Secondary Logo

Journal Logo

Original Article

A retrospective study of changes in cerebral oxygenation using a cerebral oximeter in older patients undergoing prolonged major abdominal surgery

Green, D. W.

Author Information
European Journal of Anaesthesiology (EJA): March 2007 - Volume 24 - Issue 3 - p 230-234
doi: 10.1017/S0265021506001645
  • Free



Several recent studies have indicated the high incidence of postoperative cognitive dysfunction (POCD) in elderly patients undergoing major surgery. In the International Study of Postoperative Cognitive Dysfunction (ISPOCD), an incidence of 25% POCD was found at 1 week and about 10% at 3 months, but no correlation could be determined between those patients who had POCD and conventional indicators such as intraoperative and perioperative blood pressure (BP) and oxygen saturation [1]. However, in the longer term 1–2 yr follow-up it would appear that POCD may eventually be reversible in the majority of patients [2].

Evidence is also now accumulating in patients undergoing liver transplantation, cardiopulmonary bypass (CPB) and major abdominal surgery that a fall in cerebral oxygen saturation (rSO2) of around 20% from baseline is associated with major cerebral dysfunction and POCD. Attempts to correct these falls in some studies have resulted in a decreased incidence of POCD. For example, Casati and colleagues showed that rSO2 monitoring in elderly patients undergoing major abdominal surgery reduces the potential exposure of the brain to hypoxia and this might be associated with decreased effects on cognitive dysfunction and shorter postanaesthesia care unit (PACU) and hospital stay [3].

The Somanetics Invos Cerebral Oximeter (SICO) has been used routinely in all elderly patients undergoing major abdominal surgery by the author since 2002. In view of the above, it was thus of interest to retrospectively analyse the data to see whether falls in rSO2 were a common accompaniment to this surgery and what the possible precipitating and alleviating factors were.


Data were obtained retrospectively on 46 consecutive patients aged 55 yr and above undergoing major abdominal surgery, who were anaesthetized by the author in the period August 2002–January 2005. King’s College Hospital Ethics Committee gave formal retrospective approval for this study. The surgery was likely to be associated with major fluid shifts and/or blood loss and thus all patients routinely received invasive monitoring, including central venous and intra-arterial pressure. All cases were performed under general anaesthesia, either total intravenous (TIVA) with propofol (13 patients) or inhalation with desflurane (33 patients). Analgesia was provided by continuous infusion of remifentanil, muscle relaxation was provided by cisatracurium and the patient’s lungs were ventilated with a combination of air and oxygen. No nitrous oxide was used. The SICO electrodes were placed on the right forehead in all patients prior to induction of anaesthesia and the values were recorded automatically at 20 s intervals until the end of the operation. Blood loss was recorded continuously using swab weighing and suction measurement and blood gases, including haemoglobin (Hb) levels, were estimated at frequent intervals throughout the procedures. All anaesthetic and physiological data were routinely recorded by the author automatically on a proprietary anaesthetic record computer and stored for future analysis. Blood gases, fluid inputs and outputs and data were recorded contemporaneously in a spreadsheet. Statistical analysis was carried out using MedcalcR (


There were 29 males and 17 females (American Society of Anesthesiologists (ASA) Grade mean 2.1 (range 1–4). Their mean ± standard deviation (SD) (range) age was 65 ± 8.8 (55–80) yr and weight 78 ± 16 (45–130) kg. The operations included abdominal aortic surgery (4), partial hepatectomy (6), radical prostatectomy (15), pancreaticoduodenectomy (9), cystectomy (4) and others (8).

Table 1 summarizes the intraoperative events, including duration, blood loss, blood replacement, starting Hb, lowest Hb recorded during the operation, mean rSO2 throughout the procedure and maximum percentage drop in rSO2. In 12 patients, the maximum drop in rSO2 equalled or exceeded 20%.

Table 1
Table 1:
Intraoperative data expressed as mean ± SD (range).

