The number of aged patients requiring general anesthesia has increased to nearly 30% of the entire surgical population (1). This patient population is at increased risk for postoperative complications because of numerous comorbidities and reduction in physiologic reserve that accompanies aging (2,3), including cerebral ischemia and neurodegenerative disorders (4–6). Anesthesia has been recognized as a potential risk factor for postoperative neurodegenerative disorders (7). Nonetheless, the central nervous system, which is the primary end-point of most general anesthetics, is seldom directly monitored and anesthesia is usually managed using indirect variables of adequate brain oxygenation, such as heart rate, arterial blood pressure, and peripheral oxygenation.
In the last decade, technological research has expanded the application of near-infrared spectroscopy to allow continuous, noninvasive, and bedside monitoring of regional cerebral oxygen saturation (rSO2) through the scalp and skull (8), providing information on the occurrence of cerebral desaturation that is as accurate as information obtained by means of invasive techniques (9). This new noninvasive monitoring of rSO2 is clinically useful during open-heart surgery, in patients with head injury, and during carotid endarterectomy (10–12). A recent study on rSO2 monitoring during orthotopic liver transplantation also demonstrated a close correlation between a decline in rSO2 during the anhepatic phase and an increase in neuron-specific enolase used as index of cerebral damage resulting from hypoxia/ischemia (13). Furthermore, actively maintaining rSO2 above a “safety threshold” in patients undergoing coronary artery bypass surgery is associated with fewer complications and shorter hospital stays (14).
Cerebral desaturations have been reported in more than 20% of cases when monitoring rSO2 in elderly patients undergoing noncardiac abdominal surgery (15,16). We therefore conducted this prospective, multicenter, randomized, blinded investigation to evaluate whether using rSO2 monitoring in elderly patients undergoing major abdominal surgery reduces the potential exposure of the brain to inadequate oxygen supply.
The study was conducted at five University Hospitals after obtaining approval from each Institutional Ethic Committee and written informed consent from all patients. Patients older than 65 years, scheduled for major abdominal, nonvascular surgery under general anesthesia (with an expected duration ≥2 h) were considered for the study. Surgical procedures were represented by major abdominal surgery with a xiphopubic skin incision and included laparotomic gastric resection, colonic resection, hepatic resection, and duodenocephalo-pancreasectomy. Patients with preexisting cerebral pathology (such as previous episodes of cerebral ischemia or stroke) and ASA physical status ≥IV or a baseline Mini Mental State Examination (MMSE) test ≤23 and patients whose follow-up was not probable (not expected to be discharged alive from the hospital or with an expected hospital stay <4 days) were excluded.
Standard monitoring was used throughout the procedure. Patients were actively warmed throughout the procedure using a forced-air warming system (Bair-Hugger, Augustine-Medical, Eden Prairie, MN) to maintain intraoperative normothermia. rSO2 was continuously monitored using near-infrared spectroscopy (INVOS 4100; Somanetics Inc., Troy, MI). Sensors for cerebral oximetry were placed bilaterally on the right and left sides of the forehead according to the manufacturer’s instructions before induction of general anesthesia; rSO2 data were continuously recorded on a 3.5" floppy disk at 10-s intervals during the entire procedure.
According to a computer-generated sequence of numbers and using a sealed-envelope technique, patients were randomly allocated to two groups: in the control group the screen of the INVOS monitor was blinded and general anesthesia was managed routinely maintaining arterial blood pressure and heart rate values within 20% of baseline values. In the treatment group the INVOS monitor was visible and anesthesia management was aimed at maintaining rSO2 more than 75% of baseline values. Cerebral desaturation was defined as a reduction of rSO2 less than 75% of baseline (80% if the baseline value was <50%) for ≥15 s. In case of cerebral desaturation in the treatment group, the attending anesthesiologist activated a two-step treatment: the first step included checking the ventilator, head position, and tubing system, increasing Fio2, increasing end-tidal CO2 partial pressure if the ETco2 was <35 mm Hg, and increasing arterial blood pressure with intravascular fluid administration (250 mL hetastarch) and vasoconstrictors (ethylephrine 2–5 mg IV) if systolic arterial blood pressure was ≤90 mm Hg. If the first step did not restore acceptable rSO2 values within 60 s, the second step included the reduction of brain oxygen consumption with an IV bolus of propofol (0.5 mg/kg).
Baseline rSO2 was defined as the average saturation value over a 1-min period before induction of general anesthesia beginning approximately 3 min after the sensors were applied.
