Cerebral hypoperfusion in immediate postoperative period following coronary artery bypass grafting, heart valve, and abdominal aortic surgery

S. M. Millar1, R. P. Alston2,3, P. J. D. Andrews2,4 and M. J. Souter5

1Department of Anaesthesia, The Alfred Hospital, Prahran Melbourne, Victoria, Australia. 2Anaesthesia, Critical Care and Pain Medicine Section, Department of Clinical and Surgical Sciences, University of Edinburgh, Edinburgh, UK. 3Department of Anaesthetics, Royal Infirmary of Edinburgh, Lauriston Place, Edinburgh EH3 9YW, UK. 4Department of Anaesthesia, Western General Hospital, Edinburgh, UK. 5Department of Neuroanaesthesia, Southern General Hospital, Glasgow, UK*Corresponding author

Accepted for publication: March 20, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Perioperative levels of jugular bulb oxyhaemoglobin saturation (SjO2) and lactate concentration (Lj), and postoperative duration of SjO2<50% were compared between patients undergoing coronary artery bypass grafting (CABG) (n=86), heart valve (n=14) and abdominal aortic (n=16) surgery. Radial artery and jugular bulb blood samples were aspirated after induction of anaesthesia, during re-warming on cardiopulmonary bypass (CPB) (36°C), on arrival in the intensive care unit (ICU) and, subsequently, at 1, 2 and 6 h after ICU admission. Most patients having heart surgery were hypocapnic at 36°C on CPB. Following CABG and heart valve surgery, many patients were hypocapnic whereas after abdominal aortic surgery, most were hypercapnic. During CPB and postoperatively, SjO2 and Lj were significantly correlated to PaCO2 and the arterial concentration of lactate (La) respectively (P<0.05). After correction for arterial carbon dioxide tension (PaCO2) and La, there were no significant changes in SjO2 or Lj on CPB. Postoperatively, having corrected for PaCO2, there were significant effects on SjO2 over all groups as a result of time from surgery (P<0.001) and its interaction with operation type (P<0.001). Following correction for La, there were no postoperative effects on Lj. No significant differences (P=0.2) in duration of SjO2<50% existed between patients undergoing CABG (1054 (82) min), abdominal aortic (893 (113) min) and heart valve (1073 (91) min) surgery. The lack of significant reciprocal effects on Lj combined with the frequency of hypocapnia and strong influence of PaCO2on SjO2, suggest that SjO2<50% during CPB and after cardiac surgery represents hypoperfusion as a consequence of hypocapnia rather than cerebral ischaemia.

Br J Anaesth 2001; 87: 229–36

Keywords: surgery, cardiopulmonary bypass; intensive care; brain, cerebral oxygenation; veins, jugular, measurement techniques oximetry; carbon dioxide; blood, lactic acid


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Despite much research, a common mechanism of injury for the brain damage that is associated with coronary artery bypass grafting (CABG) surgery has yet to be determined.1 This failure may in part be because of the focus of research on cardiopulmonary bypass (CPB) as the epoch when damage occurs. We have found, using fibreoptic measurements of jugular bulb oxyhaemoglobin saturation (SjO2), that many patients appear to experience episodes, often multiple and sometimes prolonged, of cerebral hypoperfusion (SjO2<50%) in the early postoperative period following CABG surgery.2 Although cerebral hypoperfusion has been reported intraoperatively,35 we have found the average duration of postoperative SjO2<50% to be in excess of six times longer.2 Therefore, if some brain damage results from occurrences during CPB,3 then the far more prolonged postoperative cerebral hypoperfusion may consequently cause more damage. However, it is not known whether postoperative cerebral hypoperfusion is a phenomenon that only follows cardiac surgery as if so, this would be a strong indication that it causes cerebral damage.