Table 2 compares two groups: Group 1, where the fall in rSO2 was 15% or more and Group 2, where the fall was 14% or less. Independent samples t-test was used to compare the two groups where data was normally distributed using the D’Agostino Pearson test. Otherwise, a U-test was used. There were no significant differences between the groups with respect to age, weight, ASA, duration, starting Hb or lowest Hb. However, there was a significant difference in total colloid replaced, blood loss and blood replaced between the two groups (P < 0.05). The type of anaesthetic agent used (propofol or desflurane) was equally distributed between the two groups. There was a not unexpected significant correlation between the percentage drop in Hb and both rSO2 and blood loss. However, there was no correlation between the fall in rSO2 and lowest Hb.

Table 2
Table 2:
Group 1 rSO2 drop 15% or greater vs. Group 2 rSO2 drop 14% or less.

Considering the six patients in whom the rSO2 drop exceeded 20% and in whom there was a significant association of the drop in rSO2 with ongoing haemorrhage, in only one patient was there a statistically significant correlation between drop in rSO2 and fall in systolic BP (Table 3). The scattergram showing the percentage drop in rSO2 vs. blood loss is shown in Figure 1.

Table 3
Table 3:
Correlation between fall in systolic BP and fall in rSO2 during haemorrhage and transfusion in six patients in whom the drop in rSO2 exceeded 20% of baseline.
Figure 1
Figure 1:
Scattergram and line of agreement between rSO2 percentage drop and blood loss in all patients. Correlation coefficient r = 0.3728, P = 0.0107.


This retrospective study has shown that a reduction in rSO2 of greater than 15% is a common finding in major surgery in elderly patients when associated with significant and rapid haemorrhage. This is an important finding as evidence is also now accumulating in patients undergoing carotid endarterectomy, CPB [4], liver transplantation [5] and major abdominal surgery [3] that a fall in rSO2 is associated with major cerebral dysfunction and POCD. Strategies to correct these falls in rSO2 may lead to improved cerebral function and a reduced incidence of POCD [3].

For the purposes of statistical analysis, it was arbitrarily decided to compare those patients in whom the fall in rSO2 was 15% or greater and those in whom it was 14% or less. It is not clear what level of fall in rSO2 is significant, but it is known from carotid endarterectomy surgery that a fall in rSO2 of greater than 20% following clamping is associated in most patients with marked deterioration in cerebral function, usually necessitating the insertion of a shunt [6]. Thus, it would seem reasonable in future studies to attempt to maintain rSO2 at values of at least 85% of the baseline.

Falls in rSO2 during cardiac surgery are significantly correlated with POCD and correction of these falls is associated with a reduced incidence of POCD [7]. However, doubt still remains as to whether there is an improvement in outcome using SICO in cardiac surgery [8]. The significant falls in rSO2 during cardiac surgery are almost always seen during the hypotension and haemodilution of CPB, which clearly represents a major change in circulatory physiology [9]. However, major blood loss and fluid shifts during other types of surgery may also cause similar falls in rSO2 sufficient to cause POCD, especially in the elderly.

A significant fall in rSO2 during the anhepatic phase of liver transplantation has been associated with raised postoperative levels of neuron-specific enolase (NSE) and S-100, which are specific variables that indicate cerebral disturbances due to hypoxia/ischaemia [5].

To evaluate whether monitoring and normalization of rSO2 minimizes intraoperative cerebral desaturation, 122 elderly patients undergoing major abdominal surgery with general anaesthesia were studied [3]. Patients were randomly allocated to an intervention group (the monitor was visible and rSO2 was maintained at >75% of pre-induction values; n = 56) or a control group (the monitor was blinded and anaesthesia was managed routinely; n = 66). Cerebral desaturation (rSO2 reduction > 75% of baseline) was observed in 11 patients of the treatment group (20%) and 15 patients of the control group (23%) (P = 0.82). To assess whether cerebral desaturation was associated with POCD, the mini-mental state examination (MMSE) was assessed. Although there was no overall statistical difference in outcome (as assessed by reduction in MMSE between the two groups), those patients who suffered a cerebral desaturation had a significantly lower MMSE at 7 days. A decline in cognitive function (defined as a reduction in MMSE ≥ 2 points from baseline) during the first week after surgery was observed in 20 patients of the treatment group (35%) and 30 patients of the control group (54%) (P = 0.137); however, when considering only those patients who actually had at least one episode of intraoperative desaturation in both groups, a decline in cognitive function was observed in 10 patients of the control group only (66%) (P < 0.001).

