- Email: [email protected]
Evaluation of cerebral oxygen perfusion during shoulder arthroplasty performed in the semi–beach chair position
J Shoulder Elbow Surg (2019) -, 1–7
www.elsevier.com/locate/ymse
Evaluation of cerebral oxygen perfusion during shoulder arthroplasty performed in the semi–beach chair position Justin H. Chan, MD*, Hector Perez, BA, Harrison Lee, BS, Matthew Saltzman, MD, Guido Marra, MD Department of Orthopaedic Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL, USA Background: The beach chair position is commonly used when performing shoulder arthroplasty. However, this position has been associated with hypotension, potentially leading to cerebral hypoperfusion, which may cause neurologic injury. In addition, shoulder arthroplasty cases are associated with longer operative times, posing a potentially greater risk of cerebral hypoperfusion. We aim to evaluate the risk of cerebral desaturation events (CDEs) during the course of total shoulder arthroplasty. Methods: Twenty-six patients undergoing shoulder arthroplasties were monitored for changes in cerebral perfusion. Seven specific time-points during the procedure were labeled for comparison of events: baseline, beach chair, incision, humeral broaching, glenoid reaming, glenoid component implantation, and humeral component implantation. Cerebral oxygen perfusion was measured using near-infrared spectroscopy. A CDE was described as a decrease of oxygen saturation greater than 20%. Results: Nineteeen of 25 subjects experienced a CDE. 42% of these patients experienced CDEs during semi–beach chair positioning. Patients experienced the largest oxygen saturation drop during semi–beach chair positioning. Transition from baseline to semi–beach chair was the only event to have a statistically significant decrease in cerebral perfusion (8%, P < .05). There was a statistically significant percentage change in mean oxygen saturation in the semi–beach chair interval (10%, P < .01) and the semi–beach chair to incision interval (7%, P < .01). Conclusions: Most patients experienced an intraoperative CDE, with greatest incidence during semi–beach chair positioning. The largest decline in cerebral oxygen saturation occurred during semi–beach chair positioning. Implant implantation was not associated with decrease in cerebral oximetry. Level of evidence: Level IV; Case Series; Treatment Study Ó 2019 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. Keywords: Total shoulder arthroplasty; beach chair position; semi–beach chair position; cerebral oxygenation; near-infrared spectroscopy
IRB approval by Northwestern University Institutional Review Board, approval number STU200834-MOD0003. *Reprint requests: Justin H. Chan, MD, Department of Orthopaedic Surgery, Northwestern University Feinberg School of Medicine, 676 N St Clair, Suite 1350, Chicago, IL 60611, USA. E-mail address: [email protected] (J.H. Chan).
During shoulder arthroplasty, the semi–beach chair position is exclusively used because it allows for anatomic shoulder positioning and easier shoulder access.19 The semi–beach chair position also allows for improved airway access, reduced bleeding risk and reduced brachial plexus injury risk.18 In addition, shoulder surgeons will often ask
1058-2746/$ - see front matter Ó 2019 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. https://doi.org/10.1016/j.jse.2019.05.022
2
J.H. Chan et al.
for induced hypotension while in the semi–beach chair position.11 Despite these advantages, it is well established in the literature that the upright beach chair position for shoulder arthroscopy or arthroplasty is associated with hypotension which can lead to cerebral hypoperfusion.4,7-9,16,17 Reduction in cerebral perfusion below a certain critical threshold may result in permanent neurologic injury.6,14 Cerebral desaturation events (CDEs), defined as a 20% reduction in frontal lobe oxygenation, have been associated with syncope in conscious patients.10 During shoulder arthroplasty, not only are patients always maintained in this position but very commonly the glenoid and/or humeral component require pressurized cementation during the surgical procedure. This has been shown to potentially cause hypotension, cardiac arrhythmias, increased pulmonary vascular resistance, and pulmonary emboli during implantation, potentially leading to hypoxemia and cardiac arrest.13,15 In addition, shoulder arthroplasty cases are associated with longer operative times and are more invasive than most arthroscopic shoulder surgeries that previous studies concentrated on, and thus potentially pose a greater risk of cerebral hypoperfusion. Near-infrared spectroscopy is a noninvasive system that has been shown to accurately monitor cerebral oxygenation and perfusion.2,3,5,12,14 The INVOs system is a US Food and Drug Administration–approved technology that has been used in patient populations undergoing procedures that have a high risk for an adverse neurologic outcome.1 Multiple small case series and case reports have described intraoperative cerebral desaturation episodes leading to rare catastrophic neurocognitive complications from shoulder surgery in a semi–beach chair position.1,3,14 Cerebral hypoperfusion has been well documented in the semi–beach chair position for shoulder arthroscopy. However, to our knowledge, there have been no previous studies evaluating cerebral perfusion during shoulder arthroplasty. The aim of this prospective observational study is to investigate the effect of total shoulder arthroplasty in the semi–beach chair position on cerebral O2 saturation at predetermined time points during the procedure. We will also investigate the incidence, timing, and magnitude of cerebral desaturation events (CDEs) during the procedure.
