Cardiovasc Eng DOI 10.1007/s10558-010-9113-0

ORIGINAL PAPER

Afterload Assessment With or Without : A Preliminary Clinical Comparison

Glen Atlas • Jay Berger • Sunil Dhar

Ó Springer Science+Business Media, LLC 2010

Abstract A clinical comparison, of two methods of both Ci and MAP, without CVP, may be adequate for use afterload assessment, has been made. The first method, in assessing afterload. systemic index (SVRi), is based upon the traditional formula for afterload which utilizes central Keywords Systemic vascular resistance index Total venous pressure (CVP), as well as (Ci), and systemic vascular resistance index Central venous mean arterial (MAP). The second method, pressure occlusion catheter total systemic vascular resistance index (TSVRi), also uses MAP and Ci. However, TSVRi ignores the contribution of Introduction CVP. This preliminary examination, of 10 randomly- selected ICU patients, has shown a high degree of corre- Examining SVRi lation (ranging from 90 to 100%) between SVRi and TSVRi (P \ 0.0001). Furthermore, there was also a high Clinicians have typically used systemic vascular resistance degree of correlation (ranging from 94 to 100%) noted index (SVRi) as a measure of afterload in the management between the hour-to-hour change in SVRi with the hour-to- of critically ill patients (Melo and Peters 1999) as well as hour change in TSVRi (P \ 0.0001). The results, of this those undergoing extensive cardiac, vascular, and thoracic pilot study, support the premise that the use of CVP may surgical procedures (Barash et al. 2009). SVRi is defined as not always be necessary for afterload evaluation in the (Li 2000, 2004): clinical setting. Minimally-invasive means of measuring ðMAP CVPÞ SVRi ¼ 80 ð1Þ Ci where MAP is , CVP is central G. Atlas (&) J. Berger venous pressure, and Ci is cardiac index. Note that the Department of Anesthesiology, University of Medicine constant 80 allows for SVRi to be expressed with the and Dentistry of NJ, Newark, NJ, USA dimensions of: dyne s cm-5 m-2 or ‘‘resistance units’’. A e-mail: [email protected] list of abbreviations, dimensions, and associated termi- G. Atlas nology is shown in Table 1. Department of Chemistry, Chemical Biology, and Biomedical Traditionally, the determination of SVRi has required Engineering, Stevens Institute of Technology, Hoboken, NJ, invasive hemodynamic monitoring with a pulmonary artery USA occlusion catheter (PAOC) to provide measurements of

S. Dhar both Ci and CVP. Note that MAP can be obtained nonin- Department of Mathematical Sciences, New Jersey Institute vasively, with a blood pressure cuff, or in a minimally- of Technology, Newark, NJ, USA invasive manner with a peripherally-placed arterial 1 S. Dhar catheter. Center for Applied Mathematics and Statistics, New Jersey Institute of Technology, Newark, NJ, USA 1 Typically, the radial artery is used for this purpose. 123 Cardiovasc Eng

Table 1 List of abbreviations, Abbreviation Meaning Dimensions dimensions, and terminology a -2 BMi Body mass index kg m CAD Coronary artery disease b -1 -2 Ci Cardiac index l min m CVP Central venous pressure mmHg DBP Diastolic blood pressure mmHg DM Diabetes mellitus EDM Esophageal Doppler monitor EF % HTN Hypertension MAP Mean arterial pressure mmHg OSA Obstructive sleep apnea PAOC Pulmonary artery occlusion catheter SBP Systolic blood pressure mmHg a Based upon patient height b -5 -2 SVRi Systemic vascular resistance index dyne s cm m (resistance units) b Based upon patient body TSVR Total systemic vascular resistance indexb dyne s cm-5 m-2 (resistance units) surface area i