However, the latter study can be criticized on several grounds. Firstly, the MMSE is not a very good discriminator of POCD. Secondly, changes in Hb as a result of blood loss were not monitored. Thirdly, bolus doses of propofol were used to lower cerebral metabolic rate for oxygen (CMRO2) rather than using other methods, such as blood transfusion and optimization of fluid status, to improve cerebral oxygen delivery. Nevertheless, they showed that in elderly patients undergoing major surgery, there is a risk that falls in cerebral oxygenation may take place, which could result in POCD. Overall, though, the incidence of POCD in the treatment arm was still high at 35%, higher than the overall figure of 26% in the ISPOCD1 trial [1].

In this study, of the 46 patients, 11 had falls in rSO2 of greater than 20% and 23 (50%) of 15% or greater. These falls were almost always associated with a major and continuing haemorrhage resulting in a large fall in Hb. Despite the fact that BP and systemic oxygenation and carbon dioxide homeostasis were maintained, blood replacement seemed to be the only satisfactory intervention that consistently restored rSO2 to near-baseline reading. However, future studies would need to define other possible interventions (such as artificially raising BP and arterial carbon dioxide) that could reverse the changes in rSO2.

The transient falls in cerebral oxygenation would be unlikely to result in POCD as they were only for short periods. However, it is clear that nearly 30% of these patients suffered a decline in rSO2, which, during carotid endarterectomy for example, would be sufficient to cause an immediate deterioration in cerebral function and possibly POCD. It is possible that the falls in rSO2 would have continued for much longer if blood had not been administered. Although some preliminary work has shown that a fall in rSO2 using the SICO during haemorrhage may simply reflect a fall in Hb concentration due to changes in the path length of the measuring device [10], in this study there was no correlation between a fall in rSO2 and the lowest level of Hb obtained during the procedure. Torella and colleagues believe that the SICO may be useful in quantifying the effects of blood loss on cerebral oxygenation and may thus develop a useful role as a transfusion trigger [11–15]. This retrospective study appears to confirm their findings. Some might be dubious that stored, 2–3 dpg depleted red cells could correct the supposed cerebral oxygenation deficit so quickly. However, recent work has confirmed not only that acute anaemia does indeed produce cognitive deficit even in healthy volunteers but also that stored erythrocytes (3 weeks) are equally efficacious as erythrocytes stored for only 3.5 h in rapidly correcting the cognitive deficit [16,17]. It is unfortunate that these workers did not record changes in cerebral oxygenation during the removal and re-infusion of erythrocytes. However, using the SICO, Han and colleagues did find that acute normovolaemic haemodilution could lead to a reduction in cerebral oxygenation when a patient’s haematocrit declines to below 30% [18]. These changes may well be more severe in the elderly patient who cannot increase regional cerebral blood flow sufficiently to compensate for a reduced blood oxygen-carrying capacity. Levels of tolerable Hb levels in these patients may need to be redefined. In a recent review of anaesthetists’ decisions to transfuse in patients with coronary artery disease undergoing cardiac bypass surgery, a relatively low level of Hb was deemed to be tolerated. The study stressed the need for organ-specific measures of adequacy of oxygenation at relatively low levels of Hb (e.g. 8.0 g dL−1) to be employed. In cardiac surgical patients, this may include monitoring of ischaemia on ECG [19]. What similar measures are applicable for assessing adequacy of cerebral oxygenation apart from devices such as the SICO?

This study can obviously be criticized on the grounds that it is retrospective and non-randomized. However, it does support recent evidence of decline in rSO2 in elderly patients being a relatively common accompaniment of major surgery, especially as in this case when associated with major blood loss. Prospective studies should evaluate other interventions such as artificially raising systolic BP and arterial carbon dioxide in maintaining rSO2.

In conclusion, this study supports recent evidence that major falls in rSO2 occur in a high percentage of elderly patients undergoing major abdominal surgery. The falls were usually associated with blood loss and were correctable by blood transfusion in most cases. These changes could go unnoticed with conventional monitoring employed during anaesthesia. Evidence from other trials suggests that, if not promptly corrected, they could result in POCD.