In other recent work,6 we have found levels of the arteriovenous difference in lactate concentration (AVDL) during the re-warming phase of CPB that, in patients who have suffered head injuries, would equate with profound cerebral ischaemia or infarction.7 8 In addition, we have found evidence of cerebral hypoperfusion using a combination estimate, the lactate-oxygen index (LOI).4 6 9 The cerebral arteriovenous difference in oxygen content (AVDO2), AVDL and LOI are related to factors that influence cerebral blood flow and oxygen delivery including arterial carbon dioxide tension (PaCO2), concentration of arterial lactate, arterial oxyhaemoglobin saturation (SaO2), mean arterial pressure (MAP) and arterial haemoglobin concentration (Hba).912 Postoperative cerebral hypoperfusion may be caused by deranged levels of these variables. However, determination of AVDO2 and LOI are error prone.13 Moreover, AVDL, AVDO2 and LOI are both mathematically coupled to, and not statistically independent from, factors that may influence them including Hba, SaO2 and the concentration of arterial lactate.1315 By contrast to these derived measures, SjO2 and concentrations of jugular bulb lactate (Lj) are robust and statically independent estimates of the adequacy of cerebral perfusion.13

Consequently, the primary aim of this study was to establish whether postoperative cerebral hypoperfusion, as estimated by levels of SjO2, Lj and duration of SjO2<50%, is subsequent only to cardiac surgery by comparing patients undergoing CABG, heart valve, and abdominal aortic surgery. The secondary aim was to investigate the effects of the concentration of arterial lactate, PaCO2, MAP, Hba, and SaO2 upon SjO2 and Lj.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The Local Ethics Committee approved the study. Patients were excluded if they had a history of previous cerebrovascular accident, transient ischaemic attack or diabetes mellitus.

Anaesthesia and postoperative management
For all groups, the anaesthetic techniques were according to individual consultant practice. Approximately 1 h before surgery, all patients were premedicated with temazepam 20–40 mg or lorazepam 1–2 mg p.o. either alone or in combination with morphine 10–15 mg with or without atropine 0.3–0.6 mg i.m. Anaesthesia was induced with thiopental 1–3 mg kg–1, etomidate 0.1–0.2 mg kg–1 or propofol 1–2 mg kg–1 in addition to fentanyl 4–10 µg kg–1 or remifentanil 1–2 µg kg–1 i.v. Neuromuscular blockade was obtained with pancuronium 0.1 mg kg–1 or rocuronium 0.9 mg kg–1. Anaesthesia was maintained with 1–2% isoflurane, propofol target controlled infusion (TCI) 2–3 µg ml–1 and remifentanil 0.02–0.1 µg kg–1 min–1, or morphine 0.25 mg kg–1 h–1 and midazolam 0.04 mg kg–1 h–1 i.v.

In patients undergoing abdominal aortic surgery, a low thoracic epidural catheter was placed. Epidural 0.125% bupivacaine with either fentanyl 2 µg ml–1 or ketamine 166 µg ml–1 was administered during and after surgery. Anaesthesia was induced with propofol 1–2 mg kg–1 and neuromuscular blockade was obtained with atracurium 0.5 mg kg–1 or vecuronium 0.1 mg kg–1. Anaesthesia was maintained with either 0.5–2% isoflurane or propofol TCI 2–3 µg ml–1 along with nitrous oxide and oxygen.

Following surgery, patients having cardiac surgery were transferred to the intensive care unit (ICU) where mechanical ventilation was weaned and the trachea extubated according to a previously described protocol.2 9 Tracheal extubation in those patients having abdominal aortic surgery was undertaken before transfer to a general surgical high dependency unit (HDU).

Cardiopulmonary bypass
Cardiopulmonary bypass was used for all patients undergoing CABG or heart valve surgery. The CPB circuit and its management have been described in detail.6 9 16 In particular, the circuit was primed with 2 litres of lactated Ringer’s solution and 50 mmol sodium bicarbonate was used. Thermal management on CPB was according to surgical practice: the lowest nasopharyngeal temperature ranged from 28.7 to 35.2°C in the patients having heart valve surgery and from 27.1 to 36.5°C in patients undergoing CABG surgery. As was our practice when the study was undertaken, re-warming was carried out with the heat exchanger set at no more than 10°C above the patient’s nasopharyngeal temperature, and not above 42°C.

