Less Invasive, Continuous Hemodynamic Monitoring During Minimally Invasive Coronary Surgery

Less Invasive, Continuous Hemodynamic Monitoring During Minimally Invasive Coronary Surgery Oliver Gödje, MD, Christian Thiel, MS, Peter Lamm, MD, Hermann Reichenspurner, MD, PhD, Christof Schmitz, MD, Albert Schütz, MD, and Bruno Reichart, MD Department of Cardiac Surgery, University Hospital Grosshadern, Ludwig-Maximilians-University of Munich, Munich Germany Background. Minimally invasive coronary surgery has gained more and more clinical acceptance. A clear contrast to the minimally invasive idea is the highly invasive pulmonary artery catheter used for hemodynamic monitoring during the operation. We evaluated a less invasive device which calculates cardiac output (CO) and hemodynamics based on arterial pulse-contour analysis. Methods. In 20 patients revascularized by the off-pump technique with the octopus system, agreement of CO by pulse-contour was compared to pulmonary arterial and femoral arterial thermodilution and hemodynamic alterations during the operation were recorded. Pulse-contour CO is computed by measuring the area under the arterial pressure waveform and dividing it by aortic impedance. Aortic impedance is determined by an arterial thermodilution at the onset of the system. Results. Correlation of pulmonary arterial and arterial thermodilution CO to pulse-contour CO was 0.91 and 0.90 respectively (both p < 0.01). Coefficients of variations were 6.2% and 6.7%. The bias was 0.1 L per minute and standard deviations were 0.42 L per minute and 0.55 L per minute. Hemodynamic changes during the operations were seen mainly during the distal anastomosis of the first diagonal branch; only slight changes occurred during the anastomosis of the left anterior descending coronary artery. Conclusions. Arterial pulse-contour analysis is easy to use and minimally invasive, thus qualifies as a reliable routine monitoring tool during minimally invasive coronary surgery with tissue stabilizers. (Ann Thorac Surg 1999;68:1532 6) 1999 by The Society of Thoracic Surgeons Hemodynamic monitoring, based on arterial pulsecontour-analysis, has been shown to be an alternative to pulmonary artery catheter thermodilution in cardiac surgical intensive care [1, 2]. Pulse-contour monitoring demonstrated accuracy comparable to that of pulmonary artery thermodilution [3] using an approach that is clearly less invasive, because it needs only arterial and central venous access. By on-line pressure waveform analysis, the computer continuously calculates stroke volume (SV), cardiac output (CO), systemic vascular resistance (SVR), and myocardial performance by means of pressure increase time (d p /d t ). Pulse-contour monitoring can thus be a minimally invasive monitoring tool during minimally invasive surgery. We investigated the usefulness of the system by comparison to standard thermodilution in 20 patients operated on by the minimally invasive technique. Presented at Evolving Techniques and Technologies in Minimally Invasive Cardiac Surgery, San Antonio, TX, Jan 22 23, 1999. Address reprint requests to Dr Gödje, Department of Cardiac Surgery, University of Ulm, Steinhövelstr 9, 81377 Ulm, Germany; e-mail: oliver.goedje@medizin.uni-ulm.de. Material and Methods Patients The study included 20 patients (14 men, 6 women). Mean age was 61.8 9.8 years, mean height was 171.1 8.9 cm, and mean weight was 79.1 13.5 kg. In 12 patients the left anterior descending coronary artery (LAD) was revascularized by the left internal thoracic artery (LITA); in 8 patients, additionally, the first diagonal (D1) branch was anastomosed with a venous bypass. In all cases a median sternotomy was performed. For mechanical stabilization of the beating heart, the Octopus tissue stabilizer system (Medtronic Inc, Minneapolis, MN) [4] was used. Methods The PiCCO device (Pulsion Medical Systems, Munich, Germany) for CO calculation from arterial pulsecontour-analysis consists of a bedside computer, an inline injectate sensor in the central venous line, and a thermistor-tipped arterial catheter. The in-line sensor and the arterial thermistor are required for transthoracic thermodilution measurement; a pressure transducer detects arterial pressure waveform and heart rate, which is also included in the system. The pressure module of the ICU-specific patient monitor is connected to the PiCCO 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(99)00956-X

