Transpulmonary versus continuous thermodilution cardiac output after valvular and coronary artery surgery

Transcription

1 doi: /icvts Interactive CardioVascular and Thoracic Surgery 9 (2009) Work in progress report - Cardiac general Transpulmonary versus continuous thermodilution cardiac output after valvular and coronary artery surgery a a, b b Rose-Marieke B.G.E. Breukers, A.B. Johan Groeneveld *, Rob B.P. de Wilde, Jos R.C. Jansen a Department of Intensive Care and Institute for Cardiovascular Research, VU University Medical Center, Amsterdam, The Netherlands b Department of Intensive Care, Leiden University Medical Center, Leiden, The Netherlands Received 5 February 2009; received in revised form 30 March 2009; accepted 31 March 2009 Abstract Residual left-sided valvular insufficiencies after valvular surgery may confound transpulmonary thermodilution cardiac output (COtp). We compared the technique with the continuous right-sided thermodilution technique (CCO) after valvular surgery (ns8) and coronary artery surgery (ns8). Patients with pulmonary and femoral artery catheters in the intensive care unit (ICU) were included. After valvular surgery, there was minimal aortic insufficiency in four patients and minimal to moderate mitral valve insufficiency in six. Five fluid loading steps (250 ml) were done in each patient. CCO and COtp were measured prior to and 15 min after each step. The cardiac output was lower after valvular than coronary artery surgery but responses to fluid loading steps were similar among surgery types and techniques. After valvular and coronary artery surgery, cardiac output was lower prior to responses than in non-responses to fluids, by either technique. After valvular surgery, COtp and CCO correlated (rs0.64, P-0.001, ns48) but fluid-induced changes did not. After coronary artery surgery, COtp and CCO correlated (rs0.81, P-0.001) and changes also did (rs0.55, P-0.001). At fluid-induced CCO increases -20%, the r for changes in cardiac output measured by both techniques was similar after valvular and coronary artery surgery. Thus, COtp and CCO were of similar value in predicting and monitoring fluid responses after both surgery types. This argues against left-sided valvular insufficiencies confounding COtp Published by European Association for Cardio-Thoracic Surgery. All rights reserved. Keywords: Cardiac function; Heart valve; Hemodynamics; Mitral valve repair; Shock (circulatory) 1. Introduction Fluid loading is a common therapeutic intervention after cardiac surgery and accurate cardiac output measurements are necessary to monitor fluid responses. The effect of cardiac output monitoring techniques on judging responses has rarely been addressed w1x. The transpulmonary and primarily left-sided thermodilution technique for cardiac output measurements (COtp) has been used to judge fluid responses w1x. However, the technique is sensitive to thermal loss, re-circulation or forwardybackward movement of the thermal indicator, for instance in the presence of left-sided valvular insufficiencies and regurgitant blood flows w2 4x. Hillis et al. w2x showed that transpulmonary dye-dilution cardiac output underestimates right-sided thermodilution cardiac output in the presence of aortic andyor mitral valve insufficiency. In contrast, the COtp has been compared right-sided measurements and results appeared interchangeable, prior to or after coronary artery surgery and only rarely combined with valvular surgery w1, 3, 5, 6x. The continuous right-sided thermodilution technique for cardiac output measurements in the pulmonary artery *Corresponding author. Department of Intensive Care, VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands. Tel.: q ; fax: q address: johan.groeneveld@vumc.nl (A.B.J. Groeneveld) Published by European Association for Cardio-Thoracic Surgery (CCO) could be less suitable for monitoring fluid responses, because of relatively long response times and, as for the bolus technique, in the presence, perhaps, of tricuspid insufficiencies w1, 7x. Nevertheless, the CCO has been claimed to be an acceptable alternative to the bolus thermodilution in coronary artery surgery patients, sometimes combined with valvular surgery w1, 6, 8 13x. Also, in patients after valvular surgery, the techniques appeared comparable w10x. In the single comparison of COtp and CCO after fluid loading, Hofer et al. w1x found, prior to coronary artery surgery, that techniques were interchangeable. This does not exclude a difference after surgery when hemodynamic stability may be less, however w6, 10 13x. In the current study we evaluated COtp and CCO and their response to fluid loading separately after coronary artery and valvular surgery. The null hypothesis was that the value in predicting and monitoring fluid responses by COtp as compared to CCO does not differ between valvular and coronary artery surgery patients, in spite of residual leftsided valvular insufficiencies in the former. 2. Patients and methods 2.1. Patients Eight consecutive, mechanically ventilated patients undergoing elective valvular repair or replacement, and