broaching, followed by preparation of the glenoid with reaming. Afterwards, the glenoid component was cemented into place, followed by press-fit placement of the humeral component. For reverse-total shoulder replacements, a similar protocol was followed, with the major difference being an addition of screw fixation into the glenoid component during impaction. Exclusion criteria included documented carotid stenosis (90% occlusion), prior neck surgery, cervical stenosis, cervical disc herniation, a history of stroke or transient ischemic attack, or sudden vision loss. We also excluded total shoulder arthroplasty performed for fracture care. Patient demographic data, including age, sex, and body mass index, were also recorded. A standardized anesthesia protocol was used in all patients. An intravenous line was inserted in the preoperative holding area. Patients were first given midazolam (2-6 mg intravenously) titrated to effect and then placed on 2 L of continuous oxygen through a facemask. An ultrasound-guided intrascalene block using 30 mL of 0.5% bupivacaine was then performed on the side ipsilateral to the operative upper extremity. The patients were transported to the operating suite. Compression devices were placed on bilateral lower extremities. The patients were transferred onto the operative table. Intraoperative monitoring included electrocardiography, arterial blood pressure via cuff placed on the nonoperative extremity, and pulse oximetry. Anesthesia was induced with 2.5-3.0 mg/kg propofol. The airway was secured and maintained using a laryngeal mask airway. After wiping the forehead with an alcohol swab, 2 noninvasive near-infrared spectroscopy sensors were placed bilaterally to the frontotemporal area with the medial margin at forehead midline and the lower margin above the eyebrow to avoid the temporalis muscle. The INVOs monitors were calibrated to match the time in anesthesia records. A preoperative baseline regional cerebral tissue oxygen saturation (RSO2) was obtained in the supine position. From this point, cerebral oxygenation was recorded continuously on the INVOs monitor. A combination of sevoflurane with nitrous oxide with 50% fraction inspired oxygen-maintained anesthesia. Semi–beach chair positioning at 30 of inclination was performed in a single step, and verified by digital inclinometer. This was followed by leg elevation. The surgeon proceeded with the total shoulder arthroplasty with no changes to his normal protocol. Specific time points were marked during the procedure in the anesthesia chart: semi–beach chair positioning, incision, humerus broaching, glenoid preparation, glenoid component impaction, and humeral component impaction. These time points were later matched to the calibrated time points in the continuous monitor recordings. Anesthesia was blinded to the RSO2 readings and was instructed to not make any changes based on the monitor output.
Methods
Statistical analysis
After institutional review board approval, informed consent was obtained from all subjects. Twenty-six consecutive patients scheduled to undergo elective total shoulder replacement in the semi–beach chair position were enrolled. All surgeries were performed at a single institution by 2 board-certified orthopedic surgeons from March 2016 to October 2016. All shoulder replacement types that required a press-fit humeral component were included. For anatomic total shoulder replacements, we would start with preparation of the humerus with humeral
The study was prospectively powered based on CDE as the primary outcome variable, and with 50 subjects, statistical testing via a chi-square analysis would yield a power of 99% (P < .001). The study was terminated early, and subsequently the goal for subjects was not achieved. Left and right hemisphere O2 measurements were averaged for analysis. Oxygen saturation data for each patient was measured in 5-second intervals. This allowed for plotting of continuous data and averaging among patients based on when the transition point occurred. Generalized least squares models,
Evaluation of cerebral perfusion during TSA Table I
3
Demographic data
Table II
Demographic Age Sex, male/female, n BMI Baseline O2
69.5 9.75 10/16 27.7 5.26 74 3
BMI, body mass index. Values are mean standard deviation unless otherwise noted.