Fibreoptic SjO2 and haemodynamics
All patients had radial artery cannulation and central venous line placement for pressure monitoring. A Baxter Edslab 4F (Baxter Healthcare Corporation, Edwards Critical-Care Division, Irvine, CA, USA) double lumen oximetric catheter was placed preferentially into the right internal jugular bulb using a previously described technique.2 The position of the catheter tip in the jugular bulb was confirmed on a plain lateral neck x-ray carried out after surgery. The first co-oximeter measurement was used to calibrate the fibreoptic catheter in vivo according to the manufacturer’s instructions. Following surgery, calibration was repeated using the first post-operative blood sample result. If the continuous SjO2 reading differed by more than 5% from the co-oximeter reading, or if there was a change in haemoglobin concentration of greater than 1.8 g dl–1, the catheter was re-calibrated.

Fibreoptic SjO2 measurements were recorded every minute from the start of surgery until its completion. Postoperative recording was continued for 18 h after arrival in the ICU for the cardiac surgery patients or from arrival in the recovery area for the abdominal aortic surgery patients. Data were logged on computer and then converted to a format compatible with the Edinburgh Data Browser (Medical Research Council Head Injuries Clinical Research Initiative, Western General Hospital, Edinburgh, UK). A Kolormon 7250 Monitoring System (Kontron Instruments Ltd., Watford, Herts, UK) was used in the CABG and heart valve surgery patients, both intra- and postoperatively. In patients undergoing abdominal aortic surgery, the Critical Care Explorer (Baxter Healthcare Ltd, Wallingford Road, Compton, Newbury, Berkshire, UK) was used to measure SjO2 throughout but other haemodynamic parameters were monitored using the Component Monitoring System (Hewlett Packard Ltd, Uxbridge, Middlesex, UK) intraoperatively and the Cardiocap2 (Datex Ohmeda Ltd, Hatfield, Herts, UK) postoperatively. Cerebral hypoperfusion was defined, using methodology developed for assessment of neurological outcome in patients with head injuries, as a SjO2<50% for at least 5 min.17

Blood samples
Blood samples were aspirated into heparinized syringes from the radial artery and jugular bulb following induction of general anaesthesia; during re-warming on CPB when nasopharyngeal temperature reached 36°C (CABG and heart valve surgery groups only); on arrival in ICU or recovery area; and, subsequently, at one, two and six hours. Sampling times were based upon a preliminary ranging study that identified the time points as having the highest frequency of SjO2<50% in the first 18 h after CABG surgery.9 Each blood sample was measured in the same three analyzers: (i) bench oximeter (IL 482 CO-oximeter, Instrumentation Laboratory, Lexington, MA, USA) which measured the concentration and saturation of oxyhaemoglobin; (ii) blood gas analyzer (ABL 4, Radiometer Ltd, Copenhagen, Denmark) which measured the partial pressure of oxygen and carbon dioxide, and (iii) lactate analyser (YSI 2300 Stat G/L, YSI Inc., Yellow Springs, OH, USA) which measured the lactate concentration.

Sample size
Calculation of sample size was based upon a mean duration of SjO2<50% to be 14% of the total monitored time in the postoperative period following CABG surgery.2 It was assumed that postoperative SjO2<50% does not occur following major abdominal vascular surgery but a 1% duration was used to give a margin for error. Assuming a one sided percentage point of the normal distribution and f=3/4, 46 patients in the abdominal aortic surgery group and 92 in the CABG group would give an 80% power of detecting a significant difference with alpha set at 0.05 in the duration of SjO2<50%.

