((Measurement of central and mixed venous to-arterial


Published on


Published in: Education, Health & Medicine
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

((Measurement of central and mixed venous to-arterial

  1. 1. Central and mixed venous-to-arterial CO2-differences 29Applied Cardiopulmonary Pathophysiology 15: 29-37, 2011Measurement of central and mixed venous-to-arterialcarbon dioxide differences in cardiac surgery patientsH. Heinze1, H. Paarmann1, M. Heringlake1, H. V. Groesdonk21 Department of Anesthesiology, University of Lübeck, Germany; 2Department of Thoracicand Cardiovascular Surgery, University of Saarland, Homburg, GermanyAbstractBackground: Measurement of central venous-to-arterial CO2 difference (p(cv-a)CO2) as an es-timate of mixed venous-to-arterial CO2 difference (p(v-a)CO2) has been recommended as ansupplementary parameter to identify the inadequacy of tissue oxygenation in septic and post-operative non-cardiac high risk surgery patients. This study investigates the agreement betweenp(cv-a)CO2 and p(v-a)CO2, and explores the relationship of p(v-a)CO2 with parameters of glob-al and regional tissue oxygenation.Methods: Simultaneous measurements of p(cv-a)CO2 and p(v-a)CO2 were performed in post-operative cardiac surgery patients immediately before and after cardiopulmonary bypass(CPB), and up to 6 hours post CPB. In addition, parameters of global blood flow and tissue oxy-genation, i.e. cardiac index, mixed venous oxygen saturation (SvO2), arterial lactate, and region-al blood flow, i.e. gastric tonometry were assessed. Pooled data were used for Bland-Altmanand correlation analysis, as appropriate.Results: Although significantly correlated, p(cv-a)CO2 and p(v-a)CO2 showed large limits ofagreement (6.7 mmHg, percentage error of 115 %). Correlation analyses revealed no mean-ingful correlation between p(v-a)CO2 and CI, SvO2, arterial lactate, and p(g-a)CO2 (R2: 0.013,0.007, 0.000, 0.006, respectively, with p > 0.05 each).Conclusions: In cardiac surgery patients p(cv-a)CO2 cannot be used as an estimate of p(v-a)CO2 with acceptable accuracy. There is no evidence that measurements of p(cv-a)CO2 orp(v-a)CO2 could help diagnose global or regional tissue hypoxia in this patient group.Key words: central-venous-to-arterial carbon dioxide difference, gastric tonometry, mixed ve-nous-to-arterial carbon dioxide difference, arterial lactate, cardiac surgery, mixed venous oxy-gen saturation, tissue oxygenation, hemodynamicIntroduction oxygen delivery/uptake relationship [2]. The physiological limitations of ScvO2, usuallySince the landmark study by Rivers et al. [1] measured in the superior venae cava, are wellmeasurement of central venous oxygen satu- known, and in certain patients large differ-ration (ScvO2) has been extensively used in ences between ScvO2 and SvO2 may occurintensive care and perioperative medicine. [3]. Beside this and the fact that Rivers et al.ScvO2 serves as an estimate of true mixed ve- studied severly septic patients during the firstnous oxygen saturation (SvO2) and therefore 6 hours after hospital admission, ScvO2 meas-
  2. 2. 30 H. Heinze, H. Paarmann, M. Heringlake, H. V. Groesdonkurements have been recommended in peri- tients, investigating intramyocardial oxygenoperative elective cardiac surgery patients monitoring [7].[4]. Following approval by the local ethical Recent studies in high risk major abdomi- committee of the University of Lübeck andnal surgery and septic shock revealed that a written informed consent, 28 patients under-ScvO2 > 70% may not rule out the inadequa- went standard CABG. The data of three pa-cy of tissue oxygenation [5,6]. In this aspect tients experiencing perioperative complica-the authors recommended the measurement tions have been excluded from this analysisof the central venous-to-arterial CO2 differ- and have been published before [8]. There-ence (p(cv-a)CO2) as a complementary target fore 25 remaining