Marys Medicine

Tais-med.ru2

Neuroscience in Anesthesiology and Perioperative Medicine Section Editor: Gregory J. Crosby
Cerebral Oxygen Desaturation Events Assessed byNear-Infrared Spectroscopy During Shoulder Arthroscopyin the Beach Chair and Lateral Decubitus Positions Glenn S. Murphy, MD,* Joseph W. Szokol, MD,* Jesse H. Marymont, MD,* Steven B. Greenberg, MD,*Michael J. Avram, PhD,† Jeffery S. Vender, MD,* Jessica Vaughn, BA,* and Margarita Nisman, BA* BACKGROUND: Patients undergoing shoulder surgery in the beach chair position (BCP) may beat risk for adverse neurologic events due to cerebral ischemia. In this investigation, we soughtto determine the incidence of cerebral desaturation events (CDEs) during shoulder arthroscopyin the BCP or lateral decubitus position (LDP).
METHODS: Data were collected on 124 patients undergoing elective shoulder arthroscopy in theBCP (61 subjects) or LDP (63 subjects). Anesthetic management was standardized in allpatients. Regional cerebral tissue oxygen saturation (SctO ) was quantified using near-infrared spectroscopy. Baseline heart rate, mean arterial blood pressure, arterial oxygen saturation, andSctO were measured before patient positioning and then every 3 minutes for the duration of the surgical procedure. SctO values below a critical threshold (ⱖ20% decrease from baseline or absolute value ⱕ55% for ⬎15 seconds) were defined as a CDE and treated using apredetermined protocol. The number of CDEs and types of intervention used to treat low SctO2values were recorded. The association between intraoperative CDEs and impaired postoperativerecovery was also assessed.
RESULTS: Anesthetic management was similar in the BCP and LDP groups, with the exceptionof more interscalene blocks in the LDP group. Intraoperative hemodynamic variables did notdiffer between groups. SctO values were lower in the BCP group throughout the intraoperative period (P ⬍ 0.0001). The incidence of CDEs was higher in the BCP group (80.3% vs 0% LDPgroup), as was the median number of CDEs per subject (4, range 0 –38 vs 0, range 0 – 0 LDPgroup, all P ⬍ 0.0001). Among all study patients without interscalene blocks, a higher incidenceof nausea (50.0% vs 6.7%, P ⫽ 0.0001) and vomiting (27.3% vs 3.3%, P ⫽ 0.011) was observedin subjects with intraoperative CDEs compared with subjects without CDEs.
CONCLUSIONS: Shoulder surgery in the BCP is associated with significant reductions in cerebraloxygenation compared with values obtained in the LDP. (Anesth Analg 2010;111:496 –505) The beach chair position (BCP) has been used for position.3Althoughthesafetyoforthopedicsurgeryinthis shoulder arthroscopic procedures since the early position has been well established,4 rare catastrophic neu- 1980s. The advantages of the conventional BCP rologic events have been reported. Pohl and Cullen5 re- (45°–90° above the horizontal plane) include lack of bra- ported 4 cases of ischemic brain and spinal cord injury chial plexus strain, a reduced risk of direct neurovascular occurring after surgery in the BCP. In an additional report, trauma compared with the lateral decubitus approach, visual loss and ophthalmoplegia were described after excellent intraarticular visualization, and ease of conver- shoulder surgery in a sitting position.6 Eight intraoperative sion to an open approach if needed.1,2 In the United States, cerebrovascular events were reported in a survey of the approximately two-thirds of arthroscopic and open shoul- American Shoulder and Elbow Surgeons Society; all events der procedures are performed with the patient in the sitting occurred during surgery in the BCP.3 The etiology of central nervous system injury after shoul- der surgery in the BCP has not been established definitively.
From the *Department of Anesthesiology, NorthShore University Health-System, University of Chicago Pritzker School of Medicine, Evanston; and Several authors have hypothesized that cerebral ischemia †Department of Anesthesiology, Northwestern University Feinberg School may occur when anesthetized patients are placed in a 45° to of Medicine, Chicago, Illinois.
90° sitting position.5,7 In awake volunteers, sympathetic ner- Accepted for publication March 29, 2010.
vous system activation occurs when assuming a sitting posi- Supported by institutional and/or departmental sources (Department ofAnesthesiology, NorthShore University HealthSystem). FORE-SIGHT cere- tion; systemic vascular resistance and heart rate (HR) are bral oximetry probes were provided to the department at cost for the project increased to maintain mean arterial blood pressure (MAP) by CAS Medical Systems, Inc., Branford, CT.
and cardiac output.8–10 In anesthetized patients, however, the Address correspondence and reprint requests to Glenn S. Murphy, MD,Department of Anesthesiology, NorthShore University HealthSystem, Uni- response of the autonomic nervous system is attenuated by versity of Chicago Pritzker School of Medicine, 2650 Ridge Ave., Evanston, the vasodilating effects of IV and volatile anesthetics. Signifi- IL 60201. Address e-mail to [email protected].
cant decreases in cardiac output, MAP, and cerebral perfusion Copyright 2010 International Anesthesia Research SocietyDOI: 10.1213/ANE.0b013e3181e33bd9 pressure (CPP) have been observed in neurosurgical patients August 2010 • Volume 111 • Number 2

when position was changed from supine to sitting.11,12 Pro- BIS values between 40 and 60. In addition, patients received 1 longed reductions in systemic pressures and CPP that exceed to 2 ␮g 䡠 kg⫺1 䡠 h⫺1 fentanyl throughout the surgical proce- critical thresholds (severity and time) may result in perma- dure. If required, rocuronium (10-mg boluses) was adminis- nent neurologic injury.
tered to maintain a train-of-four count of 2 to 3. Ventilation Near-infrared spectroscopy (NIRS) is a noninvasive tech- was controlled to maintain end-tidal carbon dioxide (Etco2) nology that provides continuous monitoring of regional cere- between 30 and 34 mm Hg. Lower body forced-air warming bral tissue oxygen saturation (Scto devices (Bair Hugger威; Augustine Medical, Minneapolis, 2). NIRS technology allows for the immediate recognition and treatment of cerebral MN) were used to maintain core temperature above desaturation events (CDEs) that would otherwise be undetec- 35.0°C. Ondansetron 4 mg was given to all patients ted with conventional intraoperative monitoring. NIRS has within 30 minutes of tracheal extubation. Neostigmine 50 been used to assess the incidence of CDEs in patients under- ␮g/kg and glycopyrrolate 10 ␮g/kg were administered going procedures at high risk for adverse neurologic at a train-of-four count of at least 2 to reverse neuromus- outcomes (cardiac, vascular, liver transplant, and major ab- cular blockade at the conclusion of surgical closure.