Statistical analysis
Statistical analysis was performed using SPSS 6.1 (SPSS Inc., Chicago, IL, USA) running on a Macintosh G3 personal computer with OS 8.6. Non-normally distributed data (duration) were transformed using the natural logarithm. A probability of 0.05 was used as the level of significance in all analysis.

As we have previously described, method comparison analysis was used to compare bench and fibreoptic measurements of SjO2.16 Duration of SjO2<50% was analysed using general factorial analysis of variance with the operation grouping entered as a factor on three levels.15 As there were between group differences in age and timing of tracheal extubation, these variances were controlled by entering them as covariates. To examine the influence of the duration of CPB, this was also entered into the model with the abdominal aortic surgery group entered as zero.

Repeated measures multivariate analysis of variance (MANOVA) was used to analyse the discrete measurements of SjO2 and Lj and these were entered as within-subject variables.15 Operation grouping was entered as a factor and, to examine their influence on SjO2and Lj, PaCO2, MAP, SaO2 and Hba were entered as varying covariates. To control for any differences between the groups, age and duration of CPB were entered as constant covariates. Using a full factorial model, a sequential approach was taken to decomposing the sum of squares. First, this adjusts covariates for all factors and interactions and then each factor is adjusted for all covariates, factors and preceding interactions. Post-hoc analysis was performed using paired and unpaired Student’s t-test as appropriate.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
From February 1997 to January 1998, 116 patients, including those from two preliminary reports,9 16 gave written informed consent and of these, six failed to enter the study because of re-scheduling of surgery. Less than the planned population were recruited because of technical difficulties adapting the data logging system and lower than expected throughput of patients undergoing vascular surgery. Eight patients were excluded. Causes for exclusion were co-oximeter failure; inability to aspirate jugular line intraoperatively; postoperative cardiovascular instability precluding blood sampling and, inability to aspirate from or premature removal of the arterial line postoperatively. Patients undergoing abdominal aortic surgery tended to be older and tracheal extubation occurred earlier than in the other two groups (Table 1). One patient had a cerebrovascular accident and two died from cardiac failure in the CABG surgery group.


View this table:
[in this window]
[in a new window]
 
Table 1 Physical characteristics of patients and operative details. CABG= coronary artery bypass grafting surgery; Aortic=abdominal aortic surgery; Valve=heart valve surgery; Excluded=patients excluded; Number of CABGs=number of coronary artery bypass grafts performed; Operation= duration of surgery; Cross-clamp=duration of aortic cross-clamping; CPB=duration of cardiopulmonary bypass; CPB temperature=minimum nasopharyngeal temperature on cardiopulmonary bypass; Time intubated= duration of tracheal intubation; ICU/HDU time=duration in intensive care unit (coronary artery bypass grafting and heart valve surgery) or high dependency unit (abdominal aortic surgery) stay; Hospital stay=duration of stay in hospital. Presented as mean (standard deviation) except * median (inter-quartile range)
 
Method comparison analysis
Whilst the limits of agreements between bench and fibreoptic measurements of SjO2 were reasonable following surgery, agreements were wide during CPB. For this reason, fibreoptic oximetry measurements were excluded from the analysis of SjO2 during CPB (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2 Method comparison analysis of fibreoptic and bench measurement of jugular bulb oxyhaemoglobin saturation. CABG=coronary artery bypass grafting, CPB=cardiopulmonary bypass, bias=mean difference bench – fibreoptic oximetry, 95% CI of bias=95% confidence intervals of bias (mean difference ±& t*standard error of the mean, values not including zero indicate a significant difference [P<0.05]), limits of agreement-mean difference ± 1.96*standard deviation
 
Discrete variables
Discrete variables are summarized in Table 3. Throughout the study, patients in all surgical groups experienced a wide range of PaCO2. During the re-warming phase of CPB, the majority of patients undergoing CABG and heart valve surgery were hypocapnic. On arrival on the ICU, and at 1 h postoperatively, many patients having CABG and heart valve surgery were also hypocapnic whereas the majority of patients having abdominal aortic surgery were hypercapnic.


