A series of case reports reporting dramatic adverse neurological outcomes in patients after shoulder surgery in the upright position have recently alarmed the surgical and anaesthetic communities. The complications ranged from cranial nerve injury1 to visual loss2 and cerebral infarction,3,4 occurring in relatively healthy middle-aged patients considered to be at low risk for cerebrovascular incidents. Although the exact pathogenesis of these events remains largely unexplained, it has been assumed that the specific positioning of the patient for shoulder surgery, that is, the beach chair position, may be responsible for these complications. The beach chair position may be associated with malrotation of the head and result in mechanical obstruction of the cerebral vessels. It has also been suggested that such positioning may induce unfavourable haemodynamic alterations with cerebral tissue hypoperfusion related to the gravitational effect of upright tilting. Finally, management of shoulder surgery often includes general anaesthesia with controlled hypotension, which may further compromise cerebral blood flow. These reports mandate that more attention should be paid to the effects of patient positioning on cerebral perfusion and that in these circumstances, routine anaesthesia monitoring may not suffice.
Near-infrared spectroscopy (NIRS) provides a non-invasive, continuous method of measuring cerebral tissue oxygen saturation. This technique has been demonstrated repeatedly to provide an early warning sign of cerebral hypoperfusion during procedures with a high risk of adverse neurological outcomes.5 Recent reports have suggested that cerebral oximetry monitoring with NIRS may be a useful indicator of cerebral hypoperfusion during shoulder surgery in the beach chair position.6,7
The aim of this prospective, observational, blinded study was to evaluate the prevalence of regional cerebral oxygen desaturation in patients undergoing shoulder surgery in the beach chair position when routine anaesthesia management and standard monitoring are used. We also aimed to identify potential physiological changes that might be associated with postural cerebral oxygen desaturation.
Materials and methods
Ethical approval for this observational study (Ethical Committee N° 2008/191) was provided by the Ethical Committee of the Ghent University Hospital, Gent, Belgium (Chairperson Professor Dr R. Rubens) on 30 April 2008. After written informed consent, 20 consecutive unpremedicated adult patients scheduled for elective shoulder surgery in the beach chair position were included. Exclusion criteria were clinically apparent neurological or cognitive dysfunction.
Standard monitoring was used throughout the procedure, including electrocardiography, pulse oximetry (SpO2), end-tidal oxygen, carbon dioxide and sevoflurane concentrations and non-invasive blood pressure measurement (AS3; Datex, Helsinki, Finland). Blood pressure measurements were performed at 3-min intervals with a non-invasive cuff placed on the arm opposite the operated side. Disposable NIRS sensors were applied on each side of the forehead for continuous registration of the regional cerebral oxygen saturation (rScO2) of the corresponding brain hemisphere (INVOS 5100; Somanetics Corporation, Troy, Michigan, USA).
Anaesthesia was induced with sufentanil (0.1–0.3 μg kg−1), propofol (2–3 mg kg−1) and cisatracurium (0.1 mg kg−1). Anaesthesia was maintained with sevoflurane (1.5–2.5% end-tidal concentration) in an oxygen/air mixture (50% oxygen) and additional doses of sufentanil (0.1–0.2 μg kg−1), if needed. All patients were raised to a 60–70° sitting position. The head of the patient was fixed in the mid-line. Management of anaesthesia and haemodynamics were left completely to the discretion of the attending anaesthesiologist, who was blinded to the rScO2 data, and who also was not informed about the purpose of the study. Routine clinical practice was used to maintain blood pressure. No deliberate hypotension was used. Systolic arterial pressure (SAP) of less than 80 mmHg or heart rate of less than 50 beats min−1 were usually treated.
Heart rate, non-invasive blood pressure, inspired and end-expiratory gas tensions, SpO2 and bilateral rScO2 were recorded continuously with RUGLOOP (Demed, Temse, Belgium). Blood loss was estimated and types and volumes of all fluids administered were recorded, as well as doses of all drugs given. At the postoperative visit by the responsible anaesthesiologist on the evening of surgery, the patient was assessed neurologically with a gross motor and sensory neurological evaluation and a gross cognitive evaluation (orientation in time and space, recall of name, date of birth and address). Any side-effects were recorded.