The clinical utility of PAOCs has been questioned as Examining TSVRi outcome studies have generally not supported their use (ASA 2003; Sandham et al. 2003; NHLBI 2006; Warzawski Another method of afterload assessment can be defined as et al. 2003). Furthermore, minimally-invasive and non- total systemic vascular resistance (TSVRi) in which the invasive means of measuring Ci have emerged. These contribution of CVP is ignored (Wang et al. 2008; Atlas technologies include the esophageal Doppler monitor 2008): Ò (EDM) or CardioQ (Deltex Medical, UK) and the Flo- MAP TracÒ (Edwards Lifesciences, USA).2 Note that the EDM TSVRi ¼ 80 ð2Þ Ci determines Ci using the velocity of blood flow in the distal (Dark and Singer 2004). Whereas the FlowTracÒ Thus, the purpose of this study was to assess the requires a peripheral arterial catheter to measure correlation of SVRi to TSVRi using the data acquired from pressure and uses a ‘‘pulse contour analysis’’ method to PAOCs. Therefore, the contribution of CVP, to afterload estimate (Manecke 2005). The assumptions assessment, could be made. In addition, the hour-to-hour made with pulse contour analysis may be erroneous as changes in both SVRi and TSVRi were also examined. This clinical comparisons, of this device with the EDM, have was done to determine how changes in CVP effect changes demonstrated significant inconsistencies in in afterload. determination (Chatti et al. 2009). Ultimately, the use of minimally-invasive and non- Measurement of the change in electrical impedance or invasive devices could possibly be utilized, for the ‘‘bioimpedance’’ of the thorax, or whole-body, as a func- assessment of afterload, without the potential complica- tion of a time, has also been developed to non-invasively tions associated with pulmonary artery cannulation. These complications are summarized in Table 2 (ASA assess Ci (Barin et al. 2000; Ouzounian et al. 1996). Unfortunately, this method has also demonstrated incon- 2003). sistent correlation with traditional methods of measuring cardiac output (Imhoff et al. 2000; Simon et al. 2009). Differential Analysis None of these minimally-invasive techniques, of Ci evaluation, allow for CVP measurement. An ideal method A differential examination of both SVRi and TSVRi was of afterload assessment would therefore also be minimally also made to assess the hour-to-hour changes in these invasive and allow for continuous measurements. It should values. It should be noted that clinicians frequently be noted that PAOCs only allow for determination of SVRi examine the ‘‘trend’’ in afterload over time. Thus, the hour- on an intermittent basis. to-hour changes in SVRi and TSVRi are defined as: DSVR ðkÞ¼SVR ðk þ 1ÞSVR ðkÞ; i i i ð3Þ k ¼ 1; 2...ðN 1Þ 2 Atys Medical (France) also manufactures an EDM. However, it is not currently available in the USA. and 123 Cardiovasc Eng

Table 2 List of complications associated with PAOCs (ASA 2003) Care Unit (CTICU) at University Hospital/Newark. Note Complication Incidence in most that this study was strictly a retrospective chart review of studies (%) preexisting data. Each patient’s demographic information is shown in Table 3. Arterial puncture B3.6 Hemodynamic data, from all 10 patients, were initially Pneumothorax 0.3–1.9 obtained each hour for at least 24 h. These data included: Minor dysrthymias [20 Ci, CVP, and MAP. Note that MAP was determined from Severe dysrthymias (ventricular 0.3–3.8 either an indwelling arterial catheter or a non-invasive tachycardia or fibrillation) blood pressure cuff. For the purposes of this study, MAP Pulmonary artery rupture 0.03–0.7 was calculated using (Li 2000, 2004): Positive catheter-tip cultures C19 2 1 Catheter-related sepsis 0.7–3.0 MAP = DBP + SBP ð5Þ Venous thrombosis 0.5–3.0 3 3 Pulmonary infarction 0.1–2.6 As this was a preliminary or ‘‘pilot’’ study, no attempt to Valvular/endocardial vegetations or endocarditis 2.2–7.1 stratify data, based upon patient acuity, severity, or survival, was made. Furthermore, the hemodynamic influence, of any of each patient’s medications, such as sedatives or vasoactive medications, was also not included. DTSVRiðkÞ¼TSVRiðk þ 1ÞTSVRiðkÞ; ð4Þ SVRi and TSVRi were calculated, for each hour, using k ¼ 1; 2...ðN 1Þ Eqs. 1 and 2 respectively. Similarly, DSVRi and DTSVRi Note that k ranges from the first hour to the hour preceding were also calculated, on an hourly basis, using Eqs. 3 and 4. the final hour (N - 1). Where N represents the final hour.