View this table:
[in this window]
[in a new window]
 
Table 3 Arterial and jugular venous oxyhaemoglobin concentrations, oxygen saturation and tension, concentrations of lactate, arterial carbon dioxide tension and mean arterial pressure after induction of anaesthesia, during cardiopulmonary bypass and following coronary artery bypass grafting, abdominal aortic and heart valve surgery. Except where indicated, results are presented as mean (standard deviation) [range]. Postinduction=baseline; CPB 36°C=at 36°C during re-warming on cardiopulmonary bypass; Postop 0 h=arrival in the intensive care/high dependency unit; Postop 1 h, Postop 2 h, Postop 6 h=1, 2 and 6 h after arrival in intensive care unit (coronary artery bypass grafting and heart valve surgery) or recovery unit (abdominal aortic surgery); CABG=coronary artery bypass grafting surgery; Aortic=abdominal aortic surgery; Valve=heart valve surgery; Hba and Hbv=arterial and venous concentrations of haemoglobin; PaO2 and PvO2=arterial and venous tensions of oxygen; SaO2 and SjO2=arterial and jugular bulb saturations of oxyhaemoglobin; {Delta}SjO2 (95% CI): change in SjO2 from postinduction presented as mean (95% confidence intervals) – values not including zero indicate a significant difference from baseline (P<0.05), * indicate a significant (P<0.05) difference compared to abdominal aortic surgery group; PaCO2=arterial tension of carbon dioxide; MAP=mean arterial pressure; La and Lj=arterial and jugular venous concentrations of lactate
 
Repeated measures MANOVA revealed that, both during CPB and in the postoperative period, there were significant positive correlations between SjO2 and PaCO2 and Lj and the arterial concentration of lactate (Table 4). Examination of the regression coefficients revealed that for a 1 kPa decrease in PaCO2 the average decrement in SjO2 was 4% during CPB and 9% following surgery. For an increase in the arterial concentration of lactate of 1 mmol, Lj increased by an average of 0.9 mmol during CPB and postoperatively. None of the other covariates correlated significantly with SjO2 or Lj.


View this table:
[in this window]
[in a new window]
 
Table 4 Covariates included in the repeated measures analysis of variance. SjO2=oxygen saturation of jugular bulb haemoglobin; Lj=jugular bulb concentration of lactate; PaCO2=arterial tension of carbon dioxide; La=arterial concentration of lactate; MAP: mean arterial pressure; Hba=arterial concentration of haemoglobin; SaO2=oxygen saturation of arterial haemoglobin; CPB time=duration of cardiopulmonary bypass; B=regression coefficient; Beta=Beta coefficient (standardized Z-score form)
 
In patients undergoing CABG and heart valve surgery, there were no significant effects on SjO2or Lj by surgical operation group, time or their interaction during CPB. Over all surgical groups, SjO2 did not differ significantly between the groups (F=1.84, P=0.67, power=5%) in the postoperative period. However, there was a significant effect of time from operation upon SjO2 (F=22.51, P<0.001, power=100%) and a significant interaction between operation group and time from operation (F=3.45, P<0.001, power=98%). In the postoperative period, there were no significant effects upon Lj resulting from operation group (F=0.58, P=0.56, power=56%), time from operation (F=0.96, P=0.42, power=31%) nor because of their interaction (F=0.58, P=0.80, power=27%).