The rScO2 values were compared at different time points (awake, last value before position change from supine to upright position, 5 min after position change to beach chair and at the minimum rScO2). Changes in cerebral oxygen saturation were described in absolute terms (absolute rScO2 <50%) and in relative terms (>20% decrease in rScO2 compared to the value before position change). A rScO2 desaturation score was calculated by multiplying rScO2 below 50% with the duration of this event (in seconds).8 Hence, the rScO2 desaturation score generated is an area under the curve measurement, which accounts for both severity and duration of desaturation.
Sample size calculation was based on the assumption that a relative decrease in saturation of 20% (the smallest effect to be clinically important9) would be detected. Based on the study of Kim et al.10 a mean and standard deviation (SD) of 71% and 6% respectively were chosen. For a power of 0.8 and an α of 0.05, a sample size of 20 patients was calculated to be appropriate to detect a clinically relevant decrease in cerebral oxygen saturation. Statistical analysis was performed using the statistical software PASW Statistics 18 (SPSS Inc., Chicago, Illinois, USA). Data were tested for normal distribution using the Shapiro–Wilk test. Normally distributed continuous data are presented as mean ± SD. The rScO2 values at different time points were compared using an analysis of variance for repeated measurements (ANOVA). Post-test pairwise comparison was performed with the Tukey test. Possible relationships between rScO2 and physiological variables were analysed using linear regression analysis and quantified using the Spearman's correlation tests. A value of P less than 0.05 was considered statistically significant.
Patient demographics are presented in Table 1. One patient had undergone a carotid endarterectomy 4 years earlier, and another patient had suffered from a transient ischaemic attack (TIA) 8 years earlier; neither patient had any residual symptoms.
Table 2 shows details of surgery and anaesthesia. All patients received crystalloid solutions. If more than 1000 ml of crystalloids were needed, colloids were administered (four patients). The surgical procedures did not necessitate the use of deliberate hypotension and, therefore, routine clinical practice was used. Seven patients received ephedrine (5–10 mg) and one patient required phenylephrine 100 μg to correct a SAP of less than 80 mmHg. Three patients needed atropine 0.5 mg for bradycardia (heart rate <50 beats min−1).
The postural changes in cerebral oxygen saturation and blood pressure are shown in Table 3. When the beach chair position was adopted, rScO2 decreased significantly from 79 ± 9 to 57 ± 9% on the left side and from 77 ± 10 to 59 ± 10% on the right side (P < 0.001 for both sides compared with baseline, no significant difference between sides). A relative decrease in rScO2 of more than 20% occurred in 80% of patients when the beach chair position was adopted. In 30% of patients, rScO2 decreased to an absolute value below 50%. Desaturation scores ranged from 0 to 725% s (mean ± SD: 89 ± 221% s) on the left side and from 0 to 3360% s (178 ± 750% s) on the right side. One patient had a desaturation score of more than 3000% s.
The patient who had undergone carotid endarterectomy had a maximum relative decrease in rScO2 of 12.7% (minimum rScO2 55%). The patient with a history of TIA had a relative decrease of 30.4% (minimum rScO2 64%).
At all time points, rScO2 was negatively correlated with age (Fig. 1a), but the magnitude of the decrease in rScO2 with position change was independent of age (Fig 1b).
Before the position change, there was a positive correlation between rScO2 and end-tidal carbon dioxide tension (EtCO2; r = 0.53, P = 0.016). No correlation was found between rScO2 and SAP (r = 0.28, P = 0.086). Five minutes after the change in position, a positive correlation was found between rScO2 and EtCO2 (r = 0.56, P = 0.013), but no correlation was found between rScO2 and SAP (r = 0.29, P = 0.076). At the minimum rScO2, there were positive correlations between rScO2 and EtCO2 (r = 0.47, P = 0.035) and between rScO2 and SAP (r = 0.60, P = 0.007).