Summary of cerebral desaturation event (CDE) data
Patients with CDE, n/N (%) Interval of first CDE Beach chair Incision Humeral broaching Glenoid reaming Glenoid impaction Humeral impaction Mean duration per event, min
19/25 (76) 8 6 1 1 1 2
(42) (32) (5) (5) (5) (11)
8.4
Values are n (%) unless otherwise noted.
with cubic splines for smoothing, were fit to the data. Akaike information criterion was used to identify the most suitable correlation structure for the model. Analyses including all positions used the average O2 value for each position for each patient. For transition models focusing on the transition from one position to another, all individual measurement times were used. Analyses were additionally performed to examine significant changes in the raw drop of RSO2 as well as to identify changes for drops as a percentage change from baseline. Analyses exclude a single patient with extreme and unusual data. All analyses were performed using the R package ‘‘rms’’ (version 3.4.0).
Results Twenty-six patients (10 males, 16 females) were observed during surgery with the mean age of 69.59.75 years and mean baseline oxygen saturation of 74% 3% (Table I). Of these patients, 1 was removed from the final analysis as a result of sensor error in recording the oxygen saturation level. The average operative time for these patients was 83 minutes (range 63-117 min). The collected cerebral oxygen saturation data was used to monitor patients for CDEs. Nineteen of the 25 observed patients experienced CDEs (Table II). Of those patients experiencing CDEs, the mean maximal decrease in oxygen saturation was 25% (range 20%-43%). During the interval in which patients were placed in the semi–beach chair position, 8 patients had their first CDE. This was the most common interval for CDE onset. The second most common interval of CDE onset was between incision and humeral broaching, where 6 patients suffered a CDE. There was no significant difference in age, sex, body mass index, or operative time in our incidence of CDEs (Table III). For each of our 25 subjects, the average of oxygen saturation levels at 7 predetermined time points (baseline, semi–beach chair, incision, humeral broaching, glenoid reaming, glenoid impaction, and humeral impaction) were obtained. The mean oxygen saturation was 77% (95% confidence interval [CI] 72-81) at baseline, which decreased to 59% (95% CI 55-64) by our final event (humeral impaction). Mean oxygen saturation levels steadily declined with each event until the fourth time point, at which point the levels stabilized (Fig. 1).
Next, we compared the mean change in oxygen saturation levels during the 6 transitions (defined as change in oxygen saturation from one event to the next). Only the first transition from baseline to semi–beach chair demonstrated a significant decrease (8%, 95% CI 2-13, P value ¼ .01) (Tables IV and V). The average oxygen saturation level during the baseline period was shown to be rising slightly before patients were shifted into the semi–beach chair position (Fig. 2). A nonlinear decrease in oxygen saturation is demonstrated with the onset of the semi–beach chair position. None of the remaining 5 transitions demonstrated a significant decrease in oxygen saturation levels (P ¼ .07, .18, .68, .90, .64). We also studied mean oxygen saturation in each interval as a percentage change from baseline measurements during transitions. The reason for pursuing an analysis regarding percentage change was based on the observation that raw mean oxygen saturation difference between patients provided large CIs. The trend for percentage change shows an overall decrease in oxygen saturation, with the steepest slope from the baseline to semi–beach chair transition (Fig. 3). When compared to baseline, there was a statistically significant decrease in mean oxygen saturation during the semi–beach chair interval (10%, 95% CI 6-13, P < .05) and the transition from semi–beach chair to incision (7%, 95% CI 2-13, P < .05) (Table VI). The most significant nonlinear change occurred with the transition from baseline to the semi–beach chair position (Fig. 4).
Table III
Summary of demographic data in relation to CDE
Number of patients Sex, male/female, n/N (%) BMI Age Operative time, min
Desaturation group
Nondesaturation group
19 5 (26.3%) / 14 (73.7%) 27.73 6.01 65.7 10.8 93 18
6 4 (66.7%) / 2 (33.3%) 27.75 2.78 69.8 6.9 104 10
CDE, cerebral desaturation event; BMI, body mass index. Values are mean standard deviation unless otherwise noted.