dominal surgery).13 Although orthopedic patients in the BCP Clinicians were instructed to maintain MAP within 20% of are at risk for cerebral hypoperfusion, no previous clinical baseline values throughout the intraoperative period, as per trials have assessed changes in Scto standard practice involving surgery in the sitting position at 2 in this patient popula- tion. The aim of this prospective cohort study was to deter- our institution. Baseline MAP was determined in the sedated mine the incidence of CDEs in the BCP and compare this patient in the supine position. MAPs below this threshold cohort to subjects undergoing shoulder surgery in the lateral were treated with phenylephrine (80 ␮g), ephedrine (5 mg), or decubitus position (LDP). In addition, the relationship be- a fluid bolus, as required clinically. Patients undergoing tween CDEs and early clinical recovery was examined.
surgery in the LDP were treated with the same MAP protocol.
Cerebral Oxygenation Measurements and Patients and Anesthesia Perioperative Data Collection This study was approved by the IRB of NorthShore Uni- Cerebral oxygen saturation was measured continuously versity HealthSystem and written informed consent was using the FORE-SIGHT system (CAS Medical Systems, Inc., obtained from all subjects. Seventy consecutive patients Branford, CT). The FORE-SIGHT device is a continuous scheduled to undergo elective arthroscopic shoulder sur- wave, spatially resolved cerebral oximeter that uses 4 gery under general anesthesia in the BCP were enrolled discrete wavelengths of laser light to calculate the absolute (BCP group). During this same time period, 70 additional value of Scto2. Four wavelengths of light allow for more consecutive patients presenting for the same surgical pro- accurate determination of oxyhemoglobin and deoxyhemo- cedure in the right or left LDP were enrolled (LDP group).
globin levels by compensating for wavelength-dependent Exclusion criteria included preexisting cerebrovascular scattering losses and reducing interference from other disease or orthostatic hypotension; age ⬍18 years; ASA background light absorbers.17,18 Sensors were applied bi- physical status IV or V; or anticipated procedure under laterally to each frontotemporal area after cleansing the interscalene blockade with monitored anesthesia care as the skin area with alcohol (the medial margin at the midline of primary anesthetic. Allocation into the BCP and LDP the forehead [avoiding the temporalis muscle] and the groups was determined by surgical preference, and use of lower margin 1–1.5 cm above the eyebrow). The cerebral interscalene blocks in each group was also primarily influ- oximetry and BIS probes were secured in the preoperative enced by surgeon preference. Anesthesia care was admin- holding area and covered with an opaque wrapping to istered to the BCP and LDP groups by the same group of prevent light interference.
providers (15 anesthesiologists).
On arrival to the operating room, MAP and HR were Cerebral oximetry values may be affected by depth of recorded. Simultaneously, arterial oxygen saturation (Spo2) anesthesia, type of anesthetic administered, arterial carbon and Scto2 values were measured in patients before induc- dioxide concentrations, inspired oxygen content, and sys- tion of anesthesia while breathing a 50% air/oxygen mix- temic blood pressure management.14–16 Therefore, anes- ture. These variables (Scto2, MAP, HR, and Spo2) were then thetic management was carefully standardized in subjects manually recorded by a research assistant every 3 minutes in both cohorts. Patients received midazolam 2 mg IV for the duration of the operative procedure. Intraoperative before being transported to the operating room. Intraopera- Scto2 data were also collected continuously on a USB tive monitoring consisted of electrocardiography, auto- device. Baseline Scto2 was the mean value observed over a matic arterial blood pressure assessment using a cuff 1-minute period after induction of anesthesia during a placed on the nonoperative upper extremity, pulse oxim- stable interval (MAP within 20% baseline values, BIS etry, capnography, bispectral index monitoring (BIS威 40 – 60, Etco2 30 –34 mm Hg, and Fio2 50%) approximately system; Aspect Medical Systems, Newton, MA), and mea- 10 minutes after induction of anesthesia. Baseline measures surement of core temperature via an esophageal probe.
for MAP, HR, Spo2, and BIS used in the analysis were also Anesthesia was induced with propofol 2.0 to 2.5 mg/kg, recorded at this time. Five minutes after these initial data fentanyl 100 ␮g, lidocaine 50 mg, and rocuronium 0.6 to 0.8 were collected, the patient was positioned for the surgical mg/kg. Maintenance of anesthesia consisted of sevoflurane procedure. In the BCP group, the head was secured in a 1% to 3% in an oxygen/air mixture (fraction of inspired neutral position to ensure that cerebral venous drainage was not impaired. The back of the operating room table of 50%). Sevoflurane concentrations were adjusted to maintain MAP within 20% of baseline values and was then raised to 80° to 90° above the horizontal plane. In August 2010 • Volume 111 • Number 2 Cerebral Desaturation in the Beach Chair Position the LDP group, patients were placed in the LDP with the Discrete data were compared using Fisher exact test head supported with towels to maintained alignment with (NCSS, Kaysville, UT). The 95% confidence intervals for the the thoracolumbar spine. The NIRS monitor was positioned differences in percentages were calculated using the Far- rington and Manning score. Ordinal data and continuous 2 data could be viewed by the research assistant but not by clinicians providing intraoperative care. If any data that were not normally distributed are presented as median and range. These data were compared between 2 values below a previously defined critical threshold (ⱖ20% decrease from baseline or absolute value ⱕ55% for groups using the Mann-Whitney U test and within groups ⬎15 seconds) were observed by the research assistant, using Wilcoxon signed rank test (StatsDirect, Cheshire, clinicians were instructed to treat the CDE. For this investiga- UK). The median differences and their 95% confidence tion, a prioritized intraoperative management protocol was intervals were calculated.
used to increase Scto Normally distributed continuous data are presented as 2 values. Interventions to treat CDEs included the following: (1) increasing MAP with phenyleph- mean and SD. These data were compared using the un- rine (80 ␮g), ephedrine (5 mg), or a fluid bolus, as clinically paired t test (NCSS), except for the hemodynamic data.
indicated; (2) increasing Etco Mean differences and their 95% confidence intervals were 2 by decreasing ventilation; or (3) calculated. Hemodynamic, Spo 2 concentrations. The number and type of 2, BIS, and Scto2 data were interventions used to treat low Scto compared within and between groups using a 2-factor 2 values were recorded by the research assistant. In addition, interventions used by analysis of variance with repeated measures on 1 factor, clinicians to treat MAP reductions unrelated to Scto with the Holm-Sidak method for pairwise multiple com- noted. All data were collected until tracheal extubation.
