SEARCH

SEARCH BY CITATION

Keywords:

  • Diagnostic techniques: cardiac output, oesophageal Doppler ultrasonography;
  • Monitoring: cardiac ouput;
  • Surgery: bowel

Abstract

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Appendix 1 Goldman Cardiac Risk score

Summary Oesophageal Doppler monitoring allows non-invasive estimation of stroke volume and cardiac output. We studied the impact of Doppler guided fluid optimisation on haemodynamic parameters, peri-operative morbidity and hospital stay in patients undergoing major bowel surgery. Fifty-seven patients were randomly assigned to Doppler (D) or control (C) groups. All patients received intra-operative fluid therapy at the discretion of the non-investigating anaesthetist. In addition, Group D were given fluid challenges (3 ml.kg−1) guided by oesophageal Doppler. Group D received significantly more intra-operative colloid than Group C (mean 28 (SD 16) vs. 19.4 (SD 14.7) ml.kg−1, p = 0.02). Cardiac output increased significantly for Group D whilst that of controls remained unchanged. The mean difference between the groups in final cardiac output was 0.87 l.min−1 (95% confidence interval 0.31–1.43 l.min−1, p = 0.003). Five control patients required postoperative critical care admission. Fluid titration using oesophageal Doppler during bowel surgery can improve haemodynamic parameters and may reduce critical care admissions postoperatively.

Surgical patients undergoing major bowel resection are athigh risk of peri-operative complications and death. Centres in the UK report mortality rates in this groupof between 4 and 9%[1, 2]. Many of these patients are elderly and suffer comorbid medical conditions as well as the local and systemic effects of colorectal disease. Studies of similar patient groups having major surgery have used fluid, inotrope and oxygen therapy to optimise cardiac output and oxygen delivery; these studies demonstrated reductions in mortality and lengthof hospital stay [3–5]. The method of cardiac output measurement in these reports, pulmonary artery catheterisation, is not commonly used for bowel surgery. Pulmonary artery catheter insertion can be time consuming and has been implicated with complications and excess mortality [6]. Furthermore some trials involving pulmonary artery catheter optimisation of oxygen delivery required pre-operative admission to intensive care which is currently not practical for colorectal resection.

The minimally invasive oesophageal Doppler monitor permits real time assessment of cardiac output [7]. Haemodynamic parameters estimated by this device can be used to guide fluid therapy during surgery [8]. A study of patients undergoing repair of femoral neckfracture demonstrated a reduction in length of hospital stay when oesophageal Doppler was used to guide fluid therapy [9].

The aim of this study was to examine the effect of oesophageal Doppler guided fluid administration during colorectal resection on haemodynamic performance, hospital stay and postoperative complications.

Patients and methods

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Appendix 1 Goldman Cardiac Risk score

The Local Research Ethics Committees approved the study. All patients gave written informed consent. Patients undergoing major bowel resections were recruited into a prospective randomised controlled trial. Those who were having emergency, intrathoracic or oesophageal surgery were not studied. Patients with a known sensitivity to starch-based colloid or any history of oesophageal disease were also not studied.

Pre-operative assessment included measurement of height, weight, haemoglobin concentration and estimations of the ASA grade and Goldman Cardiac Risk Index [10, 11] (Appendix 1). Pre-operative therapy included the routine use of bowel purgatives. All patients received a standardised general anaesthetic. This comprised intravenous induction, muscle relaxation and tracheal intubation. Anaesthesia was maintained using isoflurane in a nitrous oxide and oxygen mixture. Peri-operative analgesia was provided by intravenous fentanyl. The investigators noted when postoperative epidural analgesia was used.

Routine peroperative cardiovascular monitoring included continuous electrocardiogram, pulse oximetry and intermittent non-invasive blood pressure every 3 mins. Central venous pressure monitoring was used at the discretion of the anaesthetist. Following induction of anaesthesia, a 12 French gauge oesophageal Doppler probe (TECO 2, Medicina, Oak House, Cookham, Berkshire, UK) was passed orally into the mid-oesophagus of all patients. A continuous wave Doppler transducer (4 MHz) was mounted at 45° to the tip of the probe. The reflected Doppler-shift signal was transformed using spectral analysis into a blood flow velocity profile. Orientation along the longitudinal axis allowed the optimal measurement of the velocity of blood flow in the descending aorta. With a nomogram incorporating the patient's age, height and weight, the characteristic waveform representing velocity against time was used to estimate the left ventricular stroke volume [7]. The systolic flow time was corrected for heart rate using a modification of Bazett's equation [8]. Corrected flow time, stroke volume, cardiac output and index was recorded every 15 mins but the anaesthetist concerned was not aware of the results.