The mathematical relationship, between DSVRi and

DTSVRi, using both an algebraic method as well as the 3000.0 SVR and TSVR definition of the total differential, is documented in the i i

) 2500.0 Patient #1 Appendix. -2 ·m

-5 2000.0

Methods 1500.0

(dyne·s·cm 1000.0 SVRi Resistance Units Data Acquisition and Analysis TSVRi 500.0 1357911131517192123 25 After institutional review board approval, the medical Time records of 10 randomly-selected patients, whose treatment (hours) had included indwelling PAOCs, were examined. All Fig. 1 SVRi and TSVRi calculated, on an hourly basis, using the data patients had been admitted to the Cardiothoracic Intensive from patient #1

Table 3 Demographic data of the ten randomly-selected cardiothoracic ICU patients

Patient Age Gender Weight Height BMi HTN CAD DM Tobacco OSA EF Procedure (years) (kg) (m) (kg m-2) use

1 46 Male 116 1.85 33.64 ??? 50% Ascending aortic aneurysm 2 65 Male 124 1.80 38.07 ?? ?55–60% Esophagectomy 3 64 Male 79 1.72 26.30 ?? ? 55% Coronary artery bypass graft(s)/ replacement 4 76 Male 73 1.70 25.06 ???? 75% Coronary artery bypass graft(s) 5 72 Female 57 1.57 22.86 ??? 70% replacement 6 47 Female 94 1.65 34.61 ??? 25–30% Coronary artery bypass graft(s)/mitral valve replacement 7 53 Male 104 1.52 44.91 ?? ?56–60% Mitral valve repair 8 70 Male 106 1.72 35.58 ???? 50% Coronary artery bypass graft(s) 9 62 Male 70 1.72 23.41 ?? 40–45% Coronary artery bypass graft(s) 10 57 Female 80 1.75 25.99 ?? ? 50% Mitral valve replacement

123 Cardiovasc Eng

Results

Based upon this preliminary study, SVRi and TSVRi are closely correlated. This is exemplified in both Figs. 1 and 2 which graphically demonstrate the data from patient #1. The correlation coefficients, from all ten patients’ data, ranged from 90 to 100%. Furthermore, all correlation coefficients were statistically significant (P \ 0.0001).

Similarly, DSVRi and DTSVRi, which represent the hour-to-hour changes in SVRi and TSVRi, were also highly

Fig. 2 A linear correlation, between SVRi and TSVRi, is illustrated correlated. These correlation coefficients ranged from 94 to using the data from patient #1 100% (P \ 0.0001). Figures 3 and 4, which are also based upon the data from patient #1, illustrate this relationship. Δ Δ All correlation coefficients and their associated 95% SVRi and TSVRi Patient #1 confidence intervals are documented in Table 4. ) -2 ·m -5 Discussion

The assessment of afterload remains an important aspect of (dyne·s·cm Resistance Units hemodynamic monitoring during anesthesia and critical care situations. Specifically, clinical conditions such as: Time septicemia, adrenal insufficiency, and pancreatitis fre- (hours) quently require afterload measurement (Melo and Peters 1999). In addition, cardiac, thoracic, and vascular surgical Fig. 3 Hour-to-hour changes in SVRi and TSVRi using the data from patient #1. These changes are referred to as DSVRi and DTSVRi procedures also may necessitate intraoperative and post- respectively operative afterload assessment (Barash et al. 2009). Furthermore, severely injured patients, particularly Statistical analysis was accomplished, on a patient-by- those with neurogenic shock (Bilello et al. 2003) or with patient basis, using a Pearson product-moment correlation systemic inflammatory response syndrome (SIRS) (Cremer coefficient (Kutner et al. 2004; Kreyszig 1999). In addition, a et al. 1996; Carrel et al. 2000; Kohsaka et al. 2005), may Fisher’s z-transformation was also used to determine statis- require afterload measurements. Other conditions requiring tical significance, and associated confidence intervals, from possible afterload assessment include: hypertensive crisis, the correlation coefficients derived from each patient’s data congestive failure, aortic dissection, hepatic trans- (Hawkins 1989). Furthermore, it was assumed that all data plantation, burn injury, and the management of pheochro- points were from a bivariate normal distribution. mocytoma (Barash et al. 2009).