Post-hoc analysis
Post-hoc analysis found that SjO2 was decreased compared to baseline in all surgical groups on arrival on the ICU and 1 h later (Table 3). Two hours after arrival on the ICU, only the valve and abdominal aortic groups had significant decreases in SjO2 whilst 6 h after arrival, the decreases were significant only in the CABG and the abdominal aortic surgery groups. There were no significant differences in these declines between the surgical groups except 2 and 6 h after arrival on the ICU. After 2 h in the ICU, the decline in SjO2 was significantly less in patients undergoing CABG than abdominal aortic surgery whilst after 6 h in the ICU, the decline in the abdominal aortic surgery patients was significantly greater than either the CABG or valve surgery groups.

Duration of SjO2<50%
Factorial analysis of variance found that the model explained little if any of the variance (adjusted r2=0.001) in duration of SjO2<50% (Table 5). There were no significant differences between the groups (F=1.65, P=0.20, power= 34%) nor were there any significant effects because of age (B=0.02, P=0.52), CPB time (B=0.01, P=0.34) nor time until tracheal extubation (B=–0.076, P=0.75).


View this table:
[in this window]
[in a new window]
 
Table 5 Duration of jugular bulb oxyhaemoglobin desaturation. CABG=coronary artery bypass grafting surgery; Aortic=abdominal aortic surgery; Valve=heart valve surgery; Monitoring=duration monitored; SjO2<50%: duration that the saturation of jugular bulb was less than 50% (minutes); Percentage=percentage of monitored time with SjO2<50%; Per 25 and Per 75=25th and 75th percentiles
 

    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The present study confirms the findings of our initial observation, preliminary report and subsequent work that SjO2<50% occurs in many patients in the early postoperative period following CABG surgery.2 9 13 Further, it indicates that SjO2<50% is also a phenomenon of heart valve surgery. What is less clear is whether SjO2<50% also occurs following abdominal aortic surgery as the findings from fibreoptic and bench oximetry are conflicting. Our assumption that SjO2<50% only happens following cardiac surgery was ill founded as patients having abdominal aortic surgery had a similar duration of SjO2<50% when using fibreoptic measurements. However, the analysis indicates our study has only a limited statistical power to exclude the possibility that there are small but true differences in the duration of SjO2<50% between the groups.

In contrast to the fibreoptic method, bench oximetry measurements of discrete jugular bulb blood samples found that no patients undergoing abdominal aortic surgery experienced SjO2<50%. This difference may be because fibreoptic measurements of SjO2<50% occurred at times other than when blood samplings were performed as all discrete measurements of SjO2 were greater than 50% in patients having abdominal aortic surgery. Indeed, inspection of the data logs of patients having abdominal aortic surgery found that most of the episodes of SjO2<50% occurred between 6 and 18 h surgery. In contrast, when using discrete sampling, most of the episodes of SjO2<50% occurred within 6 h in patients having heart surgery. As measurements were only compared for 6 h after arrival in the ICU, it cannot be excluded that fibreoptic measurements of SjO2<50% beyond this time were artefactual. This may have arisen because tracheal extubation was performed earlier in patients having abdominal aortic surgery than in those having cardiac surgery. Patients having abdominal aortic surgery may therefore have been more likely to move and so cause the fibreoptic catheter to give artefactual signals when it impinged on the wall of the jugular bulb. However, this hypothesis is not supported by the lack of correlation between the time to tracheal extubation and duration of SjO2<50%.

Although none of the patients undergoing abdominal aortic surgery experienced SjO2<50% when discrete blood sampling and bench oximetry was used, they did have the highest mean decreases in SjO2 compared to baseline. This fits with the baseline values of SjO2 following induction of anaesthesia being higher, and postoperative values being lower, in patients undergoing abdominal vascular surgery than those having either the CABG or heart valve surgery. Thus, there are between group differences in both baseline values and changes in SjO2 following surgery that require explanation.