patients were analysed.to identify persistent ischaemic hypoxia [5,6].P(cv-a)CO2 is an estimate of mixed venous-to-arterial CO2 difference (p(v-a)CO2) and may Anaesthesia and surgical techniquehave the same limitations as the ScvO2, e.g.the exact sampling site, the haemodynamic Anaesthesia was induced with etomidatecondition, and the level of sedation. Patients (0.3-0.5 mg·kg–1) and sufentanyl (0,5-1after cardiac surgery exhibit certain problems µg·kg–1) and maintained with continuous infu-after hypothermic cardiopulmonary bypass sions of propofol (5-8 mg·kg–1h–1) and sufen-(CPB), namely rewarming, reperfusion, in- tanyl (0.5-1 µg·kg–1h–1). Muscle relaxationflammation, and are often haemodynamically was achieved with pancuronium bromid (0.1compromised to a certain degree. This may mg·kg–1). Patients were mechanically ventilat-have a profound effect on regional and glob- ed in a volume controlled mode with 100%al perfusion in the immediate postoperative oxygen throughout the surgical procedure;period, influencing p(cv-a)CO2. Thus, the respiratory rate was adjusted to achievewhether a widening of the p(cv-a)CO2 in the normocapnia. Intraoperative fluid manage-face of a ScvO2 > 70% could also serve as a ment was adjusted to achieve and maintain acomplementary target to identify inadequate central venous pressure between 8 to 12global or regional tissue oxygenation in car- mmHg and a pulmonary artery capillary oc-diac surgery patients is yet unknown. clusion pressure (PAOP) between 15 to 18 Therefore we re-analyzed a study pub- mmHg. Volume replacement was performedlished before [7] where simultaneous meas- with Ringer’s solution and gelatine polysucci-urements of ScvO2, SvO2, pcvCO2, pvCO2, nate, as appropriate. Naso-pharyngeal andarterial carbon dioxide pressure (paCO2), gas- rectal temperature was continuously meas-tric luminal carbon dioxide pressure (pgCO2), ured during the whole observation time. Stan-cardiac index (CI), and arterial lactate levels dard CABG was performed in moderate hy-were performed. pothermia (32°C naso-pharyngeal) with a The aim of this study was to study the membrane oxygenator (Hilite; Medos, Stol-agreement between p(cv-a)CO2 and p(v- berg, Germany) and a roller pump (Stöckert,a)CO2, and explore the relationship of p(v- München, Germany). The pump was primeda)CO2 with parameters of regional blood with 1.430 ml Ringer`s solution, 250 ml offlow, i.e. gastric tonometry and global tissue 20% mannitol, and 20 ml 8.4% (1 M) natri-oxygenation, namely SvO2, and arterial lac- umbicarbonate. After median sternotomytate levels. and harvesting of the bypass grafts, patients were fully heparinized according to body weight in a dose of 300 IE/kg body weight.Material and methods Activated clotting time was kept greater than 450 seconds. Aortic and two stage venousThis is a re-analysis of a prospective observa- cannulation was used and after cross-clamp-tional study in elective cardiac surgery pa- ing the heart was arrested using antegrade
  3. 3. Central and mixed venous-to-arterial CO2-differences 31cold cristalloid cardioplegia which was re- the tonometer balloon is automatically filledpeated every 20 minutes. Mean arterial with approximately 6 mL air and pCO2 isblood pressure (MAP) during CPB was main- measured every 10 min in a recirculatingtained at 60 to 80 mmHg. Vasopressors (nor- mode [9]. The difference of arterial and gas-epinephrine) were applied, if necessary. tric pCO2 (p(g-a)CO2) was calculated which has been shown to reflect gastro-intestinal stagnant hypoxia secondary to hypovolaemiaMeasurements [10,11]. Negative values of p(g-a)CO2 were set to zero. There was no enteral feeding asIntra- and postoperatively all parameters were long as the tonometric tube was in place. Totaken before CPB and every hour up to 6 rule out intra-gastric CO2 generation follow-hours after CPB. All patients were mechani- ing the buffering of acid with bicarbonate pa-cally