A representative case is shown in Figure 2. The rScO2 decreased promptly when the position was changed from supine to the beach chair position and immediately recovered when the supine position was restored. Changes in rScO2 closely paralleled changes in blood pressure, which is also apparent from the effects of ephedrine.
None of the patients developed gross neurological or cognitive dysfunction postoperatively.
In the present study, we observed cerebral desaturation (a relative decrease in rScO2 of >20%) in 80% of patients when an upright position was adopted during shoulder surgery. Postural decreases in cerebral oxygenation were consistent and related to blood pressure and EtCO2.
Recently, three reports have described the value of NIRS in monitoring the adequacy of cerebral perfusion during shoulder surgery in the beach chair position. Fischer et al.6 reported a case showing causality between rScO2, mean arterial pressure (MAP) and EtCO2. Murphy et al.7 evaluated the incidence of cerebral desaturation during shoulder surgery in the beach chair position compared to the lateral decubitus position. They used standardised anaesthetic management and aimed to optimise cerebral perfusion by maintaining MAP within 20% of baseline values and controlling EtCO2 between 30 and 34 mmHg. When cerebral oxygen desaturation was observed, they used a predetermined management protocol to increase rScO2. In the study of Tange et al.,11 perioperative management included support stockings, fluid administration at a rate of 10 ml kg−1 h−1 throughout the study period, gradual head-up tilt position and careful blood pressure management to maintain MAP above 60 mmHg. All these measures aimed to minimise the impact of position change. Our study is different in that the objective was to evaluate the actual prevalence of regional cerebral oxygen desaturation in patients undergoing surgery in the beach chair position when standard anaesthesia management and routine anaesthesia monitoring (which currently does not include cerebral oximetry) were employed. For this reason, anaesthetic management was left to the discretion of the attending clinicians who did not participate in the study and were blinded to the rScO2 data.
It is assumed that the upright position induces significant haemodynamic changes that may impair cerebral circulation. Compared to the supine position, assuming an upright position has been shown to decrease systolic and mean arterial pressure, stroke volume and cardiac output, inducing a cerebral blood flow decrease of 12%.12 In conscious individuals, these effects are compensated for by an increase in systemic vascular resistance, but during anaesthesia this autonomic response may be attenuated or blocked. The combination of the sitting position and general anaesthesia may, therefore, be potentially deleterious to cerebral perfusion. Reports describing cerebral ischaemia in the beach chair position have, therefore, stressed the risk of hypotension.13 It is often suggested that a systemic MAP between 50 and 150 mmHg lies within the range of cerebral autoregulation and, therefore, guarantees adequate cerebral perfusion. However, this assumption has been challenged. First, the concept of cerebral autoregulation is questioned, because there seems to be considerable individual variability in the autoregulation limits.14 Second, it has been claimed that blood pressure measured at the brachial artery may overestimate the pressure at the level of the brain when the sitting position is adopted. Some authors, therefore, propose an arithmetic correction of blood pressure to determine pressure at the level of the brain (1 mmHg for each 1.25 cm difference in height between the external meatus and the middle of the blood pressure cuff).13 Applying this assumption to the present study, the mean difference between the brain and the site of the blood pressure cuff on the arm of 31 ± 3 (24–38) cm, would overestimate arterial pressure at the level of the head by 24 ± 2 (18–29) mmHg. Third, cerebral perfusion pressure (CPP) depends not only on inflow pressure but also on outflow pressure: CPP = MAP−central venous pressure (CVP) or CPP = MAP−intracranial pressure (ICP) if ICP is more than CVP. In the upright position, the jugular veins either partly or fully collapse, and the vertebral venous plexus becomes the main pathway for venous drainage.12 Flow through this plexus might be impeded during head rotation and head tilt.