parisons in post hoc analysis (SigmaPlot 11.0; Systat Soft- After discontinuation of sevoflurane at the end of the ware, Inc., San Jose, CA).
surgical procedure, the time required to achieve the following Multiple logistic regression analysis (StatsDirect) was end points were recorded: open eyes on verbal command; performed to determine predictors of nausea. Variables squeeze hand on verbal command; tracheal extubation; and included in the initial analysis were those identified in admission to postanesthesia care unit (PACU). The Aldrete univariate analyses as having a value of P ⬍ 0.10. In the score was recorded on arrival to the PACU and then assessed final analyses, variables with high P values were removed every 15 minutes until discharge. Hydromorphone was used from the model 1 at a time and were excluded from the for postoperative analgesia and titrated to achieve pain scores final model if their removal either did not diminish the fit of ⬍2 on a scale of 0 to 10 (0 ⫽ no pain, 10 ⫽ worst pain of the model or actually improved it, as determined by the imaginable). Any episodes of nausea and vomiting during the correct prediction of both positive and reference responses.
PACU admission were noted, and the severity of events The sensitivity and specificity of the logistic model were graded on a 3-point scale (1 ⫽ mild, 1 episode; 2 ⫽ moderate, calculated from the model-predicted reference and model- 2–3 episodes; and 3 ⫽ severe, ⬎3 episodes). The need for predicted positive responses (using the default threshold rescue antiemetics was also assessed. The times needed to probability for positive classification of 0.5) and the actual meet discharge criteria and achieve actual discharge were reference and actual positive responses. No external vali- noted. All recovery data were collected by PACU nurses dation was attempted.
blinded to intraoperative cerebral oximetry information.
The criterion for rejection of the null hypothesis estab- However, PACU nurses and research assistants were not lished a priori was a 2-tailed P ⬍ 0.05.
blinded to patient positioning.
Patient demographic data that were recorded included age, sex, height, weight, preoperative hemoglobin, ASA One hundred forty patients were enrolled in this clinical trial.
physical status, and preexisting medical conditions. Details Sixteen subjects were excluded from final analysis because of of the intraoperative anesthetic management included du- the following factors: changes in patient positioning (45° ration of anesthesia, administration of crystalloids, doses of beach chair) (4 in BCP group); protocol violations (3 in BCP opioids and rocuronium provided intraoperatively, and group and 3 in LDP group); incomplete data collection (2 in core temperatures at the conclusion of the anesthetic.
BCP group and 3 in LDP group); and procedure canceledbefore entering the operating room (1 in LDP group). As a Statistical Analysis result, data analysis was performed on 61 patients in the BCP Sample size was determined based on the primary outcome group and 63 patients in the LDP group. The BCP and LDP variable, the incidence of CDEs. Scto groups were similar in terms of demographic characteristics.
predetermined critical threshold (ⱖ20% decrease from There were no differences between groups in age, weight, baseline or absolute value ⱕ55% for ⬎15 seconds) were height, sex, preoperative hemoglobin values, preexisting used to define these events. In a pilot study of patients medical conditions, or ASA physical status (Table 1). Intraop- undergoing surgery in the BCP, CDEs were observed in erative management data are presented in Table 2. The 50% of the subjects. We hypothesized that we would duration of anesthesia was longer in the LDP group. A higher observe 50% fewer CDEs in patients having surgery in the percentage of patients in the LDP group received interscalene LDP. Group sample sizes of 66 in each group achieve 80% blocks (71.4%) compared with the BCP group (8.2%, P ⬍ power to detect a difference of 0.25 between the null 0.0001) and a lower dose of intraoperative fentanyl was used hypothesis that both group proportions are 0.50 and the in the LDP group (P ⬍ 0.0001).
alternative hypothesis that the proportion of the LDP group Hemodynamic data are presented in Figures 1 and 2.
is 0.25 with a 2-tailed significance level (␣) of 0.05 using ␹2 The ANOVA statistics revealed that whereas HR and MAP or Fisher exact test with continuity correction.
decreased in both the LDP and BCP groups after induction ANESTHESIA & ANALGESIA

Table 1. Patient Characteristics Beach chair group Difference (95% CI) Sex (male/female) 38 (62.3%)/23 (37.7%) 40 (63.5%)/23 (36.5%) ⫺1.2% (⫺18.0% to 15.6%) 1.9 (⫺3.2 to 7.1) ⫺3.8 (⫺10.5 to 3.0) ⫺3.1 (⫺6.8 to 0.6) ASA physical status Hemoglobin (g/dL) 14.0 ⫾ 1.5a 13.9 ⫾ 1.4b 0.1 (⫺0.5 to 0.6) 1.7% (⫺5.5% to 9.8%) ⫺3.1% (⫺11.7% to 4.5%) 6.1% (⫺11.0% to 22.9%) ⫺1.6% (⫺8.5% to 4.4%) 5.2% (⫺6.7% to 17.6%) 1.9% (⫺8.2% to 12.4%) ⫺3.0% (⫺13.1% to 6.6%) ⫺4.6% (⫺15.1% to 5.3%) ⫺1.2% (⫺13.3% to 10.9%) ⫺6.1% (⫺17.5% to 4.8%) CI ⫽ confidence interval; MI ⫽ myocardial infarction; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebrovascular accident; TIA ⫽ transient ischemicattack.
Data are mean ⫾ SD, median (range), or number of patients (%).
a n ⫽ 60.
b n ⫽ 52.
Table 2. Perioperative Variables Beach chair group Difference (95% CI) Interscalene block ⫺63.2% (⫺74.6% to ⫺48.5%) ⫺18.0 (⫺32 to 4) Dose fentanyl (␮g) Dose rocuronium (mg) ⫺0.09 (⫺0.23 to 0.05) Final OR temperature (°C) ⫺0.1 (⫺0.4 to 0.1) Times to recovery landmarks (min) Tracheal extubation PACUAldrete scores 8 (2–10)a 9 (2–10)a 9 (8–10)b 10 (8–10)c 10 (8–10)d 10 (8–10)e 9 (8–10)f 10 (9–10)g 31.3% (17.1%–45.1%) Nausea severity (1–3 scale) 1.5 (1–3)h 1 (1–2)i 19.7% (9.8%–31.8%) Vomiting severity (1–3 scale) 1 (1–2)j 33 (52.4%)a 32.9% (17.0%–47.3%) Pain medication dose (mg hydromorphone) 1.5 (0.5–4)l 1 (0.5–4.0)m PACU discharge (min) 83.5 (39–145)a ⫺1 (⫺13 to 10) 94 (55–181)a CI ⫽ confidence interval; OR ⫽ operating room; PACU ⫽ postanesthesia care unit.