Standard general anesthesia was then induced with 2 μg/kg IV fentanyl and 5 mg/kg sodium thiopental IV. After administration of 0.2 mg/kg cisatracurium, tracheal intubation was performed and mechanical ventilation was started using a 40% oxygen-air mixture (tidal volume, 7 mL/kg; respiratory rate, 12 bpm; inspiratory-to-expiratory time, 1/2, with a 10% inspiratory pause). Mechanical ventilation was adjusted to maintain an ETco2 partial pressure ranging between 32 and 36 mm Hg. General anesthesia was maintained with sevoflurane by adjusting the end-tidal concentrations between 1% and 4% to maintain heart rate and mean arterial blood pressure within 15% of baseline values. Supplemental IV boluses of fentanyl (1 μg/kg) were administered if required to maintain cardiovascular stability, and muscle relaxation was maintained with 2-mg boluses of cisatracurium bromide at the discretion of the attending anesthesiologists and according to the surgeon’s needs. In case of blood loss exceeding 20% of whole circulating volume, the actual hemoglobin concentration was measured and homologous blood transfused if hemoglobin concentration was <9 g/dL.
After discontinuation of sevoflurane at the end of the procedure the time required to recover spontaneous ventilation, open eyes on verbal command, squeeze the observer’s hand, and time to tracheal extubation were recorded. Patients were then transferred to the postanesthesia care unit (PACU) and the Aldrete score was recorded every 15 min until achievement of readiness for PACU discharge (Aldrete score ≥9) (17). In each patient the anesthetic exposure was also calculated normalizing the end-tidal concentration of sevoflurane using the age-adjusted MAC values of sevoflurane (18).
Postoperative analgesia consisted of IV ketoprofen (100 mg every 8 h) and a patient-controlled analgesia infusion of morphine (1 mg bolus; lockout time, 6 min; maximum dose, 6 mg/h) for the first 72 h after surgery. Occurrence of any adverse event, including nausea, vomiting or other undesirable side effect, was also recorded.
Cognitive function of the patients was assessed using the MMSE test the day before surgery and then repeated 1 wk after surgery (19). In each hospital the same research fellow blinded to intraoperative treatment performed the MMSE test in a standardized manner. A decrease in MMSE score ≥2 points from baseline was also considered as an index of decline in cognitive function (20,21).
Patients were then followed daily after surgery until hospital discharge by the surgeons, ward nurses, and research fellows, who were blinded as to intraoperative management and patient grouping. Any occurrence of major respiratory, cardiovascular, surgical, or neurological complications requiring unplanned treatment was recorded. Duration of hospital stay was also recorded.
To calculate the required study size, we considered the results of previous studies performed in a similar population (15,16). To detect a difference of 4 points in mean rSO2 values recorded during surgery accepting a two-tailed α error of 5% and a β error of 10%, 54 patients per group were required (22).
Data from the cerebral oximeter were analyzed using Microsoft Excel 5.0 (Microsoft, Redmond, WA). Cerebral desaturation was considered to occur when rSO2 values decreased to <75% of baseline values for 15 s. If baseline rSO2 was less than 50%, the threshold for defining cerebral desaturation was a reduction to less than 80% of baseline values. Mean and minimum values of rSO2 were then calculated in each patient, as were the area under the curve of rSO2 readings less than 50% (AUCrSO2 < 50%) and the area under the curve of rSO2 readings less than 75% of baseline (AUCrSO2 < 75% of baseline).
Statistical analysis was performed using Systat 7.0 (SPSS Inc, Chicago, IL). The distribution of data was first evaluated using the Kolmogorov-Smirnov test. The Student’s t-test and Mann-Whitney U-test were used to compare continuous variables as indicated. Categorical variables were analyzed using contingency table analysis and Fisher’s exact test. Correlation and linear regression analysis were also used if indicated. Continuous variables are presented as mean ± sd or 95% confidence intervals (95% CIs) for normally distributed data or median (range) for not normally distributed data. Categorical variables are presented as numbers (percentage). A value of P ≤ 0.05 was considered as significant.
A total of 146 patients were screened for the study. Ten patients did not give their informed consent; 5 patients had a MMSE ≤ 23 and were excluded. A total of 122 patients completed the study in the 5 participating hospitals, 56 in the treatment group and 66 in the control group (Figure 1).
A center effect was excluded before performing complete statistical analysis of the whole population. No differences in anthropometric characteristics, duration of surgery, and anesthetic exposure were observed between the two groups (Table 1).
Surgery was performed uneventfully in all studied patients, and no severe intraoperative complication occurred in any patient. One patient of the control group died 20 days after surgery (colonic resection) (1.5%) as a result of the development of a surgical complication (rupture of the colonic anastomosis) leading to severe septic shock and multiple organ failure (P = 0.45).