Fig. 4 A linear correlation, between DTSVRi and DSVRi,is documented using the data from patient #1 Δ

Δ

123 Cardiovasc Eng

However, the use of PAOCs has been associated with questionable benefit as well as significant morbidity, mor-

(0.88, 0.97) (0.98, 1.00) tality, and economic cost (ASA 2003; Domino et al. 2004; Stewart et al. 1998). In particular, the rising incidence of 0.94* 0.99* antibiotic-resistant nosocomial infections may further question, and limit, the use of these highly invasive hemo- dynamic monitors (Blot et al. 2005; Richards et al. 2000).

(0.98, 1.00) (0.99, 1.00) Minimally-invasive hemodynamic monitors, specifically

0.99* 1.00* the EDM, have become well established as reliable, safe,

i and accurate (Madan et al. 1999; DiCorte et al. 2000; Dark

SVR and Singer 2004). Furthermore, the EDM can be placed D either orally, in intubated patients, or nasally, in awake (0.87, 0.97) (0.93, 0.98) patients (Atlas and Mort 2001; Dodd 2002). Whereas the

versus Ò i 0.93* 0.97* FlowTrac requires a peripheral arterial cannula. It should be noted that both of these instruments provide approxi- TSVR

D mations of cardiac output. However, in clinical trials, the EDM has been shown to provide reproducible and reliable (0.91, 0.98) (0.93, 0.99) and i measurements of cardiac output (Singer et al. 1989; Valtier 0.96* 0.97* et al. 1998). In addition, the EDM is extremely accurate in measuring changes in cardiac output (DiCorte et al. 2000;

versus SVR Dark and Singer 2004). i

(0.99, 1.00) (0.99, 1.00) Therefore, the ability to assess afterload, with mini- mally-invasive devices, without CVP measurement, could 1.00* 1.00* offer significant benefit in patient safety. This preliminary study has shown that CVP appears to play a minor role in afterload assessment. Furthermore, the hour-to-hour chan- (0.93, 0.99) (0.94, 0.99) ges in afterload appear to be minimally affected by the

0.97* 0.97* exclusion of CVP. However, specific clinical conditions may nonetheless require CVP measurement for management purposes. These would include such pathological states as: superior (0.97, 0.99) (0.99, 1.00) vena cava syndrome, congenital cardiac anomalies, intra-

0.99* 0.99* cardiac tumors, and cardiac tamponade. These conditions can create significant increases in CVP by restricting venous blood flow (Magder 2005).

(0.83, 0.97) (0.85, 0.97) Furthermore, jugular venous distention (JVD), which is a diagnostic ‘‘hallmark’’ of CHF, is also associated with an 0.92* 0.94* increase in central venous pressure (Drazner et al. 2001). However, invasive monitoring of CVP is rarely necessary for the clinical diagnosis and management of this common

(0.77, 0.95) (0.91, 0.98) condition (Senni et al. 1998).