Data are mean ⫾ SD, median (range), or number of patients (%).
a n ⫽ 62, b n ⫽ 58, c n ⫽ 57, d n ⫽ 18, e n ⫽ 27, f n ⫽ 9, g n ⫽ 4, h n ⫽ 24, i n ⫽ 5, j n ⫽ 13, k n ⫽ 1, l n ⫽ 52, m n ⫽ 33.
of anesthesia, intraoperative HR and MAP values did not Scto2 data are presented in Figure 6 and Table 3. Scto2 differ between groups. No differences between groups values before (75.5 ⫾ 4.0 vs 75.9 ⫾ 3.9) and after (baseline: were noted in Spo 80.4 ⫾ 5.0 vs 81.1 ⫾ 5.1) induction of anesthesia were 2 (Fig. 3), end-tidal sevoflurane concen- tration (Fig. 4), or BIS (Fig. 5) data throughout the intraop- similar between the LDP and BCP groups. The ANOVA erative period.
statistics revealed that Scto2 not only decreased over time August 2010 • Volume 111 • Number 2

Cerebral Desaturation in the Beach Chair Position Figure 1. Heart rate (bpm) for the patients in the beach chair position Figure 3. Arterial oxygen saturation (SpO , %) for the patients in the group and in the lateral decubitus position group. The data are beach chair position group and in the lateral decubitus position presented as mean ⫾ SD. The horizontal line indicates the time group. The data are presented as mean ⫾ SD. There were no during which the heart rates in the patients of the 2 groups differed differences between the groups at any time. The number of patients from their baseline heart rates (9 –90 minutes, overall P ⬍ 0.05).
in the beach chair position group decreased from 61 at baseline to There were no differences between the groups at any time. The 60 at 45 minutes and then progressively over time to 51 at 1 hour number of patients in the beach chair position group decreased from and to 20 at 90 minutes, whereas the number of patients in the 61 at baseline to 60 at 45 minutes and then progressively over time lateral decubitus position group decreased from 63 at baseline to 61 to 50 at 1 hour and to 20 at 90 minutes, whereas the number of at 48 minutes and then progressively over time to 54 at 1 hour and patients in the lateral decubitus position group decreased from 63 at to 40 at 90 minutes.
baseline to 61 at 48 minutes and then progressively over time to 54at 1 hour and to 40 at 90 minutes.
Figure 4. End-tidal sevoflurane concentrations for the patients in thebeach chair position group and in the lateral decubitus position Figure 2. Mean arterial pressure (mm Hg) for the patients in the group. The data are presented as mean ⫾ SD. There were no beach chair position group and in the lateral decubitus position differences between the groups at any time. The number of patients group. The data are presented as mean ⫾ SD. The horizontal line in the beach chair position group decreased from 61 at 15 minutes indicates the time during which the blood pressures in the patients to 60 at 45 minutes, to 58 at 60 minutes, to 45 at 75 minutes, and of the 2 groups differed from their baseline blood pressures (6 –90 to 25 at 90 minutes, whereas the number of patients in the lateral minutes, overall P ⬍ 0.05). There were no differences between the decubitus position group decreased from 63 at 15 minutes to 58 at groups at any time. The number of patients in the beach chair 45 minutes, to 55 at 60 minutes, to 49 at 75 minutes, and to 43 at position group decreased from 61 at baseline to 60 at 45 minutes and then progressively over time to 51 at 1 hour and to 20 at 90minutes, whereas the number of patients in the lateral decubitusposition group decreased from 63 at baseline to 61 at 48 minutesand then progressively over time to 54 at 1 hour and to 40 at 90 in the BCP group). Nine of the episodes resolved within 1 minute of treatment. The duration of the episodes rangedfrom 30 seconds to 9 minutes.
but also was lower in the BCP group than the LDP group Recovery data are presented in Table 2. Lower doses of across time (P ⬍ 0.0001). The percentage of patients devel- opioids were used in the LDP group in the operating room oping a CDE was higher in the BCP group (80.3%) com- and PACU, which was likely attributable to the higher use pared with the LDP group (0%, P ⬍ 0.0001). In addition, the of interscalene blocks in this group. The incidence of nausea median number of CDEs was greater in the BCP group: and vomiting was also lower in this group. The time from ⱖ20% decreases in Scto the end of the procedure to PACU admission was shorter, 2 from baseline (4 [0 –38] BCP; 0 [0 – 0] LDP; P ⬍ 0.0001) and Scto ⱕ and early Aldrete scores were higher, in the LDP group. All 55% (0 [0 – 4] BCP; 0 [0 – 0] LDP; P ⫽ 0.003). Furthermore, the median number of other immediate and early recovery variables were similar interventions required to treat CDEs was also greater in the between groups.
BCP group (2 [0 –11] BCP; 0 [0 – 0] LDP; P ⬍ 0.0001). Scto Further analysis was performed to determine the impact values increased after initial treatment interventions in 61% of intraoperative CDEs on postoperative recovery (Tables 4 of patients, with most responses occurring within 30 to 45 and 5). Data were analyzed only on subjects who did not seconds. Eight patients had 12 episodes of Scto ⱕ receive interscalene blocks because use of this technique ANESTHESIA & ANALGESIA was not equally distributed between the BCP and LDPgroups and interscalene blocks can beneficially influencerecovery from anesthesia. Patient characteristics did notdiffer between subjects with and without CDEs. Perioper-ative management variables, including opioid dosing in theoperating room and PACU, also did not differ betweengroups. A significantly higher incidence of nausea (50.0%vs 6.7%, P ⫽ 0.0001) and vomiting (27.3% vs 3.3%, P ⫽0.011) was observed in subjects with intraoperative CDEscompared with subjects with no CDE. All other recoverylandmarks (with the exception of Aldrete scores on PACUadmission) did not differ between groups.