No differences in arterial blood pressure values (systolic and diastolic), heart rate, or peripheral oxygen saturation were reported between the two groups. The maximum percentage decrease of systolic arterial blood pressure from baseline values was –24% ± 13% in the treatment group and -28% ± 13% in the control group (P = 0.12), and no differences in the median (range) number of observation times in which systolic arterial blood pressure was less than 80% of baseline values were reported between the two groups (2 [0–9] in the treatment group and 2 [0–12] in the control group; P = 0.16).
Table 2 shows the rSO2-derived variables recorded in the two groups and MMSE score recorded before surgery and 7 days after surgery. Although there was a large patient-to-patient variability in rSO2 values, baseline rSO2 values were similar in the two groups. Cerebral desaturation (defined as an 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). All episodes of cerebral desaturation occurred during the maintenance period of general anesthesia and the episodes were never associated with a concomitant reduction in arterial oxygen saturation. In 8 of the 11 patients of the treatment group with intraoperative desaturation, rSO2 recovered to acceptable values with only the first treatment step (increasing Fio2, ETco2 partial pressure, and arterial blood pressure), whereas in 3 of them a propofol bolus followed by a further increase in the sevoflurane end-tidal concentration was necessary to restore acceptable rSO2 values.
Postoperative complications requiring unplanned patient examination and diagnostic tests were observed in 12 patients of the control group (19%) and 6 patients of the treatment group (10%) (P = 0.20). Table 3 shows the distribution of different complications observed in the two groups. No statistically significant differences were observed between the two groups.
We then extrapolated from the whole population the 26 patients (11 in the treatment group and 15 in the control group) who actually had an episode of intraoperative desaturation. These two subgroups were similar with respect to anthropometric variables, intraoperative hemodynamics, duration of surgery, and mean baseline rSO2 values (61% [95% CI, 56%–72%] in the treatment group and 58% [95% CI, 53%–70%] in the control group (P = 0.169); however, rSO2 was higher in the treatment group (65% [95% CI, 59%–71%) than in the control group (52% [95% CI, 47%–57%]; P = 0.005), and mean AUCrSO2 <75% of baseline was much smaller in the treatment group (1.7 min% [95% CI, 0.8 min%–4.3 min%]) as compared with patients of the control group (292 min% [95% CI: 75 min%–660 min%]; P = 0.006).
No differences in MMSE score were observed between groups throughout the study (Table 2). However, when considering only those patients who had at least one episode of intraoperative desaturation, no differences were observed in median (range) baseline MMSE (29 [27–30]) in the treatment group and 29 [27–30] in the control group; P = 0.82), at the seventh postoperative day those patients of the control group who had intraoperative desaturation showed a significantly lower value of MMSE (26 [25–30]) as compared with patients of the treatment group (28 [26–30]) (P = 0.02). A significant correlation was also observed between the AUCrSO2 <75% of baseline and postoperative decrease in MMSE score from preoperative values in the subgroup of patients who had intraoperative cerebral desaturation (Figure 2) (2 outliers were excluded before performing the regression analysis).
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).
No differences in times for recovery from anesthesia after discontinuation of sevoflurane were reported between groups in the whole population (Fig. 3); however, patients of the control group who had intraoperative desaturation also showed a longer time for PACU discharge and hospital stay (47 min [13–56 min] and 24 days [7–53 days]) as compared with those patients of the treatment group (25 min [15–35 min] and 10 days [7–23 days]; P = 0.01 and P = 0.007, respectively). Interestingly, in the 26 patients who actually had intraoperative cerebral desaturation the AUCrSO2 <75% of baseline also showed a positive correlation with the duration of hospital stay after surgery (Figure 4).
All general anesthetics produce cardiovascular depression that can be augmented in the aged patient, potentially exposing him or her to inadequate brain perfusion (2,3), and this might be related to the occurrence of postoperative cognitive decline often reported in these patients (23,24). Nonetheless, the brain is rarely monitored in routine practice. Near-infrared spectroscopy monitoring has been reported to provide information on the balance between oxygen demand and supply to the brain (8,9), and the usefulness of intraoperative rSO2 monitoring has been shown during cardiac and vascular surgery (10–12,14). This multicenter, prospective, randomized, blinded investigation showed that in a population of relatively healthy, elderly patients cerebral desaturation can occur in one of every five patients. Adjusting the anesthesia plan according to rSO2 monitoring allowed us to minimize the exposure of the brain to a potentially inadequate oxygen supply.