ventilated as long as clinically appropri- tients received 300 mg ranitidine [12].ate and were sedated with continuous infu-sion of propofol. Additionally to standard monitoring with a Statisticsthree lead electrocardiogram, a transcuta-neous oxygen sensor, a radial arterial and a Agreement between p(cv-a)CO2 and p(v-central venous line, all patients were aCO2), and ScvO2 and SvO2 were analyzedequipped with a pulmonary artery catheter using the method by Bland and Altman [13].for continuous determination of pulmonary The percentage error was calculated as 1.96artery pressures, automated semicontinuous * SD/mean of reference method [14]. Corre-measurement of cardiac output (CO)/cardiac lations were assessed by calculating Pear-index (CI), and continuous measurement of son’s coefficient (r) and the coefficient of de-mixed venous oxygen saturation (SvO2) (Vig- termination (R2). All data are presented asilance®; Edwards Lifescience, Irvine, USA). Ar- mean (SD) unless stated otherwise. A p valueterial blood samples were drawn for determi- less than 0.05 was regarded significant.nation of oxygen (paO2) and carbon dioxide(PaCO2) tension, base excess (BE), arterial bi-carbonate, and arterial pH (ABL 505 bloodgas analyzer, Radiometer, Copenhagen, Den- Resultsmark), as well as haemoglobin concentrationand arterial lactate concentration. Simultane- The demographic and operative data of theously, central venous and mixed venous studied 25 patients are listed in table 1. Theblood samples were drawn for determination perioperative course of p(v-a)CO2 and p(cv-of pcvCO2, ScvO2, and pvCO2. p(cv-a)CO2 a)CO2 is shown in fig.1. Pooled data of 127and p(v-aCO2) were calculated. simultaneous measurements could be ana- A nasogastric tonometry catheter (TRIP lyzed. P(v-a)CO2 and p(cv-a)CO2 showed aNGS catheter, Tonometric Division, Instru- significant correlation with a correlation in-mentarium, Helsinki, Finnland) was placed in dex of r = 0.600. The Bland-Altman transfor-the stomach. Correct positioning was verified mation revealed a bias of p(cv-a)CO2 of 1.2by auscultation over the gastric region, aspira- (3.4) mmHg, LOA of 6.7 mmHg with a per-tion of gastric contents and radiologically centage error of 152% (fig. 2). SvO2 andwith the next routine chest X-ray. The ScvO2 showed a significant correlation astonometer was connected to a Tonocap® well. Bias of ScvO2 was -0.3 (7.8) % with LOA(Datex, Helsinki, Finnland). The Tonocap® of 15.3 %. This resulted in a percentage errormeasures gastric luminal pCO2 (pgCO2) of 21,9 % (fig. 3).which is in equilibrium with gastric mucosal Correlation analysis between p(v-a)CO2pCO2 by automatic gas capnometry. Briefly, and CI, SvO2, arterial lactate, and p(g-a)CO2
  4. 4. 32 H. Heinze, H. Paarmann, M. Heringlake, H. V. GroesdonkGender [female / male] 7 / 18 Table 1: Demographic and periope- rative dataAge [years] 69 (8)Height [cm] 171 (7)Weight [kg] 77 (14)LVEF [%] 61 (13)Number of grafts performed 2-ACVG [n] 1 2-ACVG + IMA [n] 8 3-ACVG [n] 5 3-ACVG + IMA [n] 10 4-ACVG + IMA [n] 1Duration of surgery [min] 226 (30)Cross clamp time [min] 45 (20) LVEF, left ventricular ejection fraction;Duration of cardiopulmonary bypass [min] 93 (17) ACVG, Aorto coronary venous graft; IMA, Internal mammaria artery graftDuration of ventilation [hours] 13 (5) Figure 1: Perioperative course of p(v-a)CO2 and p(cv-a)CO2revealed no meaningful significant correla- – There is no reasonable association be-tions (see fig. 4). tween p(v-a)CO2 and CI, SvO2, or arterial lactate levels, as markers of the adequacy of global tissue oxygenation, and p(g-Discussion a)CO2, as a marker of regional perfusion. Therefore neither p(v-a)CO2, nor p(cv-The main findings of this study are as follows: a)CO2 measurement can be recommended– In cardiac surgery patients p(cv-a)CO2 can- to rule out tissue hypoxia in postoperative not be used as an estimate of p(v-a)CO2 cardiac surgery patients. with acceptable accuracy.