2 R.-M.B.G.E. Breukers et al. / Interactive CardioVascular and Thoracic Surgery 9 (2009) 4 8 eight after coronary artery surgery only, -85 years of age, were included 2 h after arrival at the intensive care unit (ICU). Other inclusion criteria were a pulmonary artery occlusion pressure (PAOP) -18 mmhg and central venous pressure (CVP) -13 mmhg and a CCO -8.0 lymin. The exclusion criteria were a life expectancy -24 h, surgical bleeding 100 mlyh, clinical evidence for pulmonary edema, known aneurysms of thoracic or abdominal aorta and atrial fibrillation. All patients underwent hypothermic cardiopulmonary bypass. None of the patients had visible V waves on their CVP curve, arguing against hemodynamically significant tricuspid insufficiency. The Local Ethics Committee approved the study protocol and written informed consent was obtained preoperatively. In the six-month period of inclusion for this study, there were 21 eligible patients who had given informed consent; we excluded three coronary artery surgery patients because of premature discontinuation of fluid loading following fear of overloading, and two valvular surgery patients, before start of study, needing a re-thoracotomy because of bleeding and tamponade. Upon arrival in the ICU, patients were sedated. Patients were mechanically ventilated by volume-controlled ventilation with tidal volumes not exceeding 10 mlykg and a positive end-expiratory pressure (PEEP) of 5 (15 patients) or 7 cm H O(1 valvular surgery patient). Ventilation frequency was breathsymin to maintain the arterial PCO in the 2 normal range. Ventilator settings were not changed during the study. Doses of sedative and vasoactive agents were unaltered during the study in each patient. In sequential atrioventricular pacemaker-dependent patients after surgery, the pacemaker frequency was set at about 80 beatsymin Measurements A pulmonary artery catheter (8.5-Fr, CCOyVIP, Edwards Life Sciences, Irvine, CA, USA) had been preoperatively inserted via the internal jugular vein. This allowed CCO measurements via warm pulses by a thermal filament and downstream detection by the thermistor of the pulmonary artery catheter connected to a computer system (Vigilance, Edwards Life Sciences, Irvine, CA, USA). Thirty seconds to 1 min averages are displayed. In pooled data (ns59), measured at 1 2-min intervals and normalized for individual means, the reproducibility of CCO, expressed as coefficient of variation (CV), was 4.2%. The transpulmonary thermodilution technique (PiCCO, Pulsion Medical Systems, Munich, Germany) was used to determine COtp and global end-diastolic volume (GEDV), an indicator of cardiac filling w3x. Reproducibility of these measurements is about 7% and 10%, respectively w3x. We found in (ns72) triplicate transpulmonary measurements, which were pooled after normalization for individual mean values with a CV of 7.9% Study protocol Within two hours after arrival in the ICU, a 5-Fr thermodilution catheter (Pulsiocath PV 2015; Pulsion Medical Systems, Munich, Germany) was inserted into the femoral artery and connected to the PiCCO-computer for computing CO. For each patient the following data were collected: demographic data, such as weight and height allowing calculation of body mass index (BMI), data to predict surgery-related mortality according to the European system for cardiac operative risk evaluation (EuroSCORE, w14)x, pressures, referenced to atmospheric pressure after calibration and zeroing, at the midchest level, i.e. mean arterial pressure (MAP) and body temperature. In each patient, five consecutive fluid loading steps were done, by R infusing 250 ml of colloid solution (Voluven, Fresenius, Germany) in 15 min. Thirty minutes after start of each fluid loading step (and 15 min after completion), cardiac output was measured by the two methods, CCO preceding COtp. For the latter, three injections of 20 ml of cold (-8 8C) normal saline