To determine whether confounding variables may be Figure 5. Bispectral index (BIS) for the patients in the beach chair contributing to the nausea, age, sex, smoking status, hemo- position group and in the lateral decubitus position group. The dataare presented as mean ⫾ SD. There were no differences between globin concentration, operative position, fentanyl dose, the groups at any time. The number of patients in the beach chair end-tidal sevoflurane concentration, and the occurrence of position group increased from 57 at baseline to 59 at 6 minutes, cerebral desaturation were compared between patients then decreased to 58 at 45 minutes after which it decreased who had not had regional anesthesia and did not become progressively over time to 49 at 1 hour and to 18 at 90 minutes,whereas the number of patients in the lateral decubitus position nauseated and patients who had not had regional anesthe- group decreased from 63 at baseline to 61 at 48 minutes and then sia but did become nauseated. Variables identified in the progressively over time to 54 at 1 hour and to 40 at 90 minutes.
univariate analyses as having a value of P ⬍ 0.10 andincluded in the initial multiple logistic regression analysisas predictors of nausea were hemoglobin concentration,operative position, and the occurrence of cerebral desatu-ration. The only variable included in the final multiplelogistic regression model as a predictor of nausea was theoccurrence of cerebral desaturation (logit P ⫽ ⫺2.639 ⫹2.639 ⫻ [0 ⫽ no desaturation, 1 ⫽ desaturation]; likelihoodratio test statistic ⫽ 17.56, df ⫽ 1, P ⬍ 0.0001). Thesensitivity of the logistic model was 91.7%, whereas itsspecificity was 56.0%.
DISCUSSIONPatients undergoing shoulder surgery in the BCP may be at Figure 6. Regional cerebral tissue oxygen saturation (SctO ) for the risk for cerebral hypoxia because of decreases in CPP. NIRS patients in the beach chair position group and in the lateral technology, which provides information on the balance decubitus position group. The data are presented as mean ⫾ SD.
The solid horizontal line indicates the time during which the SctO between oxygen supply and demand in the frontal region values in the patients of the beach chair position group differed from of the brain, has not been used previously to determine their baseline SctO values (3– 0 minutes, overall P ⬍ 0.05) whereas changes in Scto2 in this patient population. Our results the dotted horizontal line indicates the time during which the SctO2 indicate that CDEs, defined as a ⱖ20% decrease in Scto2 values in the patients of the lateral decubitus position group differedfrom their baseline SctO values (12– 0 minutes, overall P values from baseline measurements or an Scto The dashed horizontal line indicates the times during which the SctO2 ⱕ55%, occurred frequently in patients having arthroscopic values in the patients of the beach chair position group differed from surgery in the BCP. Despite the use of a protocol designed those in the lateral decubitus position group (3–90 minutes, overall to optimize CPP, CDEs were observed in 80.3% of subjects.
P ⬍ 0.05). The number of patients in the beach chair position group In contrast, no CDEs were noted in a similar cohort of decreased from 61 at baseline to 60 at 45 minutes and thenprogressively over time to 51 at 1 hour and to 20 at 90 minutes, patients undergoing shoulder arthroscopy in the LDP. An whereas the number of patients in the lateral decubitus position association between intraoperative CDEs and postopera- group decreased from 63 at baseline to 61 at 48 minutes and then tive nausea and vomiting was also observed.
progressively over time to 54 at 1 hour and to 40 at 90 minutes.
Table 3. Primary Outcome Variables Difference or median Beach chair group difference (95% CI) Patients with cerebral desaturation events 80.3% (68.7%–88.4%) Interventions for SctO decreases Interventions for MAP decreases Episodes SctO ⱕ55 Episodes ⱖ20% decrease SctO CI ⫽ confidence interval; SctO ⫽ regional cerebral tissue oxygen saturation; MAP ⫽ mean arterial blood pressure.
Data are number of patients (%) or median (range).
August 2010 • Volume 111 • Number 2 Cerebral Desaturation in the Beach Chair Position Table 4. Patient Characteristics No interscalene block, no cerebral No interscalene block, cerebral desaturation events group desaturation events group Difference (95% CI) Sex (male/female) 18 (60.0%)/12 (40.0%) 28 (63.6%)/16 (36.4%) ⫺3.6% (⫺25.9% to 18.2%) 1.7 (⫺4.8 to 8.2) 3.9 (⫺5.8 to 13.6) 2.4 (⫺2.4 to 7.2) ASA physical status Hemoglobin (g/dL) 13.8 ⫾ 1.4a 13.8 ⫾ 1.6b ⫺0.1 (⫺0.8 to 0.7) 6.7% (⫺1.7% to 21.3%) 1.1% (⫺8.9% to 14.6%) ⫺4.2% (⫺25.7% to 18.4%) 3.3% (⫺4.9% to 16.7%) 3.0% (⫺13.3% to 21.8%) 4.2% (⫺10.4% to 21.8%) ⫺6.8% (⫺18.2% to 4.9%) ⫺0.2% (⫺12.8% to 15.3%) 8.8% (⫺4.3% to 25.8%) 12.1% (⫺1.6% to 29.7%) CI ⫽ confidence interval; MI ⫽ myocardial infarction; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebrovascular accident; TIA ⫽ transient ischemicattack.
Data are mean ⫾ SD, median (range), or number of patients (%).
a n ⫽ 29.
b n ⫽ 43.
Table 5. Perioperative Variables No interscalene block, no No interscalene block, cerebral desaturation cerebral desaturation Difference (95% CI) 60.0% (74.1%–90.6%) 10 (⫺7.8 to 27.4) Dose fentanyl (␮g) Dose rocuronium (mg) 0.13 (⫺0.07 to 0.33) Final OR temperature (°C) ⫺0.2 (⫺0.5 to 0.1) Times to recovery landmarks (min) Tracheal extubation PACUAldrete scores 10 (8–10)a 9 (8–10)b 9 (9–10)c 10 (8–10)d 9.5 (9–10)e 9 (8–10)f ⫺43.3% (⫺59.1% to ⫺24.0%) Nausea severity (1–3 scale) 1.5 (1–3)h ⫺23.9% (⫺39.3% to ⫺7.8%) Vomiting severity (1–3 scale) 1 (1–2)j 4.7% (⫺11.3% to 18.6%) Pain medication dose (mg 1.5 (0.5–4.0)k 1.5 (0.5–4.0)l PACU discharge (min) 2.5 (⫺12 to 17) 2.5 (⫺12 to 22) CI ⫽ confidence interval; OR ⫽ operating room; PACU ⫽ postanesthesia care unit.
Data are mean ⫾ SD, median (range), or number of patients (%).
a n ⫽ 29, b n ⫽ 43, c n ⫽ 16, d n ⫽ 12, e n ⫽ 4, f n ⫽ 6, g n ⫽ 2, h n⫽ 22, i n⫽ 1, j n ⫽ 12, k n ⫽ 28, l n ⫽ 39.