  5. 5. Central and mixed venous-to-arterial CO2-differences 33Figure 2: Correlation and Bland-Altman analyses between p(cv-a)CO2 or p(v-a)CO2.p(cv-a)CO2: Central venous-to-arterial CO2 difference; p(v-a)CO2: Venous-to-arterial CO2 difference;LOA: Limits of agreement (1.96 * SD); p: Significance; R2: Coefficient of determination (Pearson).Figure 3: Correlation and Bland-Altman analyses between ScvO2 and SvO2.ScvO2: Central venous oxygen saturation; SvO2: Mixed venous oxygen saturation; LOA: Limits ofagreement (1.96 * SD); p: Significance; R2: Coefficient of determination (Pearson). These results are in contrast with some of crease very much during reperfusion after hy-the existing literature. In a heterogenous pothermic CPB. This will result in false lowgroup of critically ill patients Cuschieri et al. CO2-differences when p(cv-a)CO2 is meas-showed a strong agreement between p(v- ured to estimate p(v-a)CO2. In our opiniona)CO2 and p(cv-a)CO2 measurements [15]. In p(cv-a)CO2 measured as an estimate of p(v-addition, they revealed a high correlation be- a)CO2 should be used very cautiously, at leasttween p(cv-a)CO2 and CI. There are possible in cardiac surgery patients.reasons for these different results. Unfortu- There is clear experimental evidence, thatnately, p(cv-a)CO2 mostly represents venous stagnant tissue hypoxia is associated with acarbon dioxide from the upper part of the widening of the venous-to-arterial pCO2-dif-body, even if the tip of the central venous ference, depending on the region of interestcatheter is located near the right atrium. [17-19]. In the present study no clear relation-Cerebral CO2-production is very different be- ship between p(v-a)CO2 and markers of glob-tween awake and sedated states, not neces- al tissue oxygenation, as SvO2 and arterial lac-sarily reflecting different states of cardiac per- tate levels, or CI could be demonstrated. Oneformance [2,16]. In addition, CO2-production reason could be the inevitable inaccuracy offrom the lower part of the body might in- blood gas analyzers, accounting for up to
  6. 6. 34 H. Heinze, H. Paarmann, M. Heringlake, H. V. GroesdonkFigure 4: Correlation between p(v-a)CO2 and CI, SvO2, arterial lactate levels, and p(g-a)CO2.p(v-a)CO2: Venous-to-arterial CO2 difference; CI: Cardiac index; SvO2: Mixed venous oxygen satura-tion; p(g-a)CO2: Gastric-to-arterial CO2 difference; p: Significance; R2: Coefficient of determination(Pearson).10% bias. In contrast to laboratory studies in tients undergoing major abdominal high riskclinical investigations much less parameters surgery [6]. 15% of the patients showed a Sc-effecting CO2-production, eg. exact body vO2 > 71%, but developed major complica-temperature, level of sedation, and alveolar tions. A p(cv-a)CO2 of more than 5mmHgventilation, can be controlled for. Especially had the most predictive threshold value inbody temperature is a very crucial point in these patients [6]. That patients with a ScvO2postoperative cardiac surgery patients. There > 70% still have an inadaquacy of regional oris a high interindividual variability in the rate global tissue oxygenation may be due to theof rewarming of peripheral tissue beds after inevitable limitations of the parameter [20-hypothermic CPB, resulting in a varying de- 22]. For example, stagnant tissue hypoxia ingree of CO2-production. These factors might the lower part of the body may be unrecog-even be worse in clincial routine, where eg. nized. While ScvO2 measurements are an es-different blood gas analyzers are used, possi- tablished monitoring parameter in septic pa-bly increasing the inaccuracy. tients [1], many questions have not been an- Vallée et al. demonstrated in patients with swered for other patient groups [21,23,24].septic shock that some degree of tissue hy- Sander et al. studied cardiac surgery patientspoxia might be present even with ScvO2 val- and revealed a reasonable correlation andues > 70% [5]. In their study these patients bias between SvO2 and ScvO2 measure-could be identified by a p(cv-a)CO2 of more ments, but large limits of aggreement [23].than 6 mmHg [5]. Futier et al. reported on pa- They conclude, that under certain conditions