Although no differences in MMSE were reported between groups in the entire population, when we considered only the 26 patients who actually had intraoperative cerebral desaturation (11 in the treatment group and 15 in the control group), the MMSE recorded at the seventh postoperative day was lower in the control group than in the treatment group, whereas the extent of MMSE reduction correlated with the exposure to potential brain hypoxia (AUCrSO2 <75% of baseline).
Evaluating cognitive function with the MMSE test is a major pitfall, as this test does not completely evaluate mental function (7,23,24); nonetheless, a reduction in the MMSE in repeated testing suggests that a decline in cognitive function is occurring (19–21). In agreement with our findings, Monk et al. (16) evaluated postoperative cognitive dysfunction with more sensitive psychometric tests in patients of different ages, reporting that in aged patients cognitive dysfunction was associated with a more frequent incidence of intraoperative desaturation and longer hospital stay.
In agreement with our findings, Austin et al. (25) reported that active interventions based on a multimodal neurophysiologic monitoring, including transcranial cerebral oximetry, decreased the incidence of postoperative neurological sequelae in a population of pediatric patients undergoing cardiac surgery, with marked reduction of hospital stay and costs. In a prospective, randomized, blinded study Iglesias et al. (14) evaluated the impact of interventions to improve rSO2 on postoperative outcome after coronary artery bypass surgery and reported that monitoring and maintaining rSO2 more than 75% of baseline was associated with a decreased length of stay after cardiac surgery.
Evaluating the intraoperative changes in rSO2 during orthotopic liver transplantation with postoperative increases in neuron-specific enolase as an index of hypoxia/ischemia-related cerebral damage, Plachky et al. (13) observed that a decrease in cardiac output alone does not predict neurological damage, but they reported a close correlation between decreased rSO2 readings and increased markers of neurological damage during the anhepatic phase. Accordingly, actively maintaining rSO2 greater than a critical threshold should potentially reduce the severity of ischemia-induced brain damage.
We observed no differences in the incidence of postoperative complications between groups. However, it must be considered that this study was powered to detect a difference in mean rSO2 between groups; whereas a post hoc power analysis based on the observed incidence of perioperative complications showed that more than 400 patients per group would be required to have a 5% α-error and 20% β-error in detecting a significant difference in the incidence of perioperative complication (22).
Importantly, the occurrence of cerebral desaturation was not associated with any difference in pulse oximetry. Although the study was not powered to evaluate the effects of pulse oximetry monitoring on patient outcome, this finding is in agreement with findings reported by Pedersen et al. (26), who failed to demonstrate any significant differences in the incidence of perioperative complications or duration of hospital stay by monitoring pulse oximetry perioperatively. Similar results have been reported by Plachky et al. (13) during liver transplantation.
In conclusion, this multicenter, prospective, randomized, blinded study shows that cerebral desaturation occurs in nearly 20% of elderly patients undergoing major abdominal nonvascular surgery; using rSO2 monitoring to adjust the anesthesia plan during surgery reduces the exposure of the brain to potential hypoxia. Interestingly, in those patients developing intraoperative cerebral desaturation this approach to anesthesia management seems to result in less cognitive decline and shorter hospital stay, suggesting that cerebral hypoxia might be involved in the physiopathology of cognitive decline observed after surgery in the aged patient
The study was entirely supported by funding of the five hospitals only. The authors thank Tyco Italy for providing the sensors used for the study at a special price.
The Collaborative Italian Study Group on Anesthesia in Elderly Patients. All clinicians participants are co-authors: Guido Fanelli, MD, Professor of Anesthesiology, Andrea Casati MD, Giorgio Danelli, MD, Massimiliano Nuzzi, MD, Federico Mentegazzi, MD, Andrea Fanelli, MD (Azienda Ospedaliera Parma, Parma); Carla Martani, MD, Emanuela Spreafico, MD (deceased) (Ospedale San Raffaele, Milano); Paolo Pietropaoli, MD, Professor of Anesthesiology, Giovanni Fierro, MD, Professor of Anesthesiology, Francesco Pugliese, MD (Policlinico Umberto I, Roma); Rodolfo Proietti, MD, Professor of Anesthesiology, Germano De Cosmo, MD, Paola Aceto, MD (Policlinico Gemelli, Roma); Rosalba Tufano, MD, Professor of Anesthesiology, Giuseppe Servillo, MD, Professor of Anesthesiology, Fabrizio Monaco, MD (Policlinico Federico II, Napoli).
1. Peduto VA, Chevallier P, Casati A, and the VIMA Group. A multicenter survey on anaesthesia practice in Italy. Minerva Anestesiol 2004; 70: 473–91.
2. Tonner PH, Kampen J, Scholz J. Pathophysiological changes in the elderly. Best Pract Res Clin Anaesthesiol 2003; 17: 163–77.