4
J.H. Chan et al.
Figure 1
Average cerebral oxygenation during arthroplasty time points.
There were no specific cerebral complications, such as weakness, nausea, or focal neurologic deficits, recorded in any of our 25 patients.
Discussion Multiple case reports and studies have shown that surgery performed in the semi–beach chair position may affect cerebral oximetry.4,7-9,16,17 Despite this, shoulder arthroplasty surgery is exclusively performed in the beach chair position because it allows for anatomic shoulder positioning and easier shoulder access, improved airway access, reduced bleeding risk, and reduced brachial plexus injury risk.18 In addition, shoulder surgeons will often ask for induced hypotension for hemostasis control and better visualization while in the beach chair position.11 In shoulder arthroplasty, the glenoid and/or humeral component require pressurized cementation during the surgical
Table IV
procedure, which may theoretically further increase their risk for a CDE in the beach chair position. Thus, the aim of this prospective observational study was to not only investigate the incidence of CDEs but also measure oxygen saturation changes at specific time points in a shoulder arthroplasty procedure. Nineteen of our 25 patients (76%) experienced CDEs in our study, with a mean maximal decrease of 25% from baseline. Previous studies measured CDEs in all patients receiving shoulder surgery in the beach chair position and showed that 18% of those patients had CDEs.17 The increase in CDEs in our patient population may be due to the fact that our anesthesia team was blinded to RSO2 results from the monitors, whereas in previous studies they were not blinded and were instructed to intervene when a CDE occurred. Our mean maximal decrease is similar to previous studies.
Table V interval
Time point legend
Time point no.
Time point
Time point abbreviation
0 1 2 3 4 5 6
Baseline Semi–beach chair positioning Incision Humeral broaching Glenoid reaming Glenoid impaction Humeral impaction
BL BC In HB GR GI HI
Average change in cerebral oxygenation during each
Events compared
Value
95% CI
P value
BL-BC* BC-In* In-HB* HB-GR* GR-GI* GI-HI*
8 5 4 1 0 –1
2, 13 0, 11 –2, 10 –5, 7 –6, 6 –7, 5
.01 .07 .18 .68 .90 .64
CI, confidence interval. Positive change values represent a decrease in oxygen saturation.
Evaluation of cerebral perfusion during TSA
Figure 2
5
Average change in cerebral oxygenation during semi–beach chair positioning. CI, confidence interval.
We saw stabilization of cerebral oxygenation after humeral broaching. Although there were no statistically significant changes to cerebral oxygenation after beach chair positioning, it is worth pointing out that there were changes to cerebral oxygenation throughout the duration of implant placement. Our study used several different methods to study change in cerebral oximetry over time during the shoulder
Figure 3
arthroplasty procedure. We hypothesized that parts of the shoulder arthroplasty procedure, along with cement pressurization, would cause decreases in cerebral oximetry over time. In every method, the largest overall and relative decrease was seen during the transition from baseline to semi–beach chair position. This is consistent with previous literature that showed most CDEs occurred an average of 8 minutes 9 seconds after transition to the semi–beach chair
Percentage change in cerebral oxygenation during arthroplasty time points.
6
J.H. Chan et al. Table VI Percentage change in cerebral oxygenation during each interval Events compared
% O2 change
95% CI
P value
BL-BC BC-In In-HB HB-GR GR-GI GI-HI
10 7 5 2 1 –2
6, 13 2, 13 0, 10 –4, 7 –4, 7 –8, 3
<.001 .01 .07 .55 .64 .43
CI, confidence interval. Positive change values represent a decrease in oxygen saturation.
position.17 We can only speculate that this may be due to the lack of a sympathetic nervous response from a general anesthetic during supine to upright positioning. This could be an interesting future follow-up study. The main limitation to our study is the sample size. Our results are underpowered for small cerebral oxygen saturation changes seen during the shoulder arthroplasty procedure. Our results are also underpowered for significant differences in incidence of CDEs in different age, sex, or body mass index. In addition, our study was a prospective observational study; thus, we do not have a control group with patients undergoing shoulder arthroplasty without cerebral oximetry recordings. We would have liked to limit our study to a single total shoulder design; however, we do not think the differences in the prosthesis design would make a large difference in our cerebral oximetry data.