Ann Thorac Surg MINIMALLY INVASIVE GÖDJE ET AL 1999;68:1532 6 MONITORING IN MICS 1533 (Pulsion Medical Systems) instead of to the routine pressure transducer. The basic algorithm for the determination of cardiac output from pulse-contour was developed by Wesseling [5] and others. According to this algorithm, left ventricular SV is computed by measuring the area under the systolic part of the arterial pressure waveform and dividing this area by the aortic impedance. A subsequent multiplication with the heart rate yields cardiac output. According to the manufacturer information, the tested PiCCO system uses an enhanced version of the Wesseling algorithm. To adjust for aortic impedance, which differs from patient to patient, an arterial thermodilution measurement for the calibration of the system is required. Because arterial pressure and heart rate are substantial for cardiac output calculation and, hence, measured beat-tobeat, a continuous determination of cardiac afterload, in terms of SVR, also becomes possible. Although the system would allow a real beat-to-beat analysis of SV, CO, and SVR, for reasons of readability, the displayed values each consist of a sliding average of the preceding 30 seconds. This sliding average however, can be adjusted to the physician s needs. Preoperatively, for arterial pressure monitoring, a 4F thermistor-tipped catheter for thermodilution and pulsecontour analysis (PV 2014L, Pulsion Medical Systems, Munich, Germany) was inserted into the femoral artery. As part of our routine monitoring, a pulmonary arterial catheter (PAC) (Ohmeda, Erlangen, Germany) was inserted upon induction of anesthesia. The PAC (Ohmeda) was connected to the cardiac output module of the patient monitor (Siemens 1281, Siemens, Erlangen, Germany); the arterial catheter was connected to the PiCCO pulse-contour computer. The indicator for pulmonary artery and arterial thermodilution consisted of 10-ml iced dextrose 5% solution at a temperature of 4 7 centigrade, as measured by the in-line-injectate sensor of the thermodilution injectate set. To minimize the influence of variations of manual injection on the accuracy of the thermodilution measurements, the bolus injections were always carried out by the same person. Because the injected bolus was detectable in the pulmonary artery and the femoral artery, both measurements were performed simultaneously. At the onset of the PiCCO application, a triplicate arterial thermodilution was performed to calibrate the pulse-contour computer; however, none of the subsequent thermodilutions were used to recalibrate the system. Measurements were performed at the following times: immediately before skin incision, during preparation of the LITA, after placement and fixation of the Octopus stabilizer, after occlusion of the coronary artery, after release of the coronary flow, during the central venous anastomosis, during closure of the chest, and at the end of the operation. At each of these time points, we used PiCCO to perform simultaneous readings of mean arterial pressure (MAP), cardiac output by pulse-contour analysis (COpc), SV, SVR and d p /d t. We also performed CO measurements by triplicate femoral arterial (COart) and pulmonary arterial (COpa) thermodilution. Because the pulse-contour based values, due to the system s on-line-character, might have changed during the thermodilution period, the average of the values immediately before and after each set of thermodilutions were the values used for statistical evaluation. The investigation was carried out in accordance with the ethics committee of our institution and the principles of the Helsinki Declaration; informed consent was obtained from all patients prior to the operation. Statistics COpc was compared to COpa and COart by means of linear regression and Bland-Altman analyses. Because SV multiplied by heart rate yields CO, we did not separately compare SV values, based on the assumption that agreement of CO values must be similar to SV value agreement, because heart rate is the same in both calculations. A similar assumption is used for SVR values based on pulse-contour and thermodilution. SVR is mathematically produced from CO and MAP. Because MAP values would be the same for both SVR calculations, any difference or agreement of SVR solely depends on CO agreement. We recorded d p /d t values, but could not compare them to a standard method at the same time (eg, tip-manometer in the left ventricle), because such a standard method was not available in our investigation. All statistical analyses were computed by SPSS for Windows (Version 8.0, 1997, SPSS Inc, Chicago, IL). Results All 20 patients left the intensive care unit between the first and second postoperative day and were discharged from the clinic between eighth and 12th postoperative days. There was no occurrence of early bypass occlusion during the hospital stay. There were 192 complete sets of CO measurements by the various methods that could be used for statistical evaluation. Eight sets of measurements in 3 patients had to be discarded due to irregular pulse-contour or thermodilution curves. Agreement of CO Measurements Regression analysis between COpa and COpc showed a correlation coefficient of 0.91 and the Bland-Altman analysis resulted in a mean difference of 0.1 L per minute with a standard deviation of 0.42 L per minute. A similar good behavior was found for comparison of COart and COpc. The correlation coefficient was 0.90, the bias was 0.12 L per minute, and the standard deviation of the bias was 0.55 L per minute. Mean coefficient of variation of the 192 triplicate CO measurements was 6.2% for COpa and 6.7% for COart with no significant difference between both methods. Due to the nature of COpc as a continuously measured parameter that changes constantly, and in which the sliding average, and hence, stability can be influenced by the user, coefficients of variation were not computed for pulse-contour values.