via the CVP port of the pulmonary artery catheter were given and values were averaged. A transesophageal echocardiogram was obtained in valvular surgery patients after surgery and before fluid loading, and compared with preoperative echocardiograms (ECHOpac and Vivid 7, GE Healthcare, Milwaukee, WI, USA) Statistical analysis We expected cardiac output increases in 30 50% of the 40 steps. Responses and non-responses among steps for either technique were defined by an increase in cardiac output of )5% vs. F5% and G10% vs. -10%, per fluid loading step, to account for the volumes administered and the literature on fluid responsiveness. Generalized estimating equations, taking repeated measurements in the same patients into account, were used to evaluate changes in time, to evaluate values to predict and monitor fluid responses, and to study potential differences between surgery types and techniques herein. Partial correlation coefficients for absolute values and fluid-induced changes, with patient number as covariate to account for repeated measurements in the same patients, were calculated and compared after Z transformation. Bland Altman analysis was used to calculate bias, precision (S.D.) and limits of agreement (bias"2 S.D.). Exact two-sided P)0.001 are given and considered statistically significant when (nssnot significant). Data are summarized as mean"standard deviation (S.D.), since they were normally distributed. 3. Results 3.1. Patient characteristics In both groups mean age, gender and BMI were similar, but the EuroSCORE was higher in valvular surgery (Table 1). Of the valvular surgery patients, four had prior mitral insufficiency, one patient had mitral insufficiency and stenosis, one patient had mitral and aortic insufficiency and two patients had aortic stenosis. After surgery, six patients had some tricuspid insufficiency (grading from minimum to grade 2 out of 4) on echocardiography. In contrast, there was some residual aortic insufficiency (grading from minimum to grade 2 out of 4) in four patients and mitral valve insufficiency (grading from minimum to grade 2 out of 4) in six patients. Four valvular surgery and one coronary artery surgery patient were pacemaker-dependent. 5 New Ideas Case Report Work in Progress Report ESCVS Article Historical Pages Negative Results Follow-up Paper Best Evidence Topic Proposal for Bailout Procedure State-of-the-art Brief Communication Nomenclature

3 6 R.-M.B.G.E. Breukers et al. / Interactive CardioVascular and Thoracic Surgery 9 (2009) 4 8 Table 1 Patient characteristics Age, years Gender, BMI, Surgical Vasoactive drugs EuroSCORE y kgym 2 procedure Valvular surgery 65"12 6y2 26"4 MVP (5), MVR (1), AVP (2), Dobu (3), dopa (1), 7"2 AVR (1), TVP (1), CABG (2), nor (1), enoximone (2) Bentall-procedure (1), aortic re-construction (1) Coronary artery surgery 68"9 6y2 29"3 CABG (2 5x) Dobu (3), dopa (4), 3"2 enoximone (1) MVP, mitral valve plasty; MVR, mitral valve replacement; AVP, aortic valve plasty; AVR, aortic valve replacement; TVP, tricuspid valve plasty; CABG, coronary artery bypass graft; Dobu, dobutamine; Dopa, dopamine; Nor, nor-epinephrine. Mean"S.D., or number, where appropriate Effect of fluid loading Table 2 shows hemodynamics prior to and after the last fluid loading step. Table 3 shows that, for the 40 fluid loading steps involved in this study, the cardiac output was indeed lower after valvular than coronary artery surgery (Ps0.049 for COtp and Ps0.015 for CCO). In responses, the baseline cardiac output was generally lower than in non-responses but increases were similar, regardless of surgery types and techniques. Fluid responsesynonresponses ()5%, F5%) were fairly concordant among cardiac output measurement techniques (62%, ns, for valvular surgery and 68%, Ps0.010 for coronary artery surgery) Relation between COs Prior to, during and after fluid loading after coronary artery and valvular surgery, COtp and CCO correlated at rs0.81 (P-0.001, ns48) and rs0.64 (P-0.001, ns48), respectively. After valvular surgery, fluid-induced changes of COtp and CCO did not correlate (rs0.26, ns, ns40), whereas fluid-induced changes of COtp and CCO correlated after coronary artery surgery (rs0.55, P-0.001, ns40). When arbitrarily selecting observations (ns34) at CCO Table 2 Hemodynamics prior to and after the last fluid loading step Valvular surgery Coronary artery surgery MAP, mmhg Before 78"15 79"11 After 75"8 77"9 GEDV, ml Before 1825" "575 After 1990"541* 1895"657 COtp, lymin Before 6.1" "1.2 After 