1. Moller JT, Cluitmans P, Rasmussen LS et al. Long-term postoperative cognitive dysfunction in the elderly ISPOCD1 study. ISPOCD investigators. International Study of Post-Operative Cognitive Dysfunction. Lancet 1998; 351: 857–861.
2. Abildstrom H, Rasmussen LS, Rentowl P et al. Cognitive dysfunction 1–2 years after non-cardiac surgery in the elderly. ISPOCD group. International Study of Post-Operative Cognitive Dysfunction. Acta Anaesthesiol Scand 2000; 44: 1246–1251.
3. Casati A, Fanelli G, Pietropaoli P et al. Continuous monitoring of cerebral oxygen saturation in elderly patients undergoing major abdominal surgery minimizes brain exposure to potential hypoxia. Anesth Analg 2005; 101: 740–747.
4. Edmonds Jr HL, Ganzel BL, Austin III EH. Cerebral oximetry for cardiac and vascular surgery. Semin Cardiothorac Vasc Anesth 2004; 8: 147–166.
5. Plachky J, Hofer S, Volkmann M et al. Regional cerebral oxygen saturation is a sensitive marker of cerebral hypoperfusion during orthotopic liver transplantation. Anesth Analg 2004; 99: 344–349.
6. Samra SK, Dy EA, Welch K et al. Evaluation of a cerebral oximeter as a monitor of cerebral ischemia during carotid endarterectomy. Anesthesiology 2000; 93: 964–970.
7. Edmonds Jr HL. Multi-modality neurophysiologic monitoring for cardiac surgery. Heart Surg Forum 2002; 5: 225–228.
8. Taillefer MC, Denault AY. Cerebral near-infrared spectroscopy in adult heart surgery: systematic review of its clinical efficacy [[La spectroscopie cerebrale par infrarouge en cardiochirurgie chez l’adulte: une etude systematique de son efficacite clinique]. Can J Anaesth 2005; 52: 79–87.
9. Lassnigg A, Hiesmayr M, Keznickl P et al. Cerebral oxygenation during cardiopulmonary bypass measured by near-infrared spectroscopy: effects of hemodilution, temperature, and flow. J Cardiothorac Vasc Anesth 1999; 13: 544–548.
10. Yoshitani K, Kawaguchi M, Iwata M et al. Comparison of changes in jugular venous bulb oxygen saturation and cerebral oxygen saturation during variations of haemoglobin concentration under propofol and sevoflurane anaesthesia. Br J Anaesth 2005; 94: 341–346.
11. Torella F, McCollum CN. Regional haemoglobin oxygen saturation during surgical haemorrhage. Minerva Med 2004; 95: 461–467.
12. Torella F, Cowley R, Thorniley MS, McCollum CN. Monitoring blood loss with near infrared spectroscopy. Comp Biochem Physiol A Mol Integr Physiol 2002; 132: 199–203.
13. Torella F, Cowley RD, Thorniley MS, McCollum CN. Regional tissue oxygenation during hemorrhage: can near infrared spectroscopy be used to monitor blood loss? Shock 2002; 18: 440–444.
14. Torella F, Haynes SL, McCollum CN. Cerebral and peripheral near-infrared spectroscopy: an alternative transfusion trigger? Vox Sang 2002; 83: 254–257.
15. Torella F, Haynes SL, McCollum CN. Cerebral and peripheral oxygen saturation during red cell transfusion. J Surg Res 2003; 110: 217–221.
16. Spahn DR, Madjdpour C. Physiologic transfusion triggers: do we have to use (our) brain? Anesthesiology 2006; 104: 905–906.
17. Weiskopf RB, Feiner J, Hopf H et al. Fresh blood and aged stored blood are equally efficacious in immediately reversing anemia-induced brain oxygenation deficits in humans. Anesthesiology 2006; 104: 911–920.
18. Han SH, Ham BM, Oh YS et al. The effect of acute normovolemic haemodilution on cerebral oxygenation. Int J Clin Pract 2004; 58: 903–906.
19. Shehata N, Wilson K, Mazer CD et al. Factors affecting perioperative transfusion decisions in patients with coronary artery disease undergoing coronary artery bypass surgery. Anesthesiology 2006; 105: 19–27.


© 2007 European Society of Anaesthesiology