(in bold), and the corresponding 95% confidence intervals for both TSVR Certain clinical situations may artificially alter CVP. The 0.90* 0.96* R use of positive end-expiratory pressure (PEEP), for patients with severe pulmonary disease, is one example (Magder 2005). In addition, the simultaneous use of the central venous (0.79, 0.96) (0.87, 0.98) cannula, for intravenous fluid administration during CVP

Patient 12345678910 0.91* 0.95* measurements, may produce erroneous data (Ho et al. 2005).

i Furthermore, CVP may not be meaningful for patients who are i

SVR in the prone or lateral decubitus positions (Soliman et al. D 1998). A malpostioned pressure transducer can also signifi- The correlation coefficient, versus

i cantly change CVP measurements (Magder 2005). Therefore, versus SVR 0.0001 i

\ in these clinical situations, TSVRi may be advantageous over P TSVR TSVR D Table 4 * SVRi due to the possible inaccuracy of CVP. 123 Cardiovasc Eng  Continuous afterload assessment would also be useful in CVP1 CVP2 CVP1 CVP2 ‘‘highly critical’’ situations. These would possible include thus: 0 ð4AÞ Ci1 Ci2 Ci1 Ci2 patients on potent vasoactive medications for treatment during shock or sepsis. Patients undergoing removal of a Finally: pheochromocytoma may also benefit from this. DSVRi DTSVRi ð5AÞ Lastly, other hemodynamic parameters and terminology, associated with afterload assessment, exist. These include: This concept can also be illustrated using the definition pulse wave velocity, impedance, and characteristic of the total differential: impedance (Segers et al. 2007; Nichols and O’Rourke oSVRi oSVRi oSVRi DSVRi ¼ DMAP þ DCi þ DCVP 2005; Milnor 1989). However, SVR and SVRi remain the oMAP oCi oCVP ‘‘mainstays’’ of clinical afterload measurement within the ð6AÞ hospital-based clinical community. and

oTSVRi oTSVRi DTSVRi ¼ DMAP þ DCi oMAP oCi ð7AÞ Conclusions

oSVRi This preliminary study has favorably compared, and Since oCVP DCVP 0 then DSVRi DTSVRi. Also: o o o o correlated, afterload assessment with CVP measurement SVRi TSVRi and SVRi TSVRi. oMAP oMAP oCi oCi (SVRi) to afterload assessment without CVP measurement (TSVRi). Further research, which would include pro- spective trials to assess the benefits, risks, and limitations of TSVRi measurement, appear indicated. In addition, References these studies could be carried out with minimally-invasive hemodynamic monitors. A clinical comparison, of various American Society of Anesthesiologists (ASA). Task force on means of afterload assessment, could then be imple- guidelines for pulmonary artery catheterization. Practice guide- lines for pulmonary artery catheterization. Anesthesiology. 2003; mented. Thus, a comparison of SVRi to TSVRi, with 99(4):988–1014. respect to patient outcome, complications, and cost, could Atlas GM. Development and application of a logistic-based systolic be made. model for hemodynamic measurements using the esophageal Doppler monitor. Cardiovasc Eng. 2008;8(3):159–73. Atlas G, Mort T. Placement of the esophageal Doppler ultrasound monitor probe in awake patients. Chest. 2001;119:319. Milnor WR. Vascular impedance. In: . 2nd ed. Appendix Baltimore: Williams & Wilkins;(1989). pp. 167–203. Barash PG, Cullen BF, Stoelting RK. Clinical anesthesia. 6th ed. The Interrelationship Between DSVR and DTSVR Phila: Lippincott Williams and Wilkins; 2009. i i Barin E, Haryadi DG, Schookin SI, Westenskow DR, Zubenko VG, Beliaev KR, Morozov AA. Evaluation of a thoracic bioimped- DSVRi is calculated using the definition of SVRi from ance cardiac output monitor during cardiac catheterization. Crit Eq. 1: Care Med. 2000;28(3):698–702. Bilello JF, Davis JW, Cunningham MA, Groom TF, Lemaster D, Sue ðÞMAP2 CVP2 ðÞMAP1 CVP1 LP. Cervical spinal cord injury and the need for cardiovascular DSVRi ¼ 80 80 Ci Ci intervention. Arch Surg. 2003;138:1127–9. 2 1 Blot SI, Depuydt P, Annemans L, Benoit D, Hoste E, De Waele JJ, ð1AÞ Decruyenaere J, Vogelaers D, Colardyn F, Vandewoude KH. Clinical and economic outcomes in critically ill patients with Rearrangement yields: nosocomial catheter-related bloodstream infections. Clin Infect Dis. 2005;41:1591–8. MAP2 MAP1 CVP1 CVP2 DSVRi ¼ 80 þ 80 Carrel T, Englberger L, Mohacsi P, Neidhart P, Schmidli J. Low Ci2 Ci1 Ci1 Ci2 systemic vascular resistance after cardiopulmonary bypass: inci- ð2AÞ dence, etiology, and clinical importance. J Card Surg. 2000;15: 347–53. Substituting the definition of TSVRi from Eq. 2 results in: Chatti R, de Rudniki S, Marque´ S, Dumenil AS, Descorps-Decle`re A,  Cariou A, Duranteau J, Aout M, Vicaut E, Cholley BP. Compar- CVP CVP Ò DSVR ¼ DTSVR þ 1 2 80 ð3AÞ ison of two versions of the Vigileo- system (1.03 and 1.07) for i i stroke volume estimation: a multicentre, blinded comparison with Ci1 Ci2 oesophageal Doppler measurements. Br J Anaesth. 2009;102: Realizing that: 463–9.