6.7"1.1** 7.9"1.2*** CCO, lymin Before 4.8" "1.2 After 5.9"1.2**** 6.6"1.2***** Temp, 8C Before 36.8" "0.7 After 37.0" "0.9 MAP, mean arterial pressure; GEDV, global end-diastolic volume; COtp, transpulmonary cardiac output; CCO, continuous cardiac output; temp, body temperature. For after vs. before: *Ps0.002, **Ps0.023, ***P (and Ps0.040 vs. valvular surgery), ****Ps0.002, *****Ps0.009 (and Ps0.024 vs. valvular surgery). Mean"S.D. increases of -20%, the correlation was 0.75 (P-0.001) for absolute values and 0.38 (Ps0.030) for fluid-induced changes after valvular surgery. When arbitrarily selecting observations (ns35) with increases in CCO of -20%, the r was 0.84 (P-0.001) for absolute values and 0.58 (P-0.001) for fluid-induced changes after coronary artery surgery (ns vs. r after valvular surgery) Bland Altman analysis (Fig. 1) After valvular surgery, the bias of the COtp as compared to CCO was 0.94 lymin (precision 0.74 lymin), so that the limits of agreement were lymin (2 S.D.ymean cardiac output of 5.8 lymin equals 25%). After coronary artery surgery, the bias was 1.00 lymin (precision 0.94 lymin), so that the limits of agreement were lymin (2 S.D.ymean cardiac output of 6.9 lymin equals 27%). 4. Discussion This study suggests that, even though cardiac output is lower after valvular than coronary artery surgery, the overestimation of CCO by COtp is comparable, and COtp and CCO are of similar value in predicting and monitoring (relatively small) fluid responses after both surgery types. This argues against residual left-sided valvular insufficiency confounding transpulmonary thermodilution. We chose to compare COtp and CCO, since both may be less affected by respiratory variations of stroke volume in mechanically ventilated patients than the bolus thermodilution technique. The precision and thus limits of agreement of cardiac output measurements, which were similar after valvular and coronary artery surgery, were within 30%, the value proposed to be associated with clinically acceptable errors w15x. The positive bias of COtp vs. pulmonary artery thermodilution cardiac output, often observed before, can be attributed to thermal loss w1, 3 6x. The comparability of the bias irrespective of surgery types argues against thermal loss induced by residual valvular insufficiencies, but we cannot exclude that forwardy backward flows had partly masked greater thermal loss and higher bias, expected at lower cardiac output w4x, after valvular surgery. Even though increases in cardiac output with fluid loading were comparable among techniques, irrespective of surgery type, there was a tendency for underestimation of relatively high (fluid-induced changes

4 R.-M.B.G.E. Breukers et al. / Interactive CardioVascular and Thoracic Surgery 9 (2009) 4 8 Table 3 Responses (R) and non-responses (NR) among fluid loading steps, defined by 5% and F5%, and G10% and -10% increase in cardiac output Valvular surgery P-value Coronary artery surgery P-value R ()5%) NR (F5%) R ()5%) NR (F5%) COtp, lymin ns14 ns26 ns21 ns19 steps Before 5.9" " " " After 6.6" " " "1.5 CCO, lymin ns19 ns21 ns16 ns24 steps Before 5.1" " " " After 5.9" " " "1.7 R (G10%) NR (-10%) R (G10%) NR (-10%) 7 New Ideas Case Report Work in Progress Report COtp, lymin ns9 ns31 ns12 ns28 steps Before 6.0" " " "1.7 After 6.9" " " "1.4 CCO, lymin ns11 ns29 ns11 ns29 Before 5.0" " " " After 6.2" " " "1.6 COtp, transpulmonary cardiac output; CCO, continuous cardiac output. Mean"S.D. Fig. 1. Bland Altman plot of difference between transpulmonary (COtp) and continuous cardiac output (CCO) vs. their mean; (a) after valvular surgery and (b) after coronary artery surgery. The solid line indicates the bias and the dashed line the limits of agreement (bias"2 S.D.). in) cardiac output by COtp vs. CCO after valvular surgery, thereby decreasing the correlation of changes after valvular vs. that after coronary artery surgery. This explains, in part, why correlation coefficients for changes seemed less different between surgery types at relatively low fluidinduced increases. The tendency argues for forwardybackward flows so that an effect of residual left-sided valvular insufficiencies on COtp measurements cannot be completely excluded w2x. Hence, a larger number of study patients might have better revealed such a small effect. In any case, the high reproducibility of CCO and similar fluidinduced changes among techniques upon fluid loading steps argue in favor of the 15 min used in this study to reach a new steady state in