3. Liu LL, Wiener-Kronish JP. Perioperative anesthesia issues in the elderly. Crit Care Clin 2003; 19: 641–56.
4. Intiso D, Stampatore P, Zarrelli MM, et al. Incidence of first-ever ischemic and hemorrhagic stroke in a well-defined community of southern Italy. 1993–1995. Eur J Neurol 2003; 10: 559–65.
5. Mori E. Impact of subcortical ischemic lesions on behavior and cognition. Ann N Y Acad Sci 2002; 977: 141–8.
6. Burton DA, Nicholson G, Hall GM. Anaesthesia in elderly patients with neurodegenerative disorders: special considerations. Drugs Aging 2004; 21: 229–42.
7. Ancelin ML, DeRoquefeuil G, Ledesert B, et al. Exposure to anaesthetic agents, cognitive functioning and depressive symptomatology in the elderly. Br J Psychiatry 2001; 178: 360–6.
8. Edmonds HL Jr. Multi-modality neurophysiologic monitoring for cardiac surgery. Heart Surg Forum 2002; 5: 225–8.
9. Pollard V, Prough DS, DeMelo AE, et al. Validation in volunteers of a near-infrared spectroscope for monitoring brain oxygenation in vivo. Anesth Analg 1996; 82: 269–77.
10. Konishi A, Kikuchi K. Cerebral oxygen saturation (rSO2) during open heart surgery and postoperative brain dysfunction [in Japanese]. Masui 1995; 44: 1322–6.
11. Higami T, Kozawa S, Asada T, et al. Retrograde cerebral perfusion versus selective cerebral perfusion as evaluated by cerebral oxygen saturation during aortic arch reconstruction. Ann Thor Surg 1999; 67: 1091–6.
12. 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: 964–70.
13. Plachky J, Hofer S, Volkman M, et al. Regional cerebral oxygen saturation is a sensitive marker of cerebral hypoperfusion during orthotopic liver transplantation. Anesth Analg 2004; 99: 344–9.
14. Iglesias I, Murkin JM, Bainbridge D, Adams S. Monitoring oxygen saturation significantly decreases postoperative length of stay: a prospective randomised blinded study. Heart Surg Forum 2003; 6: 204.
15. Casati A, Fanelli G, Pietropaoli P, et al. In a population of elderly patients undergoing elective non-cardiac surgery cerebral oxygen desaturation is associated with prolonged length of hospital stay [abstract]. Anesthesiology 2003; 99: A551.
16. Monk TG, Weldon BC, Weldon JE, van der Aa MT, et al. Cerebral oxygen desaturations are associated with postoperative cognitive dysfunction in elderly patients. [abstract] Anesthesiology 2002; 97: A40.
17. Aldrete JA, Kroulik D. A postanesthetic recovery score. Anesth Analg 1970; 49: 924–34.
18. Nickalls RW, Mapleson WW. Age-related iso-MAC charts for isoflurane, sevoflurane and desflurane in man. Br J Anaesth 2003; 91: 170–4.
19. Folstein M, Folstein S, McHugh P. “Mini Mental State”. A practical method for grading the cognitive state of patients for clinician. J Psychiatr Res 1975; 12: 189–9.
20. Mondimore FM, Damlouji N, Folstein MF, Tune L. Post-ECT confusional states associated with elevated serum anticholinergic levels. Am J Psychiatry 1983; 140: 930–1.
21. Casati A, Aldegheri G, Vinciguerra E, et al. Randomized comparison between sevoflurane anaesthesia and unilateral spinal anaesthesia in elderly patients undergoing orthopaedic surgery. Eur J Anaesthesiol 2003; 20: 640–6.
22. Browner WS, Black D, Newman B, Hulley SB. Estimating sample size and power. In: Hulley, SB, Cummings SR, eds. Designing clinical research - an epidemiologic approach. Baltimore: Williams & Wilkins, 1988; 139–50.
23. Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term postoperative cognitive dysfunction in the elderly: ISPOCD1 study. Lancet 1998; 351: 857–61.
24. Dijkstra JB, Houx PJ, Jolles J. Cognition after major surgery in the elderly: test performance and complaints. Br J Anaesth 1999; 82: 867–74.
25. Austin EH, Edmonds HL, Auden SM, et al. Benefit of neurophysiologic monitoring for pediatric cardiac surgery. J Thorac Cardiovasc Surg 1997; 114: 707–15.
© 2005 International Anesthesia Research Society
26. Pedersen T, Pedersen D, Moller AM. Pulse oximetry for perioperative monitoring. Cochrane Database Syst Rev 2003; 3::CD002013.