  7. 7. Central and mixed venous-to-arterial CO2-differences 35these parameters may differ significantly. We mate of p(v-a)CO2 with acceptable accuracy.confirmed these results with the present There is no evidence that measurements ofstudy showing a high, significant correlation, p(cv-a)CO2 or p(v-a)CO2 could help diagnoseand a low bias, but large LOA between the global or regional tissue hypoxia in this pa-two parameters. Perz et al. demonstrated that tient group. Unless more information on thein elective cardiac surgery patients a low relationship between central-venous parame-(< 60 %) and a supranormal (> 77 %) ScvO2 ters with other parameters of tissue oxygena-were associated with an unfavourable out- tion and especially the therapeutic implica-come, while patients with normal values (61- tions of pathological values of these parame-76 %) showed no complications. Values for ters are known, p(cv-a)CO2 or p(v-a)CO2 andp(cv-a)CO2 were not reported [25]. ScvO2 should be interpreted very cautiously. Thus, up to now ScvO2 monitoring shouldbe used very cautiously, especially in cardiacsurgery patients. Financial Support In the present study no reasonable rela- This study was supported by institutional re-tionship between p(v-a)CO2 and lactate or sources of the Department of Anesthesiolo-p(g-a)CO2 could be demonstrated. While ar- gy, University of Luebeck, Germany.terial lactate is a well accepted routine mark-er of global tissue oxygenation [26], gastrictonometry has been critizised of being un- Referencesspecific to tissue hypoxia in critically ill pa- 1. Rivers E, Nguyen B, Havstad S, Ressler J,tients [27]. In addition, interventional trials Muzzin A, Knoblich B, Peterson E, Tom-aimed at improving an abnormal gastric mu- lanovich M. Early goal-directed therapy incosal pH, showed no impact on outcome the treatment of severe sepsis and septic[28]. Therefore, routine use of gastric tonom- shock. N Engl J Med 2001; 345: 1368-1377etry is not recommended [29]. Nevertheless, 2. Shepherd SJ, Pearse RM. Role of central andgastric tonometry has been proven to be an mixed venous oxygen saturation measure-early warning sign of low regional oxygen de- ment in perioperative care. Anesthesiologylivery [27,30-32]. As neither lactate nor p(g- 2009; 111: 649-656a)CO2 increased in parallel with widening of 3. Heinze H, Hanke T, Misfeld M. A mysteri-p(v-a)CO2, this parameter cannot be recom- ous difference between central and mixedmended for routine use in cardiac surgery pa- venous oxygen saturation after cardiac sur- gery. Applied Cardiopulmonary Pathophysi-tients, until further information about thresh- ology 2008; 12: 20-22olds, sensivity, and timing are elaborated. 4. Carl M, Alms A, Braun J, Dongas A, Erb J, There are limitations of the study. Goetz A, Goepfert M, Gogarten W, Grosse The studied patients group is very small, J, Heller AR, Heringlake M, Kastrup M,and the study is a re-analysis of an older Kroener A, Loer SA, Marggraf G, Markewitzstudy. Although the data were prospectively A, Reuter D, Schmitt DV, Schirmer U,collected, they were not intended to evaluate Wiesenack C, Zwissler B, Spies C. S3 guide-the relationship between p(v-a)CO2 and p(cv- lines for intensive care in cardiac surgery pa-a)CO2. Therefore no definite conclusions can tients: hemodynamic monitoring and car-be drawn. diocirculatory system. Ger Med Sci; 8: No patients showed an unfavourable Doc12 5. Vallee F, Vallet B, Mathe O, Parraguette J,course. As can be seen in the Bland Altman Mari A, Silva S, Samii K, Fourcade O, Gen-diagram, a wide range of p(v-a)CO2 and SvO2 estal M. Central venous-to-arterial carbonvalues were observed. But these did not dioxide difference: an additional target forseem to have an effect on outcome. goal-directed therapy in septic shock? Inten- We conclude that in cardiac surgery pa- sive Care Med 2008; 34: 2218-2225tients p(cv-a)CO2 cannot be used as an esti-
  8. 8. 36 H. Heinze, H. Paarmann, M. Heringlake, H. V. Groesdonk 6. Futier E, Robin E, Jabaudon M, Guerin R, Pe- and association with outcome. Crit Care tit A, Bazin JE, Constantin JM, Vallet B. Cen- 2005; 9: R694-699 tral venous O2 saturation and venous-to-ar- 17. Vallet B, Teboul JL, Cain S, Curtis S. Venoar- terial CO2 difference as complementary terial CO(2) difference during regional is- tools for goal-directed therapy during high- chemic or hypoxic hypoxia. J Appl Physiol risk surgery. Crit Care; 14: R193 2000; 89: 1317-1321 7. Misfeld M, Heinze H, Sievers HH, Kraatz 18. Neviere R, Chagnon JL, Teboul JL, Vallet B, EG. Intramyocardial oxygen monitoring in Wattel F. Small