Although intuitively one would suppose that older patients have an increased risk of cerebral desaturation, our data showed no difference in the response to position change with age, which is consistent with the results of Gatto et al.15 and Edlow et al.16
The cerebral circulation is very responsive to changes in carbon dioxide tension. In a patient with normal carbon dioxide reactivity, cerebral blood flow changes by 1–2 ml 100 g−1 min−1 per mmHg change in carbon dioxide tension.17 This was also clear in our study, which showed a lower rScO2 with lower EtCO2.
Given the positive correlations between rScO2 and blood pressure and EtCO2, a simple recommendation could be to avoid hypotension and low EtCO2. The study by Tange et al.11 demonstrated that, with appropriate measures, cerebral desaturation can be avoided. However, three caveats should be considered. First, the ability of the circulatory system to compensate for sudden position changes varies considerably between patients and is unpredictable. Second, if deliberate hypotension is required, the impact on cerebral perfusion is unknown unless specific monitoring is employed. Finally, the high prevalence of cerebral desaturation in this study, together with the dramatic case reports, indicates that there is a compelling need for more wariness and for more specific monitoring. Due to the advantage of simple, continuous and non-invasive monitoring, NIRS could have the potential to optimise patient care in these situations.
Several limitations should be considered. First, because NIRS technology does not distinguish between arterial and venous haemoglobin saturations, changes in the proportion of cerebral arterial and venous blood volume may confound measurements.18 Changes in body position affect both arterial and venous pressures and can alter the ratio of arterial to venous compartments in the cerebral circulation. The measured changes in rScO2 in the upright position may, therefore, be a consequence of true changes in tissue oxygen tension, or may be the effect of the changes in the relative fractions of arterial and venous blood in the cerebral tissue. Real-time measurements of changes in cerebral blood flow using transcranial Doppler might be helpful in solving this question, but is difficult to establish properly during head-up table tilting and during surgical manipulation of the patient. Second, the prevalence of desaturation is dependent on the baseline value and the threshold used to define cerebral desaturation. In most clinical studies, a 20% reduction from awake values or an absolute decrease below 50% oxygen saturation is used as the threshold.9 Slater et al.8 also took the duration of desaturation into account by introducing the term ‘desaturation score’. Third, the relevance of the changes in cerebral oxygenation parameters might be questioned because of the absence of gross postoperative neurological dysfunction. However, without extensive neurocognitive monitoring, subtle changes induced by cerebral hypoxia may go unnoticed until functional organ damage becomes evident. Several randomised controlled trials have demonstrated that detection and treatment of cerebral oxygen desaturation results in better clinical outcomes.8,19–21
Finally, there is the interesting controversy with regard to the exact determination of CPP and its relationship with flow. The practice of correcting the blood pressure for position (raising the transducer or adjusting arterial pressure estimation to allow for the height between the heart and the brain) is dictated by the assumption that the heart lifts the blood to the elevated brain without a corresponding and equivalent venous return limb. In a closed loop system, such as the intact circulation, there is a continuous and balanced fluid column and no net work is performed against the effect of gravity.22 Therefore, as long as the hydrostatic gradient from the measurement site to the brain remains the same for inflow and outflow pressures, there is no significant flow-related pressure drop between the measurement site and the brain.23 Hence, there is no need for adjustment of pressure measurements if the assumption is correct that venous outflow depends only on the outflow pressure, although this does not take into account the more complex nature of the venous resistance and the possibility of venous collapse when assuming the upright position. Therefore, it has been proposed that the CPP formula should be adapted to take into account the effect of atmospheric pressure on the jugular veins. It has been suggested that measurement level adjustments are probably not necessary if CVP can be maintained above 18 mmHg.24
In conclusion, the results of the present study showed an incidence of 80% of cerebral oxygen desaturation when the beach chair position was adopted in patients undergoing shoulder surgery. Together with the high prevalence of cerebral desaturation also found in other studies, this underlines the need for awareness of this problem and suggests the need for perioperative monitoring of cerebral tissue oxygenation in this type of surgery. Monitoring, of course, should not preclude all customary measures being taken to avoid an abrupt decrease in blood pressure with position change.
