements into a simulator that predicts Ees, V30, and PVA. Analysis of simulated data was performed before and after correction of the estimated LVV at which pressure would be 0 mmHg (V0), which was initially fixed in the model. Before V0 correction, accuracy and precision of Ees, V30, and PVA tended to fall outside predefined limits for method interchangeability, but utility for qualitative assessment of acute changes was evident. After V0 correction, the accuracy and precision of simulated data were within the defined limits for method interchangeability. CONCLUSIONS: These data support the potential for clinical utility of simulation models populated with data readily available at the bedside to characterize left ventricular mechanical performance and energetics.
OBJECTIVE: To test the hypothesis that a simulation algorithm populated with data readily available from hemodynamic monitors and echocardiography can accurately model cardiac contractility, diastolic capacitance, and energetics. DESIGN: Bland-Altman analysis of paired data sets. SETTING: University laboratory. PARTICIPANTS: Archived data previously recorded from 7 anesthetized swine. MEASUREMENTS AND MAIN RESULTS: Left ventricular pressure and volume (LVV) data that had been continuously recorded over a range of inotropic conditions were used as reference data. One investigator performed conventional analysis of measured pressure/volume loops during preload reduction to derive reference values for end-systolic elastance (Ees-a measure of contractility), the predicted LVV at an end-diastolic pressure of 30 mmHg (V30-an index of diastolic capacitance and chamber dilation), and pressure-volume area (PVA-a correlate of myocardial oxygen consumption). Other investigators blinded to these results entered pressure, cardiac output, and ejection fraction measurements into a simulator that predicts Ees, V30, and PVA. Analysis of simulated data was performed before and after correction of the estimated LVV at which pressure would be 0 mmHg (V0), which was initially fixed in the model. Before V0 correction, accuracy and precision of Ees, V30, and PVA tended to fall outside predefined limits for method interchangeability, but utility for qualitative assessment of acute changes was evident. After V0 correction, the accuracy and precision of simulated data were within the defined limits for method interchangeability. CONCLUSIONS: These data support the potential for clinical utility of simulation models populated with data readily available at the bedside to characterize left ventricular mechanical performance and energetics.