Figure 4
Finally, our study is very specific in its scope and is by design not generalizable. It is worth mentioning that despite several case reports with catastrophic neurologic deficits after CDEs, this did not occur in any of our patients. The clinical relevance of intraoperative CDEs thus is still not well understood, and the incidence of CDEs did not correlate with clinical outcomes in our study.
Conclusion The largest decline in cerebral oxygen saturation during shoulder arthroplasty performed in the beach chair position is during semi–beach chair positioning itself. Shoulder arthroplasty, cement, and implant impaction during the procedure do not cause any additional risk for cerebral hypoperfusion. Noninvasive blood pressure monitoring does not guarantee normal cerebral perfusion, and in fact a majority (76%) of our patients experienced an intraoperative CDE. Though this did not correlate with any clinical findings in our study, it is clear that oxygen saturation is compromised in the semi–beach chair position, and additional monitoring such as our near-infrared spectroscopy system allows for prompt identification of cerebral hypoperfusion in cases where the beach chair position is necessary. In cases such as shoulder arthroplasty where the semi–beach chair position is used exclusively, protocols aimed at detecting and reversing CDE especially at the time of positioning may improve patient safety.
Percentage change in cerebral oxygenation during semi–beach chair positioning. CI, confidence interval.
Evaluation of cerebral perfusion during TSA
7
Disclaimer 8.
Guido Marra reports that he has received royalties and consultant fees from Zimmer; is on the editorial/governing board for Clinical Orthopaedics & Related Research; and is a board member/committee appointments for the American Academy of Orthopaedic Surgeons, American Shoulder and Elbow Surgeons, and the Association of Bone & Joint Surgeons. Matthew Saltzman reports that he has received royalties from Wright Medical Technology and consultant fees from Medacta and Wright Medical Technology. The other authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
9.
10. 11.
12.
13.
References 1. Bhatti MT, Enneking FK. Visual loss and ophthalmoplegia after shoulder surgery. Anesth Analg 2003;96:899-902. https://doi.org/10. 1213/01.ANE.0000047272.31849.F9 2. Carlin RE, McGraw DJ, Calimlim JR, Mascia MF. The use of nearinfrared cerebral oximetry in awake carotid endarterectomy. J Clin Anesth 1998;10:109-13. 3. Casati A, Spreafico E, Putzu M, Fanelli G. New technology for noninvasive brain monitoring: continuous cerebral oximetry. Minerva Anestesiol 2006;72:605-25 [in English, Italian]. 4. Dippmann C, Winge S, Nielsen HB. Severe cerebral desaturation during shoulder arthroscopy in the beach-chair position. Arthroscopy 2010;26(9 Suppl):S148-50. https://doi.org/10.1016/j.arthro.2010.03.012 5. 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 2009;9:304-7. https://doi.org/10.1111/j.1533-2500.2009.00282.x 6. Friedman DJ, Parnes NZ, Zimmer Z, Higgins LD, Warner JJ. Prevalence of cerebrovascular events during shoulder surgery and association with patient position. Orthopedics 2009;32. https://doi.org/10. 1213/ane.0b013e3181a9d8b5 7. Gillespie R, Shishani Y, Streit J, Wanner JP, McCrum C, Syed T, et al. The safety of controlled hypotension for shoulder arthroscopy in the
14.
15.
16.
17.
18.
19.