Prior to induction of anaesthesia, patients were individually randomised into Doppler (D) or control (C) groups. All patients received fluid therapy deemed appropriate for the patient by the anaesthetist. There was no restriction on quantity nor type of fluid used. Group D received additional fluid boluses of 3 ml.kg−1 Hydroxyethyl starch solution according to an algorithm based upon oesophageal Doppler recordings. The algorithm was designed to optimise the stroke volume and maintain the corrected flow time > 0.35 s as described by Sinclair et al. [9] (Fig. 1). If stroke volume fell inresponse to a fluid challenge and flow time and aortic velocity waveform indicated that the patient was volume overloaded, the trial was discontinued and the anaesthetist was made aware of the measurements. The oesophageal Doppler parameters at skin preparation and during skin closure were recorded for analysis as ‘initial’ and ‘final’, respectively.

image

Figure 1. Fluid algorithm guided by oesophageal Doppler measurements. (SV: stroke volume, FTc: corrected flow time).

Download figure to PowerPoint

The location and conduct of postoperative care was the joint decision of the anaesthetist and surgeon in charge of the case. Medical and nursing ward staff were unaware of randomisation. The investigators followed each patient, noting time spent on the ward until discharge, the time taken to tolerate oral diet and any time spent in high dependency or intensive care. Outcomes were analysed on an intention-to-treat basis for 28 days postoperative mortality for all causes. Any complication that developed postoperatively was recorded, as was the cause of any death.

A sample size calculation revealed that 26 patients would be required in each group, to detect an increase in final cardiac output of 1 l.min−1. Using the standard deviation 1.3 l.min−1 from the most similar recent study, this gave the study a power of 80%, with a p-value of 0.05[9]. Normally distributed data were analysed using Students' t-test for unpaired samples and other continuous data were analysed using the Mann–Whitney U-test. Haemodynamic parameters were analysed by linear regression to calculate confidence intervals for the group effect, adjusting for initial value of the relevant measure. For the comparison of categorical data, the Fisher's exact test was used. A p-value of < 0.05 was considered statistically significant. Normally distributed data are expressed as means (standard deviations (SD)) whilst other data are summarised as median (range).

Results

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Appendix 1 Goldman Cardiac Risk score

Fifty-seven patients agreed to participate in the trial and 29 were randomly assigned to the Doppler group. Both groups were similar in terms of age, gender, weight, height and pre-operative haemoglobin concentration (Table 1). ASA and Goldman cardiac risk indices were similar. The pre-operative length of hospital stay, duration and type of surgery was comparable between the groups. One patient in each group was withdrawn due to problems with the probe. Group D received 28.0 (SD 16) ml.kg−1 of colloid intra-operatively, significantly more than 19.4 (SD 14.7) ml.kg−1 given to controls (p = 0.02). The total fluid therapy in Group D 64.6 (SD 36.4) ml.kg−1 was more than Group C 55.2 (SD 24) ml.kg−1 although this did not reach statistical significance. Stroke volume, corrected flow time and cardiac output increased significantly in Group D whilst remaining stable in controls (Table 2). The mean difference in final cardiac output was 0.87 l.min−1 (95% confidence interval 0.31–1.43, p = 0.003). The cardiac output of nine control patients fell by the end of surgery; this did not occur in the Doppler group.

Table 1.  Patient characteristics. Values are means (standard deviations) or medians (ranges).
 ControlDoppler
n2829
Age; year67.5 (10.1)66.5 (12.5)
Height; m1.67 (0.08)1.68 (0.11)
Weight; kg68.3 (13)69.9 (19)
Pre-operative haemoglobin; g.dl−112.8 (1.5)12.8 (1.8)
ASA GradeII (I–III)I (I–III)
Goldman Score3 (3–29)3 (3–11)
Operation Time: h, mins2, 22 (60)2, 12 (34)
Table 2.  Haemodynamic variables measured by oesophageal Doppler at the beginning and end of surgery. Values are means (standard deviations).
 nControl 28Doppler 29
  • *p

     < 0.05 compared with final control.