123 Cardiovasc Eng

Cremer J, Martin M, Redl H, Bahrami S, Abraham C, Graeter T, artery versus central venous catheter to guide treatment of acute Haverich A, Schlag G, Hans-Beorg B. Systemic inflammatory lung injury. NEJM. 2006;354(21):2213–23. response syndrome after cardiac operations. Ann Thorac Surg. Nichols W, O’Rourke M. Vascular impedance. In: McDonald’s blood 1996;61:1714–20. flow in arteries. Theoretical, experimental and clinical principles. Dark PM, Singer M. The validity of trans-esophageal Doppler 5th ed. London: Hodder Arnold, Oxford University Press; 2005. ultrasonography as a measure of cardiac output in critically ill pp. 233–67. adults. Intensive Care Med. 2004;30:2060–6. Ouzounian JG, Masaki DI, Abboud TK, Greenspoon JS. Systemic DiCorte CJ, Latham P, Greilich PE, Cooley MV, Grayburn PA, Jessen vascular resistance index determined by thoracic electrical ME. Esophageal Doppler monitor determinations of cardiac bioimpedance predicts the risk for maternal hypotension during output and during cardiac operations. Ann Thorac Surg. regional anesthesia for cesarean delivery. Am J Obstet Gynecol. 2000;69(6):1782–6. 1996;174(3):1019–25. Dodd TEL. Nasal insertion of the oesophageal Doppler probe. Richards MJ, Edwards JR, Culver DH, Gaynes RP. The National Anaesthesia. 2002;57(4):412. Nosocomial Infections Surveillance System. Nosocomial infec- Domino KB, Bowdle TA, Posner KL, Spitellie PH, Lee LA, Cheney tions in combined medical and surgical intensive care units in the FW. Injuries and liability related to central vascular catheters: a United States. Infect Control Hosp Epidemiol. 2000;21:510–5. closed claims analysis. Anesthesiology. 2004;100(6):1411–8. Sandham JD, Hull RD, et al. A randomized, controlled trial of the use Drazner MH, Rame JE, Stevenson LW, Dries DL. Prognostic of pulmonary-artery catheters in high-risk surgical patients. importance of elevated and a third heart NEJM. 2003;348(1):5–14. sound in patients with . NEJM. 2001;345(8):574–81. Segers P, Rietzschel ER, De Buyzere ML, Vermeersch SJ, De Hawkins DL. Using U-statistics to derive the asymptotic distribution Bacquer D, Van Bortel LM, De Backer G, Gillebert TC, of Fisher’s Z statistic. Am Stat. 1989;43:235–7. Verdonck PR. Noninvasive (Input) impedance, pulse wave Ho AMH, Dion PW, Karmaker MK, Jenkins CR. Accuracy of central velocity, and wave reflection in healthy middle-aged men and venous pressure monitoring during simultaneous continuous women. Hypertension. 2007;49(6):1248–55. infusion through the same catheter. Anaesthesia. 2005;60(10): Senni M, Tribouilloy CM, Rodeheffer RJ, Jacobsen SJ, Evans JM, 1027–30. Bailey KR, Redfield MM. Congestive heart failure in the Imhoff M, Lehner JH, Dietrich Lo¨hlein D. Noninvasive whole-body community: a study of all incident cases in Olmsted county, electrical bioimpedance cardiac output and invasive thermodi- Minnesota, in 1991. Circulation. 