beach-chair position. J Bone Joint Surg Am 2012;94:1284-90. https:// doi.org/10.2106/JBJS.J.01550 Jeong H, Lee SH, Jang EA, Chung SS, Lee J, Yoo KY. Haemodynamics and cerebral oxygenation during arthroscopic shoulder surgery in beach chair position under general anaesthesia. Acta Anaesthesiol Scand 2012;56:872-9. https://doi.org/10.1111/j.1399-6576.2012. 02716.x Lee JH, Min KT, Chun YM, Kim EJ, Choi SH. Effects of beach-chair position and induced hypotension on cerebral oxygen saturation in patients undergoing arthroscopic shoulder surgery. Arthroscopy 2011; 27:889-94. https://doi.org/10.1016/j.arthro.2011.02.027 Madsen PL, Secher NH. Near-infrared oximetry of the brain. Prog Neurobiol 1999;58:541-60. Morrison DS, Schaefer RK, Friedman RL. The relationship between subacromial space pressure, blood pressure, and visual clarity during arthroscopic subacromial decompression. Arthroscopy 1995;11: 557-60. Murphy GS, Szokol JW, Marymont JH, Greenberg SB, Avram MJ, Vender JS, et al. Cerebral oxygen desaturation events assessed by near-infrared spectroscopy during shoulder arthroscopy in the beach chair and lateral decubitus positions. Anesth Analg 2010;111:496-505. https://doi.org/10.1213/ANE.0b013e3181e33bd9 Orsini EC, Byrick RJ, Mullen JB, Kay JC, Waddell JP. Cardiopulmonary function and pulmonary microemboli during arthroplasty using cemented or non-cemented components. The role of intramedullary pressure. J Bone Joint Surg Am 1987;69:822-32. Papadonikolakis A, Wiesler ER, Olympio MA, Poehling GG. Avoiding catastrophic complications of stroke and death related to shoulder surgery in the sitting position. Arthroscopy 2008;24:481-2. https://doi. org/10.1016/j.arthro.2008.02.005 Parvizi J, Holiday AD, Ereth MH, Lewallen DG. The Frank Stinchfield Award. Sudden death during primary hip arthroplasty. Clin Orthop Relat Res 1999;369:39-48. Salazar D, Sears B, Acosta A, Aghdasi B, Francois A, Tonino P, et al. Effect of head and neck positioning on cerebral perfusion during shoulder arthroscopy in beach chair position. J Surg Orthop Adv 2014; 23:83-9. Salazar D, Sears BW, Andre J, Tonino P, Marra G. Cerebral desaturation during shoulder arthroscopy: a prospective observational study. Clin Orthop Relat Res 2013;471:4027-34. https://doi.org/10.1007/ s11999-013-2987-6 Skyhar MJ, Altchek DW, Warren RF, Wickiewicz TL, O’Brien SJ. Shoulder arthroscopy with the patient in the beach-chair position. Arthroscopy 1988;4:256-9. Tange K, Kinoshita H, Minonishi T, Hatakeyama N, Matsuda N, Yamazaki M, et al. Cerebral oxygenation in the beach chair position before and during general anesthesia. Minerva Anestesiol 2010;76: 485-90.
www.elsevier.com/locate/ymse
Evaluation of cerebral oxygen perfusion during shoulder arthroplasty performed in the semi–beach chair position Justin H. Chan, MD*, Hector Perez, BA, Harrison Lee, BS, Matthew Saltzman, MD, Guido Marra, MD Department of Orthopaedic Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL, USA Background: The beach chair position is commonly used when performing shoulder arthroplasty. However, this position has been associated with hypotension, potentially leading to cerebral hypoperfusion, which may cause neurologic injury. In addition, shoulder arthroplasty cases are associated with longer operative times, posing a potentially greater risk of cerebral hypoperfusion. We aim to evaluate the risk of cerebral desaturation events (CDEs) during the course of total shoulder arthroplasty. Methods: Twenty-six patients undergoing shoulder arthroplasties were monitored for changes in cerebral perfusion. Seven specific time-points during the procedure were labeled for comparison of events: baseline, beach chair, incision, humeral broaching, glenoid reaming, glenoid component implantation, and humeral component implantation. Cerebral oxygen perfusion was measured using near-infrared spectroscopy. A CDE was described as a decrease of oxygen saturation greater than 20%. Results: Nineteeen of 25 subjects experienced a CDE. 42% of these patients experienced CDEs during semi–beach chair positioning. Patients experienced the largest oxygen saturation drop during semi–beach chair positioning. Transition from baseline to semi–beach chair was the only event to have a statistically significant decrease in cerebral perfusion (8%, P < .05). There was a statistically significant percentage change in mean oxygen saturation in the semi–beach chair interval (10%, P < .01) and the semi–beach chair to incision interval (7%, P < .01). Conclusions: Most patients experienced an intraoperative CDE, with greatest incidence during semi–beach chair positioning. The largest decline in cerebral oxygen saturation occurred during semi–beach chair positioning. Implant implantation was not associated with decrease in cerebral oximetry. Level of evidence: Level IV; Case Series; Treatment Study Ó 2019 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. Keywords: Total shoulder arthroplasty; beach chair position; semi–beach chair position; cerebral oxygenation; near-infrared spectroscopy
IRB approval by Northwestern University Institutional Review Board, approval number STU200834-MOD0003. *Reprint requests: Justin H. Chan, MD, Department of Orthopaedic Surgery, Northwestern University Feinberg School of Medicine, 676 N St Clair, Suite 1350, Chicago, IL 60611, USA. E-mail address: [email protected] (J.H. Chan).