1534 MINIMALLY INVASIVE GÖDJE ET AL Ann Thorac Surg MONITORING IN MICS 1999;68:1532 6 Fig 1. Course of means of cardiac output and stroke volume based on pulse-contour analysis and mean arterial pressure of 20 patients during the Octopus operation. Data sets at time of stabilizer placing and coronary occlusion for D1 consist of 8 patients (COpc cardiac output derived from pulse-contour analysis; SVpc stroke volume derived from pulse-contour analysis; MAP mean arterial pressure; D1 first diagonal branch; LITA left internal thoracic artery). Course of Parameters During the Operation The courses of mean values of COpc, SVpc and MAP are shown in Figure 1. The courses of SVRpc and d p /d t together with MAP are shown in Figure 2. COpc and SVpc increased during LITA harvesting (Fig 1): SVR decreased (Fig 2). During placement of the stabilizer and occlusion of the LAD, SVR increased, whereas COpc, SVpc, and MAP decreased. After releasing the coronary flow, COpc, SVpc, and MAP normalized to nearly initial values. During placement of the stabilizer and occlusion of the first diagonal branch, COpc, SVpc, and SVRpc conditions occurred that were similar to those of the LAD anastomosis. The negative changes however, were more evident without reaching critical values. During sideclamping of the aorta for the central bypass anastomosis, all values showed a large standard deviation. This is due to the fact that during side-clamping, the aortic impedance changes drastically. Because no recalibrations were performed to compensate for these impedance changes, pulse-contour values at this point of time were less reliable. After release of the side clamp, until the end of the operation, all parameters constantly increased to values within ranges of those at the time of the skin incision. During the anastomoses, d p /d t also decreased, but after finishing the last anastomosis, they showed a clear increase to values more favorable than the preoperative values. Comment With off-pump coronary surgery, new technologies have been developed for reduced invasiveness [4, 6, 7]. One of those is local cardiac wall stabilization by stabilizer systems [4]. This introduces the need for monitoring tools, because worsening myocardial conditions [8, 9] can not easily be covered by extracorporeal circulation. Hemodynamic monitoring by pulmonary artery catheter lacks continuous qualities, and it lacks properties that would qualify it to be called minimally invasive. Monitoring by pulse-contour analysis, calibrated with arterial thermodilution, is more adequate, because this technique is less invasive [1, 2]. To adjust the pulse-contour computer for the aortic impedance, an initial COart measurement is necessary. COart correlates to COpa with coefficients between 0.9 and 0.99 as shown in the past. It was concluded that calibrating the pulse-contour computer with COart is justified [3]. A correlation coefficient of 0.91 and a bias of 0.1 L per minute express a good agreement of COpc with conventional thermodilution in our study, standard deviations Fig 2. Course of means of systemic vascular resistance and d p /d t based on pulse-contour analysis and mean arterial pressure of 20 patients during the Octopus operation. Data sets at time of stabilizer placing and coronary occlusion for D1 consist of 8 patients (SVRpc systemic vascular resistance derived from pulse-contour analysis; d p /d t pressure increase time derived from pulsecontour analysis; MAP mean arterial pressure; D1 first diagonal branch; LITA left internal thoracic artery).