1998;98:2282–9. lution cardiac output in high-risk surgical patients. Crit Care Simon R, Desebbe O, Henaine R, Bastien O, Lehot JJ, Cannesson M. Med. 2000;28:2812–8. Comparison of ICG thoracic bioimpedance cardiac output Kohsaka S, Menon V, Lowe AM, Lange M, Dzavik V, Sleeper LA, monitoring system in patients undergoing cardiac surgery with Hochman JS. Systemic inflammatory response syndrome after pulmonary artery cardiac output measurements. [French (English acute myocardial infarction complicated by . abstract)] Annales Franc¸aises d’Anesthe´sie et de Re´animation. Arch Intern Med. 2005;165:1643–50. 2009;28(6):537–41. Kreyszig E. Advanced engineering mathematics. 8th ed. New York: Singer M, Clarke J, Bennett ED. Continuous hemodynamic monitor- Wiley; 1999. ing by esophageal Doppler. Crit Care Med. 1989;17:447–52. Kutner MH, Nachtsheim CJ, Neter J. Applied linear regression Soliman DE, Maslow AD, Bokesch PM, Strafford M, Karlin L, Marx models. 4th ed. Chicago: McGraw-Hill; 2004. GR. Transoesophageal during scoliosis repair: Li JK-J. The arterial circulation. Physical principles and clinical comparison with CVP monitoring. Can J Anaesth. 1998;45(10): applications. Totowa: Humana Press; 2000. 925–32. Li JK-J. Dynamics of the vascular system. Singapore: World Stewart RD, Psyhojos T, Lahey SJ, Levitsky S, Campos CT. Central Scientific; 2004. venous catheter use in low-risk coronary artery bypass grafting. Madan AK, UyBarreta VV, Aliabadi-Wahle S, Jesperson R, Hartz Ann Thorac Surg. 1998;66(4):1306–11. RS, Flint LM, Steinberg SM. Esophageal Doppler ultrasound Valtier B, Cholley BP, Belot JP, de la Coussaye JE, Mateo J, Payen monitor versus pulmonary artery catheter in the hemodynamic D. Noninvasive monitoring of cardiac output in critically ill management of critically ill surgical patients. J Trauma Injury patients using transesophageal Doppler. Am J Respir Crit Care Infect Critic Care. 1999;46(4):607–11. Med. 1998;158:77–83. Magder S. How to use central venous pressure measurements. Curr Wang GY, Ma B, Tang HT, Zhu SH, Lu J, Wei W, Ge SD, Xia ZF. Opin Crit Care. 2005;11(3):264–70. Esophageal echo-Doppler monitoring in burn shock resuscita- Manecke GR. EdwardsÒ sensor and VigileoÒ monitor: easy, accurate, tion: are hemodynamic variables the critical standard guiding reliable cardiac output assessment using the arterial pulse wave. fluid therapy? J Trauma. 2008;65:1396–401. Expert Rev Med Devices. 2005;2(5):523–7. Warzawski RC, Deye AN, et al. Early use of the pulmonary artery Melo J, Peters JI. Low systemic vascular resistance: differential catheter and outcomes in patients with shock and acute diagnosis and outcome. Crit Care. 1999;3:71–7. respiratory distress syndrome: a randomized controlled trail. National Heart, Lung, and Blood Institute (NHLBI). Acute Respiratory JAMA. 2003;290(20):2713–20. Distress Syndrome (ARDS) Clinical Trials Network. Pulmonary

123