ANESTHESIA & ANALGESIA Alterations in systemic hemodynamics occur when pos- 1.35 cm).5,21 If the average vertical distance between the tural conditions are changed from supine to sitting. Reduc- brachial artery measurement site and cerebral oximeter tions in cardiac output, MAP, and CPP may subsequently probe was 38 cm (measured in 10 sitting patients), and the compromise cerebral oxygen delivery. In awake volunteers, average MAP intraoperatively was 80 mm Hg, a "cor- assumption of a head-up posture results in a 10% to 15% rected" MAP value of 51.8 mm Hg at the level of the frontal increase in HR, a 5- to 10-second decrease in MAP and cortex is derived. This "corrected" pressure might repre- systemic vascular resistance followed by a rapid rebound sent an inadequate CPP in some subjects. Although the and overshoot (15%– 40%), and a sustained reduction in lower limit of cerebral autoregulation is generally accepted thoracic blood volume and cardiac output (15%–30%).8–10 to be a MAP of approximately 50 mm Hg, some studies Compensatory increases in sympathetic nervous system have demonstrated that this lower threshold may be as activation and systemic vascular resistance to maintain high as 70 to 80 mm Hg in awake, normotensive subjects.22 systemic perfusion pressures are attenuated under general Consequently, if the "waterfall" theory is correct, mainte- anesthesia. Raising anesthetized neurosurgical patients nance of a MAP of 80 mm Hg using a manual blood from the supine to the sitting position resulted in signifi- pressure cuff might have represented suboptimal blood cant reductions in cardiac output, MAP, and CPP, which pressure management and accounted for the lower Scto2 persisted for up to 30 minutes after positioning.11,12 values in the BCP group.
The influence of alterations in positioning on cerebral Despite the frequent occurrence of CDEs in patients oxygenation in anesthetized surgical patients has been undergoing surgery in the sitting position, no obvious examined in only 1 previous trial. In these 12 subjects, no neurologic deficits were observed in this study cohort. This finding is not unexpected because major adverse cerebro- 2 values were observed in the LDP, but a small (6%), statistically significant decrease in Scto vascular events after orthopedic procedures in the BCP curred after assuming the sitting position.19 However, have been rarely described in the literature. Only 13 cases cerebral oximetry data were only collected for 5 minutes of stroke, coma, or blindness have been reported in this after each change in position. In the present investigation, patient population.3,5,6 At the present time, the incidence of clinically significant reductions in Scto permanent neurologic events after BCP surgery has not 2 values were absent when position was altered from supine to right or left been assessed in a prospective or retrospective investiga- lateral decubitus; Scto tion. However, a survey of the membership of the Ameri- 2 remained near baseline measures throughout the operative procedure. No episodes of CDEs can Shoulder and Elbow Surgeons yielded an estimated were recorded in any of the 63 LDP subjects, and no rate of stroke of 0.00382% to 0.00461% during shoulder interventions to treat low Scto surgery, with all events occurring in the BCP.3 2 were required. In contrast, pronounced reductions in Scto incidence of adverse neurologic outcomes is likely related 2 were observed in the BCP to the relatively limited duration of the surgical procedure.
2 decreased from a baseline mean of 80% to mean values of ⬍70% by 9 minutes, and the mean re- Severity and duration of ischemia are critical determinants mained between 66% and 70% throughout the operative of tissue damage, and viability-time thresholds must be procedure. The median number of CDEs was significantly exceeded to produce stroke. In a pig model, low Scto2 higher in the BCP group, despite the use of a similar values that persisted for ⬍2 hours did not result in neuro- protocol to maintain MAP. In addition, more interventions logic injury.23 An analysis of NIRS data from 265 coronary to treat reductions in Scto artery bypass graft patients revealed a desaturation-time 2 were required in the BCP group.
These findings suggest that CDEs (as defined in this threshold of 50 minutes that was associated with cognitive investigation) occur frequently during sitting position sur- decline and longer hospital length of stay.24 In the present gery and that cerebral oxygenation in the frontal cortices investigation, the duration of CDEs was limited, with no may potentially be compromised in these patients under- CDEs exceeding previously defined viability-time thresh- going general anesthesia.
olds. However, the degree and duration of cerebral isch- Hemodynamic and systemic oxygenation variables (HR, emia required to produce overt neurologic symptoms in a relatively healthy patient population is unknown at the 2) were not different between the BCP and LDP groups throughout the intraoperative period. HR and MAP present time. The use of a protocol designed to detect, treat, decreased from initial values in both groups over time, but and reduce the duration of CDEs (as used in this study) were within 20% of baseline measures. MAP is a primary would likely minimize the risk of obvious neurologic determinant of CPP and oxygenation, and reductions in MAP are associated with comparable decreases in A number of recovery variables were assessed in the PACU to determine whether CDEs in the operating room 2.15,20 MAP values, measured at the brachial artery, did not differ between the 2 groups of patients. It is possible, were associated with impaired early recovery from anes- however, that MAP measured at the brachial artery may thesia and surgery. A larger number of patients in the LDP overestimate the actual pressure at the level of the brain group received interscalene blocks, per surgeon preference.
when the sitting position is assumed. According to the The use of regional anesthesia may facilitate recovery from "open model" or "waterfall" theory, as blood flows verti- anesthesia and surgery; therefore, we performed an analy- cally from the heart, there is a reduction in arterial pressure sis only on subjects not administered regional anesthesia.
directly related to the weight of the column of blood.5,21 The times required to achieve immediate recovery land- When the sitting position is used, an arithmetic correction marks (time to open eyes, squeeze hand, tracheal extuba- of MAPs obtained from other sites is required to determine tion, and arrive in the PACU) and meet PACU recovery blood pressure at the level of the brain (1 mm Hg for each criteria did not differ between patients with and without August 2010 • Volume 111 • Number 2 Cerebral Desaturation in the Beach Chair Position CDEs. However, an approximately 7-fold higher incidence and subsequent Scto2 measures under very different physi- of both nausea and vomiting was observed in subjects with ologic conditions (awake versus anesthetized) would make CDEs. Some authors have suggested that an important the determination of changes in Scto2 due to alterations in perioperative cause of nausea and vomiting is anesthetic- patient positioning extremely difficult. Third, neurocogni- induced systemic hypotension, which produces a reduction tive testing to assess the presence or absence of subtle in cerebral perfusion and oxygenation.25 In patients under- neurologic dysfunction potentially related to CDEs was not going prostate resection surgery, spinal anesthesia resulted performed. In addition, transcranial Doppler, which can be in decreases in CPP and oxygenation, and an association used to indirectly measure cerebral blood flow changes between intraoperative CDEs and nausea at the end of related to alterations in position, was not used intraopera- surgery was observed.26 Our findings provide further tively. Fourth, arterial carbon dioxide levels were not support for an association between CDEs in the operating measured (ventilation was determined on the basis of Etco2 room and nausea and vomiting during early recovery.