The authors have no conflicts of interest to declare.
1. Boisseau N, Rabarijaona H, Grimaud D, Raucoules-Aimé M. Tapia's syndrome following shoulder surgery. Br J Anaesth
2. Bhatti MT, Enneking FK. Visual loss and ophthalmoplegia after shoulder surgery. Anesth Analg
3. Pohl A, Cullen DJ. Cerebral ischemia during shoulder surgery in the upright position: a case series. J Clin Anesth
4. Verbrugge SJ, Klimek M, Klein J. A cerebral watershed infarction after general anaesthesia in a patient with increased anticardiolipin antibody level. Anaesthesist
5. Casati A, Spreafico E, Putzu M, Fanelli G. New technology for noninvasive brain monitoring
: continuous cerebral oximetry
. Minerva Anestesiol
6. Fischer GW, Torrillo TM, Weiner MM, Rosenblatt MA. The use of cerebral oximetry
as a monitor of the adequacy of cerebral perfusion in a patient undergoing shoulder surgery in the beach chair position. Pain Pract
7. Murphy GS, Szokol JW, Marymont JH, et al. Cerebral oxygen desaturation events assessed by near-infrared spectroscopy
during shoulder arthroscopy in the beach chair and lateral decubitus positions. Anesth Analg
8. Slater JP, Guarino T, Stack J, et al. Cerebral oxygen desaturation predicts cognitive decline and longer hospital stay after cardiac surgery. Ann Thorac Surg
9. Edmonds HL Jr, Ganzel BL, Austin EH. Cerebral oximetry
for cardiac and vascular surgery. Semin Cardiothorac Vasc Anesth
10. Kim MB, Ward DS, Cartwright CR, et al. Estimation of jugular venous O2
saturation from cerebral oximetry
or arterial O2
saturation during isocapnic hypoxia. J Clin Monit Comput
11. Tange K, Kinoshita H, Minonishi T, et al. Cerebral oxygenation in the beach chair position before and during general anesthesia. Minerva Anestesiol
12. Alperin N, Hushek SG, Lee SH, et al. MRI study of cerebral blood flow and CSF flow dynamics in an upright posture: the effect of posture on the intracranial compliance and pressure. Acta Neurochir
13. Cullen DJ, Kirby RR. Beach chair position may decrease cerebral perfusion; catastrophic outcomes have occurred. APSF Newsletter
14. Drummond JC. The lower limit of autoregulation: time to revise our thinking? Anesthesiology
15. Gatto R, Hoffman W, Paisansathan C, et al. Effect of age on brain oxygenation regulation during changes in position. J Neurosci Methods
16. Edlow BL, Kim MN, Durduran T, et al. The effects of healthy aging on cerebral hemodynamic responses to posture change. Physiol Meas
17. Brian JE Jr. Carbon dioxide and the cerebral circulation. Anesthesiology
18. Pollard V, Prough DS, DeMelo AE, et al. The influence of carbon dioxide and body position on near-infrared spectroscopic assessment of cerebral hemoglobin oxygen saturation. Anesth Analg
19. Goldman S, Sutter F, Ferdinand F, Trace C. Optimizing intraoperative cerebral oxygen delivery using noninvasive cerebral oximetry
decreases the incidence of stroke for cardiac surgical patients. Heart Surg Forum
20. 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
21. Murkin JM, Adams SJ, Novick RJ, et al. Monitoring
brain oxygen saturation during coronary bypass surgery: a randomized, prospective study. Anesth Analg
22. Munis JR, Lozada LJ. Giraffes, siphons, and Starling resistors: cerebral perfusion pressure revisited. J Neurosurg Anesthesiol
23. Munis J. The problems of posture, pressure, and perfusion: cerebral perfusion pressure defined. APSF Newsletter
24. Pranevicius M, Pranevicius O. Modified calculation of the cerebral perfusion pressure in a sitting position: jugular Starling resistor and related clinical implications. APSF Newsletter