intestine intramucosal coronary artery bypass surgery. J Clin Monit PCO(2) and microvascular blood flow dur- Comput 2004; 18: 93-101 ing hypoxic and ischemic hypoxia. Crit Care 8. Kraatz E, Uhlig T, Heinze H. The connection Med 2002; 30: 379-384 between myocardial oxygen partial pressure 19. Groeneveld AB. Interpreting the venous-ar- and intragastric CO2 – three cases. Anäs- terial PCO2 difference. Crit Care Med thesiol Intensivmed Notfallmed 1998; 26: 979-980 Schmerzther 1998; 33 (Suppl 2): S106-108 20. Dueck MH, Klimek M, Appenrodt S, 9. Creteur J, De Backer D, Vincent JL. Monitor- Weigand C, Boerner U. Trends but not indi- ing gastric mucosal carbon dioxide pressure vidual values of central venous oxygen satu- using gas tonometry: in vitro and in vivo val- ration agree with mixed venous oxygen sat- idation studies. Anesthesiology 1997; 87: uration during varying hemodynamic condi- 504-510 tions. Anesthesiology 2005; 103: 249-25710. Chapman MV, Mythen MG, Webb AR, Vin- 21. Varpula M, Karlsson S, Ruokonen E, Pettila cent JL. Report from the meeting: Gastroin- V. Mixed venous oxygen saturation cannot testinal Tonometry: State of the Art. Inten- be estimated by central venous oxygen sat- sive Care Med 2000; 26: 613-622 uration in septic shock. Intensive Care Med11. Schlichtig R, Mehta N, Gayowski TJ. Tissue- 2006 arterial PCO2 difference is a better marker 22. Marx G, Reinhart K. Venous oximetry. Curr of ischemia than intramural pH (pHi) or ar- Opin Crit Care 2006; 12: 263-268 terial pH-pHi difference. J Crit Care 1996; 23. Sander M, Spies CD, Foer A, Weymann L, 11: 51-56 Braun J, Volk T, Grubitzsch H, von Heymann12. Bams JL, Kolkman JJ, Roukens MP, Douma C. Agreement of central venous saturation DP, Loef BG, Meuwissen SG, Groeneveld and mixed venous saturation in cardiac sur- AB. Reliable gastric tonometry after coro- gery patients. Intensive Care Med 2007 nary artery surgery: need for acid secretion 24. Pearse RM, Hinds CJ. Should we use central suppression despite transient failure of acid venous saturation to guide management in secretion. Intensive Care Med 1998; 24: high-risk surgical patients? Crit Care 2006; 1139-1143 10: 18113. Bland JM, Altman DG. Statistical methods 25. Perz S, Uhlig T, Kohl M, Bredle DL, Reinhart for assessing agreement between two meth- K, Bauer M, Kortgen A. Low and „supranor- ods of clinical measurement. Lancet 1986; mal“ central venous oxygen saturation and 1: 307-310 markers of tissue hypoxia in cardiac surgery14. Critchley LA, Critchley JA. A meta-analysis patients: a prospective observational study. of studies using bias and precision statistics Intensive Care Med 37: 52-59 to compare cardiac output measurement 26. Bakker J. Lactate: May I have your votes techniques. J Clin Monit Comput 1999; 15: please? Intensive Care Med 2001; 27: 6-11 85-91 27. Hamilton MA, Mythen MG. Gastric tonom-15. Cuschieri J, Rivers EP, Donnino MW, Katilius etry: where do we stand? Curr Opin Crit M, Jacobsen G, Nguyen HB, Pamukov N, Care 2001; 7: 122-127 Horst HM. Central venous-arterial carbon 28. Gomersall CD, Joynt GM, Freebairn RC, dioxide difference as an indicator of cardiac Hung V, Buckley TA, Oh TE. Resuscitation of index. Intensive Care Med 2005; 31: 818- critically ill patients based on the results of 822 gastric tonometry: a prospective, random-16. Pearse R, Dawson D, Fawcett J, Rhodes A, ized, controlled trial. Crit Care Med 2000; Grounds RM, Bennett ED. Changes in cen- 28: 607-614 tral venous saturation after major surgery,
  9. 9. Central and mixed venous-to-arterial CO2-differences 3729. Antonelli M, Levy M, Andrews PJ, Chastre J, 32. Marik P. Gastric tonometry: the canary sings Hudson LD, Manthous C, Meduri GU, once again. Crit Care Med 1998; 26: 809- Moreno RP, Putensen C, Stewart T, Torres A. 810 Hemodynamic monitoring in shock and im- plications for management. International Consensus Conference, Paris, France, 27-28 April 2006. Intensive Care Med 2007; 33: 575-590 Correspondence address:30. Kolkman JJ, Otte JA, Groeneveld AB. Gas- Hermann Heinze, M.D. trointestinal luminal PCO2 tonometry: an Department of Anesthesiology update on physiology, methodology and University of Lübeck clinical applications. Br J Anaesth 2000; 84: Ratzeburger Allee 160 74-86 23538 Lübeck31. Mythen MG, Webb AR. Gastrointestinal Germany tonometry comes of age? Br J Anaesth 1998; 81: 667-668 hermann.heinze@uk-sh.de