values). Large inter- and intraindividual variations in arte- A CDE was defined in this investigation as a ⱖ20% rial to Etco2 gradients have been reported, and the degree of this variability may be influenced by patient position- 2 values from baseline measures or an Scto2 of ⱕ55%. At the present time, there is not a universally ing.39 Finally, cerebral oxygenation data were recorded accepted threshold used to identify pathological cerebral using the FORE-SIGHT cerebral oximeter. Another Food saturation. The threshold for identifying cerebral ischemia and Drug Administration–approved NIRS device (INVOS; may be influenced by a number of patient-specific (pres- Somanetics Corp., Troy, MI) has been used in the majority ence of cerebrovascular disease, incomplete circle of Willis) of previous perioperative studies, and there are fewer data or technology-dependent variables. Because of wide supporting a beneficial effect of the FORE-SIGHT cerebral patient-to-patient variability in baseline Scto oximeter on clinical outcomes. However, a recent volunteer study has demonstrated that the FORE-SIGHT monitor has authors recommend monitoring changes from baseline greater precision with respect to measuring absolute Scto measurements; a reduction of 15% to 20% from baseline has than the INVOS monitor.32 been used as a critical threshold in many investigations. In In conclusion, our findings demonstrate that significant awake patients undergoing carotid endarterectomy, a 20% reductions in Scto 2 occur when position is changed from 2 was associated with symptoms of cerebral supine to sitting in patients undergoing general anesthesia.
ischemia.27 In another group of carotid endarterectomy These changes occurred despite the use of a protocol patients, a 15% to 20% decrease in Scto2 was associated designed to maintain systemic MAP within 20% of baseline with a 20-fold increase in the odds for developing cerebral values. Furthermore, intraoperative CDEs were associated ischemia on electroencephalography.28 In addition, 15% to with a higher incidence of nausea and vomiting in the 25% decreases in Scto2 have been significantly correlated PACU. Future larger-scale investigations are required to with cognitive dysfunction after cardiac surgery,29 longer define the degree and duration of reduction in Scto PACU and hospital admissions after abdominal surgery,30 associated with permanent neurologic injury.
and greater release of biochemical markers of brain injuryafter liver transplantation.31 Preclinical studies suggest that AUTHOR CONTRIBUTIONS when quantitative NIRS technology is used, absolute Scto2 GSM helped design and conduct the study, analyze the data, values of ⱕ55% represent cerebral ischemia.18,32 Recent and write the manuscript. This author has seen the original studies have demonstrated that FORE-SIGHT– derived val- study data, reviewed the analysis of the data, approved the ues below this threshold were associated with adverse final manuscript, and is the author responsible for archiving outcomes after cardiac and aortic surgery.29,33 the study files. JWS helped design and conduct the study and There are several limitations to the present investigation.
write the manuscript. This author approved the final manu- First, NIRS devices measure saturation in an uncertain mix script. JHM helped design and conduct the study. This author of arterial, venous, and capillary compartments. In the approved the final manuscript. SBG helped conduct the study.
supine position, the venous contribution to cerebral oxim- This author approved the final manuscript. MJA helped designthe study, analyze the data, and write the manuscript. This etry predominates, with 70% to 84% of Scto2 values deter- author has seen the original study data, reviewed the analysis mined by venous blood.34,35 Changes in body position may of the data, and approved the final manuscript. JSV helped alter venous and arterial blood pressure and affect the ratio design the study. This author approved the final manuscript.
of the compartments in the cerebral circulation.36 There- JV helped conduct the study. This author approved the final fore, reductions in Scto2 may not only reflect decreases in manuscript. MN helped conduct the study. This author has oxygen supply but also changes in cerebral blood seen the original study data and approved the final manu- volumes/compartments. Second, baseline Scto2 values used to define a clinically significant reduction in cerebraloxygenation were measured after induction of anesthesia, because we were interested in assessing position-related GSM received honoraria from CASMED. All other authors report no conflicts of interest.
under identical anesthetic conditions (similar Fio2, MAP, Etco2, and BIS values). At the presenttime, there is no consensus on the setting under which baseline measures should be obtained; previous investiga- 1. Gelber PE, Reina F, Caceres E, Monllau JC. A comparison of risk between the lateral decubitus and the beach-chair position tors have collected these data before24,30 and after37,38 when establishing an anteroinferior shoulder portal: a cadav- induction of anesthesia. We believe that assessing baseline eric study. Arthroscopy 2007;23:522– 8 ANESTHESIA & ANALGESIA 2. Skyhar MJ, Altchek DW, Warren RF, Wickiewicz TL, O'Brien 24. Slater JP, Guarino T, Stack J, Vinod K, Bustami RT, Brown JM III, SJ. Shoulder arthroscopy with the patient in the beach-chair Rodriguez AL, Magovern CJ, Zaubler T, Freundlich K, Parr GV.
position. Arthroscopy 1988;4:256 –9 Cerebral oxygen desaturation predicts cognitive decline and 3. Friedman DJ, Parnes NZ, Zimmer Z, Higgins LD, Warner JJ.
longer hospital stay after cardiac surgery. Ann Thorac Surg Prevalence of cerebrovascular events during shoulder sur- gery and association with patient position. Orthopedics 25. Borgeat A, Ekatodramis G, Schenker CA. Postoperative nausea and vomiting in regional anesthesia: a review. Anesthesiology 4. Weber SC, Abrams JS, Nottage WM. Complications associated 2003;98:530 – 47 with arthroscopic shoulder surgery. Arthroscopy 2002;18:88–95 26. Atallah MM, Hoeft A, El-Ghorouri MA, Hammouda GE, Saied 5. Pohl A, Cullen DJ. Cerebral ischemia during shoulder surgery MM. Does spinal anesthesia affect cerebral oxygenation during in the upright position: a case series. J Clin Anesth transurethral prostatectomy? Reg Anesth Pain Med 1998; 6. Bhatti MT, Enneking FK. Visual loss and ophthalmoplegia after shoulder surgery. Anesth Analg 2003;96:899 –902 27. Samra SK, Dy EA, Welch K, Dorje P, Zelenock GB, Stanley JC.