Ann Thorac Surg MINIMALLY INVASIVE GÖDJE ET AL 1999;68:1532 6 MONITORING IN MICS 1535 Fig 3. Course of cardiac output and stroke volume based on pulse-contour analysis and mean arterial pressure of a patient in whom a conversion to extracorporeal circulation became necessary (COpc cardiac output derived from pulse-contour analysis; SVpc stroke volume derived from pulse-contour analysis; MAP mean arterial pressure; ECC extracorporeal circulation; LITA left internal thoracic artery). being also within acceptable limits. This confirms results from previous studies, in which a thermodilution COpa measurement was used for impedance calibration. We also found a good agreement between COart and COpc (r 0.89, bias 0.1 L per minute). Thus, COpc agrees well with COpa and COart and has the potential to replace both. Another method for cardiac monitoring is transesophageal echocardiography (TEE). With this method, a continuous monitoring of myocardial condition is available [10], which clearly surpasses the possibilities of the PiCCO system. However, with TEE no information about the general hemodynamic situation is possible. In our opinion the pulse-contour method and TEE do not stand in contrast, but would be an ideal combination. Additionally the PiCCO can be used until the patient is discharged from the intensive or intermediate care unit, which is not possible with TEE. Looking at Figures 1 and 2, the question arises of whether both methods are really necessary. Pressure monitoring would have been sufficient, because no patient developed any problem. Figures 3 and 4, however, show a patient (not included) in whom conversion to extracorporeal circulation became necessary. If the arterial pressure measurement alone had been available, other threatening circulatory changes would not have been recognized. During placement of the stabilizer, COpc dramatically decreased; as a reaction, SVRpc increased which maintained sufficient arterial pressure. In fact, the need for conversion was not recognized (because in this investigation we only evaluated the PiCCO system and did not yet rely on it) before a further decrease of MAP occurred at the time of LAD occlusion. With regard to patient security, the use of the pulse-contour device seems to be justified. The pulse-contour method is an easy-to-use technique. Because of its continuity and less invasive qualities, it has great potential for use in monitoring patients during and after minimally invasive surgery. The continuous measurement of hemodynamics offers physicians a Fig 4. Course of systemic vascular resistance and d p /d t based on pulse-contour analysis and mean arterial pressure of a patient in whom a conversion to extracorporeal circulation became necessary. (SVRpc systemic vascular resistance derived from pulse-contour analysis; d p /d t pressure increase time derived from pulse-contour analysis; MAP mean arterial pressure; ECC extracorporeal circulation; LITA left internal thoracic artery).

1536 MINIMALLY INVASIVE GÖDJE ET AL Ann Thorac Surg MONITORING IN MICS 1999;68:1532 6 closer monitoring and a faster reaction to a patient s situation, at the same time relieving them of timeconsuming, repeated, intermittent measurements. References 1. Gödje O, Höke K, Lamm P, et al. Continuous, less invasive, hemodynamic monitoring in intensive care after cardiac surgery. Thorac Cardiovasc Surg 1998;46:242 49. 2. Gödje O, Höke K, Lichtwarck-Aschoff M, Lamm P, Reichart B. Less invasive continuous cardiac output determination by femoral arterial thermodilution calibrated pulse contour analysis a comparison to conventional pulmonary arterial cardiac output. Crit Care Med 1999: in press. 3. Gödje O, Peyerl M, Seebauer T, Dewald O, Reichart B. Accuracy and reproducibility of pulmonary artery and arterial thermal and thermal dye dilution variables in cardiac surgery patients. Chest 1998;13:1070 7. 4. Jansen EW, Borst C, Lahpor JR, et al. Coronary artery bypass grafting without cardiopulmonary bypass using the octopus method: results in the first one hundred patients. J Thorac Cardiovasc Surg 1998;116:60 7. 5. Wesseling KH, de Wit B, Weber JAP, Smith NT. A simple device for the continuous measurement of cardiac output. Adv Cardiovasc Phys 1983;5:16 52. 6. Diegeler A, Falk V, Matin M, et al. Minimally invasive coronary artery bypass grafting without cardiopulmonary bypass. Early experience and follow-up. Ann Thorac Surg 1998;66:1022 5. 7. Society of Thoracic Surgeons/American Association for Thoracic Surgery Committee on New Technology. Policy statement: minimally invasive coronary artery bypass surgery. Ann Thorac Surg 1998;66:1848 9. 8. Burfeind WR, Duhaylongsod FG, Samuelson D, Leone BJ. The effects of mechanical stabilization on left ventricular performance. Eur J Cardiothorac Surg 1998;14:285 9. 9. Grundemann PF, Borst C, van Herwaarden JA, Verlaan CW, Jansen EW. Vertical displacement of the beating heart by the octopus tissue stabilizer: influence on coronary flow. Ann Thorac Surg 1998;65:1348 52. 10. Perrino AC, Harris SN, Luther MA. Intraoperative determination of cardiac output using multiplane transesophageal echocardiography: a comparison to thermodilution. Anesthesiology 1998;89:350 7.