Flow time (corrected); sInitial 0.352 (0.038) 0.341 (0.025)
 Final 0.354 (0.045) 0.383 (0.03)*
Stroke volume; mlInitial71.5 (19.9)70.25 (16.6)
 Final67.8 (18.7)82.9 (24.4)*
Cardiac output; l.min−1Initial 4.6 (1.3) 4.5 (1.4)
 Final 5.0 (1.4) 6.1 (1.9)*

One control patient died in the postoperative period. This man had significant cardiac comorbidity, with a pre-operative Goldman cardiac score of 29. He was admitted to the high dependency unit and died as a result of surgical complications and cardiac failure on the third postoperative day. Five patients required high dependency care. Three patients were admitted to the surgical high dependency unit for postoperative care for 2, 3 and 4 days, respectively. Two patients were admitted to the coronary care unit after developing severe tachyarrhythmias, which compromised organ perfusion. All five patients requiring critical care were in the control group (p = 0.02). The cardiac output in four of these patients had decreased by the end of surgery. Nine control patients and five in the Doppler group had at least one postoperative complication including chest infection, delerium, pulmonary embolus, surgical problems requiring re-operation, cardiac failure and arrhythmias. The length of hospital stay and time to tolerate oral diet was similar between the groups (Table 3).

Table 3.  Post-operative stay. Values are medians (ranges).
 ControlDoppler
Tolerating oral diet; days 6 (5–8) 7 (4–37)
Length of hospital stay; days11 (7–30)12 (7–103)
Critical care days (n = 5) 3 (1–5) 0

No Group D patient developed signs of fluid overload or cardiac failure. There were no complications associated with probe placement nor subsequent haemodynamic management.

Discussion

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Appendix 1 Goldman Cardiac Risk score

This study describes a practical method of fluid management using oesophageal Doppler during surgery. We have demonstrated improvements in cardiovascular performance and reduced complications leading to critical care admission using Doppler guided fluid challenges, although no change in hospital length of stay was detected.

Despite evidence it may improve outcome, only 3% of surgical patients who die have cardiac output measured peroperatively [12]. Recent studies have questioned the safety of pulmonary artery catheterisation and highlighted the time taken to insert the device [6, 13]. Oesophageal Doppler is comparable with other methods for estimating cardiac output and stroke volume [7]. The corrected flow time has been used as an indicator of intravascular volume status and left ventricular preload [14]; however, certain assumptions have to be made, particularly regarding uniformity of blood flow and aortic wall diameter. The fluid challenge algorithm in this study was designed to respond to trends, which hopefully would reduce systematic errors caused by discrepancies between the actual and calculated values. Monitors that measure aorticcross-sectional diameter using 10 MHz M-Echo mode may improve accuracy [15]. Minimally invasive oesophageal Doppler is simple to use with clinical proficiency being gained relatively quickly.

This study highlights the risk of death and serious complication faced by patients undergoing colorectal resection. The morbidity and mortality we report is consistent with UK studies of colorectal surgery [1, 2]. Peri-operative fluid management of the high risk surgical patient is a challenging area of clinical practice. This is undoubtedly the case during colorectal surgery where patients can experience large fluid shifts as a result of bowel purgation, blood, third space and evaporative losses. Previous studies drove oxygen delivery towards a predefined goal of 600 ml.min−1.m−2 whereas our algorithm responded to changes in intravascular volume status. The fluid requirements in both groups were similar yet we were able to give additional colloid (15 (SD 8.9) ml.kg−1) to the intervention group. This would imply that patients in the control group were relatively hypovolaemic during surgery. Hypovolaemia increases the risk of postoperative complications [16]. The gut is thought to play a critical role in initiating this process [17]. Previous studies have emphasised the importance of optimal fluid provision to improve oxygen delivery and reduce complications [3–5, 9]. Oesophageal Doppler provides an instantaneous representation of beat-to-beat haemodynamic function and so allows rapid correction of hypovolaemia and oxygen debt. The timing of volume loading can be crucial in reducing hypoperfusion and subsequent tissue reperfusion injury. Delaying fluid replacement until the postoperative phase may compromise the critically perfused gut mucosa during surgery.

We observed similar increases in cardiac output to those reported in other studies [9]. However, we were unable to demonstrate an impact on length of hospital stay. It is likely that this study was underpowered to demonstrate any effect on this endpoint. Also most patients stay in hospital until they can eat, drink and pass stool, having followed a rigid programme of oral intake. It is possible that the benefits of improved peri-operative intravascular volume were masked by the structure of postoperative care. Further trials with larger sample sizes are required to demonstrate any impact on hospital stay in this population.