7. Papadonikolakis A, Wiesler ER, Olympio MA, Poehling GG.
Evaluation of a cerebral oximeter as a monitor of cerebral Avoiding catastrophic complications of stroke and death re- ischemia during carotid endarterectomy. Anesthesiology lated to shoulder surgery in the sitting position. Arthroscopy 2000;93:964 –70 28. Rigamonti A, Scandroglio M, Minicucci F, Magrin S, Carozzo 8. Smith JJ, Porth CM, Erickson M. Hemodynamic response to the A, Casati A. A clinical evaluation of near-infrared cerebral upright posture. J Clin Pharmacol 1994;34:375– 86 oximetry in the awake patient to monitor cerebral perfusion 9. Frey MA, Tomaselli CM, Hoffler WG. Cardiovascular re- during carotid endarterectomy. J Clin Anesth 2005;17:426 –30 sponses to postural changes: differences with age for women 29. MacLeod D, White W, Ikeda K, Newman M, Mathew J.
and men. J Clin Pharmacol 1994;34:394 – 402 Decreased forebrain cerebral tissue oxygen saturation is asso- 10. Van Lieshout JJ, Wieling W, Karemaker JM, Secher NH.
ciated with cognitive decline after cardiac surgery. Anesth Syncope, cerebral perfusion, and oxygenation. J Appl Physiol Analg 2009;108(SCA Suppl):SCA 8 2003;94:833– 48 30. Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Danelli 11. Smelt WL, de Lange JJ, Booij LH. Cardiorespiratory effects of G, Fierro G, De Cosmo G, Servillo G. Continuous monitoring of the sitting position in neurosurgery. Acta Anaesthesiol Belg cerebral oxygen saturation in elderly patients undergoing major abdominal surgery minimizes brain exposure to poten- 12. Dalrymple DG, MacGowan SW, MacLeod GF. Cardiorespira- tial hypoxia. Anesth Analg 2005;101:740 –7 tory effects of the sitting position in neurosurgery. Br J Anaesth 31. Plachky J, Hofer S, Volkmann M, Martin E, Bardenheuer HJ, 1979;51:1079 – 82 Weigand MA. Regional cerebral oxygen saturation is a sensi- 13. Casati A, Spreafico E, Putzu M, Fanelli G. New technology for tive marker of cerebral hypoperfusion during orthotopic liver noninvasive brain monitoring: continuous cerebral oximetry.
Minerva Anestesiol 2006;72:605–25 transplantation. Anesth Analg 2004;99:344 –9 14. Fassoulaki A, Kaliontzi H, Petropoulos G, Tsaroucha A. The 32. MacLeod D, Ikeda K, Vacchiano C. Simultaneous comparison effect of desflurane and sevoflurane on cerebral oximetry of FORE-SIGHT and INVOS cerebral oximeters to jugular bulb under steady-state conditions. Anesth Analg 2006;102: and arterial co-oximetry measurements in healthy volunteers.
Anesth Analg 2009;108(SCA Suppl):SCA 56 15. Chu LC, Hsu YW, Lee TC, Lin YC, Huang CJ, Chen CC, Hung 33. Fischer G, Lin H, DiLuozzo G, Griepp R, Reich D. Decreased YC. Changes of regional cerebral oxygen saturation during cerebral tissue oxygen saturation during aortic surgery in- spinal anesthesia. Acta Anaesthesiol Taiwan 2007;45:155– 62 creases risk of postoperative complications. Anesth Analg 16. Lovell AT, Owen-Reece H, Elwell CE, Smith M, Goldstone JC.
2009;108(SCA Suppl):SCA 54 Continuous measurement of cerebral oxygenation by near 34. Kurth CD, Steven JM, Benaron D, Chance B. Near-infrared infrared spectroscopy during induction of anesthesia. Anesth monitoring of the cerebral circulation. J Clin Monit 1993;9:163–70 Analg 1999;88:554 – 8 35. Watzman HM, Kurth CD, Montenegro LM, Rome J, Steven JM, 17. Strangman G, Boas DA, Sutton JP. Non-invasive neuroimaging Nicolson SC. Arterial and venous contributions to near- using near-infrared light. Biol Psychiatry 2002;52:679 –93 infrared cerebral oximetry. Anesthesiology 2000;93:947–53 18. Fischer GW. Recent advances in application of cerebral oxim- 36. Pollard V, Prough DS, DeMelo AE, Deyo DJ, Uchida T, etry in adult cardiovascular surgery. Semin Cardiothorac Vasc Widman R. The influence of carbon dioxide and body position Anesth 2008;12:60 –9 on near-infrared spectroscopic assessment of cerebral hemo- 19. Fuchs G, Schwarz G, Kulier A, Litscher G. The influence of globin oxygen saturation. Anesth Analg 1996;82:278 – 87 positioning on spectroscopic measurements of brain oxygen- 37. Reents W, Muellges W, Franke D, Babin-Ebell J, Elert O.
ation. J Neurosurg Anesthesiol 2000;12:75– 80 Cerebral oxygen saturation assessed by near-infrared spectros- 20. Dunham CM, Sosnowski C, Porter JM, Siegal J, Kohli C.
copy during coronary artery bypass grafting and early post- Correlation of noninvasive cerebral oximetry with cerebralperfusion in the severe head injured patient: a pilot study.
operative cognitive function. Ann Thorac Surg 2002;74:109 –14 J Trauma 2002;52:40 – 6 38. Kussman BD, Wypij D, DiNardo JA, Newburger JW, Mayer JE 21. Drummond JC, Hargens AR, Patel PM. Hydrostatic gradient is Jr, del Nido PJ, Bacha EA, Pigula F, McGrath E, Laussen PC.
important: blood pressure should be corrected. APSF Newslett Cerebral oximetry during infant cardiac surgery: evaluation and relationship to early postoperative outcome. Anesth Analg 22. Drummond JC. The lower limit of autoregulation: time to revise our thinking? Anesthesiology 1997;86:1431–3 39. Grenier B, Verche re E, Mesli A, Dubreuil M, Siao D, Vanden- 23. Kurth CD, McCann JC, Wu J, Miles L, Loepke AW. Cerebral driessche M, Cale s J, Maurette P. Capnography monitoring oxygen saturation-time threshold for hypoxic-ischemic injury during neurosurgery: reliability in relation to various intraop- in piglets. Anesth Analg 2009;108:1268 –77 erative positions. Anesth Analg 1999;88:43– 8 August 2010 • Volume 111 • Number 2

Source: http://tais-med.ru/pdf/Murphy.pdf