Hemodynamic optimization is a major issue in the perioperative and intensive care settings. The titration of fluid in the operating room and in intensive care guided by the variations of the stroke volume (SV) and therefore of the cardiac output (CO) improves morbimortality. Esophageal Doppler is currently considered as the reference method for monitoring cardiac output in the operating room. However, when its use is not possible, there are devices that allow continuous estimation of cardiac output from the pulse wave contour (i.e. an invasive blood pressure signal). The PRAM method (pressure recording analytical method) with the MostCareUp monitor (Vytech, Padova, Italy) and the LTIA method (long time interval analysis) using an analog-to-digital converter connected to the patient's scope seem to be promising pulse wave analysis techniques. The majority of studies investigating these devices have focused on cardiac output values from a purely static perspective. Few studies have looked at changes in cardiac output during hemodynamic intervention (fluid challenge or use of vasopressors). However, in daily clinical practice, it is the dynamic information on CO that informs doctors on the effectiveness of therapeutic challenges. Thus, our hypothesis is to compare the variations in cardiac output estimated by PRAM et LTIA methods with those obtained using esophageal Doppler, during surgical procedures under general anesthesia.
The main goal of this study is to compare the relative changes in cardiac output obtained during hemodynamic interventions (fluid challenge or vasopressors) using PRAM and LTIA versus esophageal doppler. Parameters measured for the primary objective: 1. Continuous measurement and recording of the invasive blood pressure signal 2. Continuous and simultaneous measurement of cardiac output by MostCareUp (Vytech, Padova, Italy) and esophageal doppler. 3. Estimation of cardiac output from the arterial pressure signal by means of a digital analog converter using the LTIA algorithm connected to the patient's scope. For all patients, data from the cardiac output monitors are recorded throughout the procedure and the resuscitation stay. The secondary objective is to study the hemodynamic parameters of ventriculoarterial coupling calculated by the PRAM method and measured by esophageal doppler, other than CO and SV during therapeutic challenges. Parameters measured for the secondary objective: Global Afterload Angle (GALA), the value of the dicrote wave, dP/dtmax (myocardial contractility index), cardiac efficiency (CCE), arterial elastance, the total impedance of the system (Ztot).
Experimental design: this is a single-center, observational prospective study The study concerns adult patients benefiting from intraoperative hemodynamic optimization under general anesthesia, including routine monitoring of blood pressure and measurement of cardiac output by esophageal Doppler.
This study is "non-interventional" on adult patients undergoing interventional radiology, surgery under general anesthesia, or admitted to the ICU for postoperative organ failure. All measurements are obtained from standard monitoring including blood pressure and cardiac output measurement by transesophageal Doppler. Monitoring includes continuous electrocardiogram, SpO2, ventilatory parameters and depth of anesthesia by Bispectral Index. During the anaesthesia consultation, the patients will be given a letter of information on the objectives and the course of the study. Their non-opposition to participating in this study will be collected at the latest during the pre-anesthetic visit, the day before the operation. The protocol starts on the day of the operation. For all patients, the procedures will be performed under general anesthesia with an intravenous anesthesia protocol with a concentration objective (Orchestra® Base Primea - Fresenius Kabi France). The protocol does not interact with the care procedure. On the day of the procedure, the MostCareUp (Vytech, Padova, Italy) will be connected in series with the pressure head already in place. The blood pressure signals will be continuously recorded to allow estimation of the cardiac output by the LTIA method using an analog-to-digital converter. This calculation is made from the signal that has been transmitted to the multiparametric scope monitor and is therefore not directly connected to the patient. The acquisition of the analog signal is done via USB2.0 thanks to a Video Capture Card equipped with a converter. The software recalculates with its algorithm an estimation of the cardiac output from the data. The clinician in charge in the operating room will use only the hemodynamic data from the esophageal Doppler to guide his therapeutic strategy. Synchronization of the 2 devices. Recording of events during therapeutic tests if there is an indication for treatment. Usual protocol of the therapeutic challenges service: Patient receiving intraoperative vascular filling titration to optimize cardiac output (maximize SV/CO), patient with persistent arterial hypotension (MAP \< 65 mmHg or \< 80% of baseline) during the procedure despite adequate depth of anesthesia (BIS 40-50) and optimized cardiac output motivating vasopressor injection. For all patients, data from the 2 cardiac output monitors and hemodynamics will be collected during and after the therapeutic tests. Under no circumstances will a decision on therapeutic strategy be made based on PRAM or LTIA data. No additional complementary examination will be performed. The physician in charge of the study collecting the data does not intervene at any time in the management of the patient. The implementation of this type of intraoperative monitoring does not prolong the duration of anesthesia.
The measure of agreement the investigators will use is the intra-class correlation coefficient. To compare 3 measurement modalities, assuming that the intraclass correlation coefficient for the measurement of VES variations will be 0.85 with a confidence interval with width of 0.1 and a first-species risk of 0.05, 84 patients must be included. It is expected that the signals of 10% of the included patients will be of insufficient quality for their exploitation. Therefore, the calculated sample size is 92 patients.
Results will be expressed as mean (standard deviation) or median (interquartile range) depending on the normality of the data distribution. A Bland-Altman analysis for repeated measures will be used to evaluate the agreement between the cardiac output measured by esophageal Doppler and the cardiac output estimated by the two pulse wave monitors. It will be associated with a calculation of the bias and its 95% confidence interval, the upper/lower limits of agreement and their 95% confidence intervals, and the percentage error. A percentage of error less than or equal to 30% indicates the agreement of the technique with the reference method. The agreement of cardiac output variations measured by esophageal Doppler and pulse wave analysis will be evaluated by a Bland-Altman analysis. The percentage of agreement will be evaluated using a four-quadrant plot. Acceptable agreement is defined as greater than 90%. The analyses will be repeated in the filler and vasopressor subgroups. The ability of pulse wave monitors to detect responders to filling will be assessed and compared with esophageal doppler. Subjects will be defined as responders to volume expansion when the filling test increases the ESV by more than 10%.