The increased postoperative complications in controls did not reach statistical significance but did cause five patients to need critical care. It was not policy at either institution for epidural analgesia to be only managed in high dependency. Five patients in the trial with postoperative epidural analgesia were managed on the general surgical ward, three from Group D. Those admitted to critical care had increased length of hospital stay (median 14 days) excluding one patient who died on the third postoperative day.

In conclusion, minimally invasive cardiac output monitoring with oesophageal Doppler can be used to guide fluid administration during bowel surgery. This was associated with increased final cardiac output and reduced critical care admission in our study. Further larger investigations are required to demonstrate an effect on mortality and hospital stay.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Appendix 1 Goldman Cardiac Risk score

The authors thank the general surgeons and anaesthetists at North Manchester General Hospital and Blackburn Royal Infirmary for their support and invaluable help with this study. A grant from the R. L. Gardner Trust was used for equipment hire.

References

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Appendix 1 Goldman Cardiac Risk score
  • 1
    Sagar PM, Hartley MN, Mancey-Jones B, Sedman PC, May J, Macfie J. Comparative audit of colorectal resection with the POSSUM scoring system. British Journal of Surgery1994; 81: 14924.
  • 2
    Singh S, Morgan MB, Broughton M, Caffarey S, Topham C, Marks CG. A 10-year prospective audit of outcome of surgical treatment for colorectal carcinoma. British Journal of Surgery1995; 82: 148690.
  • 3
    Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee T. Prospective trial of supranormal values of survivors as therapeutic goals in high risk surgical patients. Chest1987; 94: 117686.
  • 4
    Boyd O, Grounds RM, Bennett ED. A randomised clinical trial of the effect of deliberate peri-operative increase of oxygen delivery on mortality in high-risk surgical patients. Journal of the American Medical Association1993; 270: 2699707.
  • 5
    Wilson J, Woods I, Fawcett J, et al. Reducing the risk of major elective surgery: randomised controlled trial of pre-operative optimisation of oxygen delivery. British Medical Journal1999; 318: 1099103.
  • 6
    Connors AF, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. Journal of the American Medical Association1996; 276: 88997.
  • 7
    Valtier B, Cholley BP, Belot JP, Coussaye JE, Mateo J, Payen DM. Non-invasive monitoring of cardiac output in critically ill patients using transoesophageal Doppler. American Journal of Respiratory and Critical Care Medicine1998; 158: 7783.
  • 8
    Singer M, Bennett ED. Non-invasive optimisation of left ventricular filling by esophageal Doppler. Critical Care Medicine1991; 9: 11327.
  • 9
    Sinclair S, James S, Singer M. Intra-operative intravascular volume optimisation and length of hospital stay after repair of proximal femoral fracture: randomised controlled trial. British Medical Journal1997; 315: 90912.
  • 10
    American Society of Anesthesiologists. New classification of physical status. Anesthesiology1963; 24: 111.
  • 11
    Goldman L, Caldera DL, Nussbaum SR, et al. Multifactorial index of cardiac risk in non-cardiac patients. New England Journal of Medicine1977; 297: 84550.
  • 12
    National Confidential Enquiry into Peri-operative Deaths. Then and Now, Report of the 1998–99 National Confidential Enquiry into Peri-operative Deaths. London: NCEPOD, 2000.
  • 13
    Lefrant JY, Muller L, Bruelle P, et al. Insertion time of the pulmonary artery catheter in critically ill patients. Critical Care Medicine2000; 28: 3559.
  • 14
    DiCorte CJ, Latham P, Greilich PE, Cooley MV, Grayburn PA, Jessen ME. Esophageal monitor determinations of cardiac output and preload during cardiac operations. Annals of Thoracic Surgery2000; 69: 17826.
  • 15
    Cariou A, Monchi M, Joly LM, et al. Non-invasive cardiac output monitoring by aortic blood flow determination: Evaluation of the Sometec Dynemo-3000 system. Critical Care Medicine1998; 26: 206672.
  • 16
    Shoemaker WC, Appel PL, Kram HB. Role of oxygen debt in the development of organ failure, sepsis and death in high risk surgical patients. Chest1992; 102: 20815.
  • 17
    Mythen MG, Webb AR. The role of gut mucosal hypoperfusion in the pathogenesis of post-operative organ dysfunction. Intensive Care Medicin.1994; 20: 2039.