One possible explanation is that these differences were the result of differences in MAP.11 12 However, unlike the studies of Yoshitake and colleagues and Grubenhofer and colleagues, we found no significant correlation between MAP and SjO2 during CPB.11 12 In addition, we also found no significant correlation between SjO2 and MAP in the early postoperative period. This latter finding is supported by cerebral blood flow studies undertaken by McNeill and colleagues, which found cerebral autoregulation to be maintained at this time.18

In the present study, PaCO2 had an extremely important influence on SjO2, both during CPB and after surgery. This has been reported by Yoshitake and colleagues during CPB and by us, following surgery.9 11 Again, this is supported by McNeill and colleagues' study which found that cerebral blood flow is highly influenced by PaCO2.18 For this reason, although there was a marked decrement in SjO2 from baseline on CPB, it was not statistically significant after correction for the concomitant decrease PaCO2. Similarly, the changes in PaCO2 from baseline varied between surgical groups, most likely as a result of differences in ventilatory management, and this variance influenced the changes in SjO2 that occurred between the groups. Variance in PaCO2 between groups may explain why, when using discrete measurements, SjO2<50% was a phenomenon of cardiac but not abdominal aortic surgery.

Notwithstanding the important influence of PaCO2, the significant effect upon SjO2 resulting from the interaction between time from surgery and surgical group remains unexplained. Differences in anaesthetic technique could have played a part as Nandate and colleagues have found that the choice of anaesthetic agent influences SjO2 not only during CPB, but also after heart surgery.19 However, this study cannot exclude the possibility that changes in SjO2 following surgery are the result of pathological processes such as cerebral oedema or increased cerebrovascular resistance.20 21

Visual inspection of the Lj data suggests that patients having abdominal aortic surgery have lower Ljs than those undergoing cardiac surgery. This difference between operative groups, as well as that within patients, is explained by the positive correlation between Lj and the concentration of arterial lactate. Unlike patients undergoing abdominal aortic surgery, those having CPB will have received a large quantity of lactate from the lactated Ringer’s solution in the crystalloid prime for the CPB pump. However, repeated measures MANOVA, which having first corrected for the concentration of arterial lactate, found no significant effects on Lj. This implies that patients did not experience cerebral ischaemia as this would have increased cerebral lactate production.

The lack of reciprocation between levels of SjO2 and Lj indicates that SjO2<50% may not be a valuable estimate of cerebral ischaemia in this setting. Rather, SjO2<50% appears to be largely caused by hypocapnia inducing cerebral vasoconstriction. This in turn decreases cerebral blood flow and any decrement in cerebral oxygen delivery is compensated for by increased oxygen extraction. Therefore, SjO2<50% in the early postoperative period may not be an important cause of cerebral injury that is associated with heart surgery. Indeed, the findings of the present study explain why, in work undertaken after the present study, we were unable to find an important relationship between either the duration of SjO2<50% or Lj and cognitive outcome three months after CABG surgery.13 Future studies investigating brain damage associated with CABG surgery should therefore consider causal factors other than postoperative SjO2<50%.


    Acknowledgements
 
Financial support for this study was provided by Scottish Hospitals Endowment Research Trust Grant Number 1424. We thank Tim Howells PhD for adapting the MRC Browser program for use in this study.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Alston RP, Souter MJ. Cerebral sequelae of cardiac surgery. Curr Opinion Crit Care 2000; 6: 92–7

2 Souter MJ, Andrews PJD, Alston RP. Jugular venous desaturation following cardiac surgery. Br J Anaesth 1998; 81: 239–41[Abstract/Free Full Text]

3 Croughwell ND, Newman MF, Blumenthal JA et al. Jugular bulb saturation and cognitive dysfunction after cardiopulmonary bypass. Ann Thorac Surg 1994; 58: 1702–8[Abstract]

4 Andrews PJD, Colqhuhoun AD. Detection of cerebral hypoperfusion during cardiopulmonary bypass. Anaesthesia 1994; 49: 949–53[ISI][Medline]

5 Nakajima T, Kuro M, Hayashi Y, Kitaguchi K, Uchida O, Takaki O. Clinical evaluation of cerebral oxygen balance during cardiopulmonary bypass: On-line continuous monitoring of jugular venous oxyhemoglobin saturation. Anesth Analg 1992; 74: 630–5[Abstract]