During shoulder arthroplasty, the semi–beach chair position is exclusively used because it allows for anatomic shoulder positioning and easier shoulder access.19 The semi–beach chair position also allows for improved airway access, reduced bleeding risk and reduced brachial plexus injury risk.18 In addition, shoulder surgeons will often ask
1058-2746/$ - see front matter Ó 2019 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. https://doi.org/10.1016/j.jse.2019.05.022
2
J.H. Chan et al.
for induced hypotension while in the semi–beach chair position.11 Despite these advantages, it is well established in the literature that the upright beach chair position for shoulder arthroscopy or arthroplasty is associated with hypotension which can lead to cerebral hypoperfusion.4,7-9,16,17 Reduction in cerebral perfusion below a certain critical threshold may result in permanent neurologic injury.6,14 Cerebral desaturation events (CDEs), defined as a 20% reduction in frontal lobe oxygenation, have been associated with syncope in conscious patients.10 During shoulder arthroplasty, not only are patients always maintained in this position but very commonly the glenoid and/or humeral component require pressurized cementation during the surgical procedure. This has been shown to potentially cause hypotension, cardiac arrhythmias, increased pulmonary vascular resistance, and pulmonary emboli during implantation, potentially leading to hypoxemia and cardiac arrest.13,15 In addition, shoulder arthroplasty cases are associated with longer operative times and are more invasive than most arthroscopic shoulder surgeries that previous studies concentrated on, and thus potentially pose a greater risk of cerebral hypoperfusion. Near-infrared spectroscopy is a noninvasive system that has been shown to accurately monitor cerebral oxygenation and perfusion.2,3,5,12,14 The INVOs system is a US Food and Drug Administration–approved technology that has been used in patient populations undergoing procedures that have a high risk for an adverse neurologic outcome.1 Multiple small case series and case reports have described intraoperative cerebral desaturation episodes leading to rare catastrophic neurocognitive complications from shoulder surgery in a semi–beach chair position.1,3,14 Cerebral hypoperfusion has been well documented in the semi–beach chair position for shoulder arthroscopy. However, to our knowledge, there have been no previous studies evaluating cerebral perfusion during shoulder arthroplasty. The aim of this prospective observational study is to investigate the effect of total shoulder arthroplasty in the semi–beach chair position on cerebral O2 saturation at predetermined time points during the procedure. We will also investigate the incidence, timing, and magnitude of cerebral desaturation events (CDEs) during the procedure.
broaching, followed by preparation of the glenoid with reaming. Afterwards, the glenoid component was cemented into place, followed by press-fit placement of the humeral component. For reverse-total shoulder replacements, a similar protocol was followed, with the major difference being an addition of screw fixation into the glenoid component during impaction. Exclusion criteria included documented carotid stenosis (90% occlusion), prior neck surgery, cervical stenosis, cervical disc herniation, a history of stroke or transient ischemic attack, or sudden vision loss. We also excluded total shoulder arthroplasty performed for fracture care. Patient demographic data, including age, sex, and body mass index, were also recorded. A standardized anesthesia protocol was used in all patients. An intravenous line was inserted in the preoperative holding area. Patients were first given midazolam (2-6 mg intravenously) titrated to effect and then placed on 2 L of continuous oxygen through a facemask. An ultrasound-guided intrascalene block using 30 mL of 0.5% bupivacaine was then performed on the side ipsilateral to the operative upper extremity. The patients were transported to the operating suite. Compression devices were placed on bilateral lower extremities. The patients were transferred onto the operative table. Intraoperative monitoring included electrocardiography, arterial blood pressure via cuff placed on the nonoperative extremity, and pulse oximetry. Anesthesia was induced with 2.5-3.0 mg/kg propofol. The airway was secured and maintained using a laryngeal mask airway. After wiping the forehead with an alcohol swab, 2 noninvasive near-infrared spectroscopy sensors were placed bilaterally to the frontotemporal area with the medial margin at forehead midline and the lower margin above the eyebrow to avoid the temporalis muscle. The INVOs monitors were calibrated to match the time in anesthesia records. A preoperative baseline regional cerebral tissue oxygen saturation (RSO2) was obtained in the supine position. From this point, cerebral oxygenation was recorded continuously on the INVOs monitor. A combination of sevoflurane with nitrous oxide with 50% fraction inspired oxygen-maintained anesthesia. Semi–beach chair positioning at 30 of inclination was performed in a single step, and verified by digital inclinometer. This was followed by leg elevation. The surgeon proceeded with the total shoulder arthroplasty with no changes to his normal protocol. Specific time points were marked during the procedure in the anesthesia chart: semi–beach chair positioning, incision, humerus broaching, glenoid preparation, glenoid component impaction, and humeral component impaction. These time points were later matched to the calibrated time points in the continuous monitor recordings. Anesthesia was blinded to the RSO2 readings and was instructed to not make any changes based on the monitor output.