CCO, as done before w7x. Also, none of the patients had hemodynamically important tricuspid insufficiency, that could confound right-sided thermodilution. Rapid re-warming in the first 45 min after hypothermic cardiopulmonary bypass possibly confounding the CCO w8, 12x is also not a likely source of error in our study. Indeed, the CCO technique is attractive to monitor fluid responses without the necessity of injections and calibrations as for pulse contour methods, which, in addition, may be confounded by aortic valve abnormalities w1, 6, 10, 13x. The COtp can be favored, however, if considered less invasive than the CCO. Otherwise, the relatively low cardiac output after valvular vs. coronary artery surgery may have been caused by more severe left ventricular dysfunction, as suggested by the lower cardiac output at similar GEDV. Apparently, dysfunction did not hamper similar fluid loading responses as after coronary artery surgery. In conclusion, our data suggest that COtp and CCO are of similar value in predicting and monitoring (relatively small) cardiac output fluid responses after valvular and coronary artery surgery, in spite of lower cardiac output after valvular surgery but comparable overestimation by COtp. Hence, residual left-sided valvular insufficiency may not confound assessment of COtp and its response to fluid loading after valvular surgery. References w1x Hofer CK, Furrer L, Matter-Ensner S, Maloigne M, Klaghofer R, Genoni M, Zollinger A. Volumetric preload measurement by thermodilution: a comparison with transoesophageal echocardiography. Br J Anaesth 2005; 94: w2x Hillis LD, Firth BG, Winniford MD. Comparison of thermodilution and indocyanine green dye in low cardiac output or left-sided regurgitation. Am J Cardiol 1986;57: w3x Gödje O, Peyerl M, Seebauer T, Lamm P, Mair H, Reichart B. Central venous pressure, pulmonary capillary wedge pressure and intrathoracic blood volumes as preload indicators in cardiac surgery patients. Eur J Cardiothorac Surg 1998;13: w4x Hüter L, Schwarzkopf KR, Preussler NP, Schubert H, Schreiber T. The level of cardiac output affects the relationship and agreement between pulmonary artery and transpulmonary aortic thermodilution measurements in an animal model. J Cardiothorac Vasc Anesth 2007;21: w5x Gödje O, Peyerl M, Seebauer T, Dewald O, Reichart B. Reproducibility of indicator dilution measurements of intrathoracic blood volume compartments, extravascular lung water, and liver function. Chest 1998; 113: w6x Mielck F, Buhre W, Hanekop G, Tirilomis T, Hilgers R, Sonntag H. Comparison of continuous cardiac output measurements in patients after cardiac surgery. J Cardiothorac Vasc Anesth 2003;17: w7x Lazor MA, Pierce ET, Stanley GD, Stanley GD, Cass JL, Halpern EF, Bode RH Jr. Evaluation of accuracy and response time of STAT-mode continuous cardiac output. J Cardiothorac Vasc Anesth 1997;11: w8x Böttiger BW, Soder M, Rauch H, Böhrer H, Motsch J, Bauer H, Martin E. Semi-continuous versus injectate cardiac output measurement in inten- ESCVS Article Historical Pages Negative Results Follow-up Paper Best Evidence Topic Proposal for Bailout Procedure State-of-the-art Brief Communication Nomenclature

5 8 R.-M.B.G.E. Breukers et al. / Interactive CardioVascular and Thoracic Surgery 9 (2009) 4 8 sive care patients after cardiac surgery. Intensive Care Med 1996;22: w9x Neto EP, Piriou V, Durand PG, Du Gres B, Lehot JJ. Comparison of two semicontinuous cardiac output pulmonary artery catheters after valvular surgery. Crit Care Med 1999;27: w10x Rödig G, Prasser C, Keyl C, Liebold A, Hobbhahn J. Continuous cardiac output measurement: pulse contour analysis vs. thermodilution technique in cardiac surgical patients. Br J Anaesth 1999;82: w11x Schmid ER, Schmidlin D, Tornic M, Seifert B. Continuous thermodilution cardiac output: clinical validation against a reference technique of known accuracy. Intensive Care Med 1999;25: w12x Bendjelid K, Schütz N, Suter PM, Romand JA. Continuous cardiac output monitoring after cardiopulmonary bypass: a comparison with bolus thermodilution measurement. Intensive Care Med 2006;32: w13x Button D, Weibel L, Reuthebuch O, Genoni M, Zollinger A, Hofer CK. Clinical evaluation of the FloTracyVigileo system and two established continuous cardiac output monitoring devices in patients undergoing cardiac surgery. Br J Anaesth 2007;99: w14x Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9 13. w15x Critchley LAH, Critchley JAJH. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999;15:85 91.