6 Souter MJ, Andrews PJD, Alston RP. Propofol does not ameliorate cerebral venous oxyhemoglobin desaturation during hypothermic cardiopulmonary bypass. Anesth Analg 1998; 86: 926–31[Abstract]

7 Robertson CS, Gopinath RK, Goodman JC, Contant CF, Valadka AB, Narayan RK. SjVO2 monitoring in head-injured patients. J Neurotrauma 1995; 12: 891–6[ISI][Medline]

8 Sheinberg M, Kanter MJ, Robertson CS, Contant CF, Narayan RK, Grossman RG. Continuous monitoring of jugular venous oxygen saturation in head-injured patients. J Neurosurg 1992; 76: 212–17[ISI][Medline]

9 Millar SA, Alston RP, Souter MJ, Andrews PJD. Aerobic, anaerobic and combination estimates of cerebral hypoperfusion during and after cardiac surgery. Br J Anaesth 1999; 81: 936–9

10 Sapire KJ, Gopinath SP, Thakar DR, Gabrielli A, Jones JW, Robertson CS. Cerebral oxygenation during warming after cardiopulmonary bypass. Crit Care Med 1997; 25: 1655–62[ISI][Medline]

11 Yoshitake A, Goto T, Baba T, Shibata Y. Analysis of factors related to jugular venous oxygen saturation during cardio pulmonary bypass. J Cardiothorac Vasc Anesth 1999; 13: 160–4[ISI][Medline]

12 Grubhofer G, Lassnigg AM, Schneider B, Rajek MA, Pernerstorfer T, Hiesmayr MJ. Jugular venous bulb oxygen saturation depends on blood pressure during cardiopulmonary bypass. Ann Thorac Surg 1998; 65: 653–8[Abstract/Free Full Text]

13 Robson MJ, Alston RP, Deary IJ, Andrews PJD, Souter MJ, Yates S. Cognition after coronary artery surgery is not related to postoperative jugular bulb desaturation. Anesth Analg 2000; 91: 1317–26[Abstract/Free Full Text]

14 Walsh TS, Lee A. Mathematical coupling in medical research: lessons from studies of oxygen kinetics. Br J Anaesth 1998; 81: 118–20[Free Full Text]

15 Norusis MJ. SPSS Advanced Statistics 6.1. Chicago: Prentice Hall, 1994

16 Millar SA, Alston RP, Souter MJ, Andrews PJD. Continuous monitoring of jugular bulb oxyhaemoglobin saturation using the Edslab dual lumen oximetry catheter during and after cardiac surgery. Br J Anaesth 1999; 82: 521–4[Abstract/Free Full Text]

17 Gopinath SP, Robertson CS, Contant CF et al. Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatr 1994; 57: 717–23[Abstract]

18 McNeill BR, Murkin JM, Farrar JK, Gelb AW. Autoregulation and the CO2 responsiveness of cerebral blood flow after cardiopulmonary bypass. Can J Anaesth 1990; 37: 313–7[Abstract]

19 Nandate K, Vuylsteke A, Ratsep I et al. Effects of isoflurane, sevoflurane and propofol anaesthesia on jugular venous oxygen saturation in patients undergoing coronary artery bypass surgery. Br J Anaesth 2000; 84: 631–3[Abstract]

20 Harris DNF, Biley SM, Smith PLC, Taylor KM, Oatridge A, Bydder GM. Brain swelling in first hour following coronary artery bypass surgery. Lancet 1993; 342: 586–7[ISI][Medline]

21 Prough DS, Rogers AT, Stump DA et al. Cerebral blood flow decreases with time whereas cerebral oxygen consumption remains stable during hypothermic cardiopulmonary bypass in humans. Anesth Analg 1991; 72: 161–8[Abstract]