Methods
Statistical analysis
After institutional review board approval, informed consent was obtained from all subjects. Twenty-six consecutive patients scheduled to undergo elective total shoulder replacement in the semi–beach chair position were enrolled. All surgeries were performed at a single institution by 2 board-certified orthopedic surgeons from March 2016 to October 2016. All shoulder replacement types that required a press-fit humeral component were included. For anatomic total shoulder replacements, we would start with preparation of the humerus with humeral
The study was prospectively powered based on CDE as the primary outcome variable, and with 50 subjects, statistical testing via a chi-square analysis would yield a power of 99% (P < .001). The study was terminated early, and subsequently the goal for subjects was not achieved. Left and right hemisphere O2 measurements were averaged for analysis. Oxygen saturation data for each patient was measured in 5-second intervals. This allowed for plotting of continuous data and averaging among patients based on when the transition point occurred. Generalized least squares models,
Evaluation of cerebral perfusion during TSA Table I
3
Demographic data
Table II
Demographic Age Sex, male/female, n BMI Baseline O2
69.5 9.75 10/16 27.7 5.26 74 3
BMI, body mass index. Values are mean standard deviation unless otherwise noted.
Summary of cerebral desaturation event (CDE) data
Patients with CDE, n/N (%) Interval of first CDE Beach chair Incision Humeral broaching Glenoid reaming Glenoid impaction Humeral impaction Mean duration per event, min
19/25 (76) 8 6 1 1 1 2
(42) (32) (5) (5) (5) (11)
8.4
Values are n (%) unless otherwise noted.
with cubic splines for smoothing, were fit to the data. Akaike information criterion was used to identify the most suitable correlation structure for the model. Analyses including all positions used the average O2 value for each position for each patient. For transition models focusing on the transition from one position to another, all individual measurement times were used. Analyses were additionally performed to examine significant changes in the raw drop of RSO2 as well as to identify changes for drops as a percentage change from baseline. Analyses exclude a single patient with extreme and unusual data. All analyses were performed using the R package ‘‘rms’’ (version 3.4.0).
Results Twenty-six patients (10 males, 16 females) were observed during surgery with the mean age of 69.59.75 years and mean baseline oxygen saturation of 74% 3% (Table I). Of these patients, 1 was removed from the final analysis as a result of sensor error in recording the oxygen saturation level. The average operative time for these patients was 83 minutes (range 63-117 min). The collected cerebral oxygen saturation data was used to monitor patients for CDEs. Nineteen of the 25 observed patients experienced CDEs (Table II). Of those patients experiencing CDEs, the mean maximal decrease in oxygen saturation was 25% (range 20%-43%). During the interval in which patients were placed in the semi–beach chair position, 8 patients had their first CDE. This was the most common interval for CDE onset. The second most common interval of CDE onset was between incision and humeral broaching, where 6 patients suffered a CDE. There was no significant difference in age, sex, body mass index, or operative time in our incidence of CDEs (Table III). For each of our 25 subjects, the average of oxygen saturation levels at 7 predetermined time points (baseline, semi–beach chair, incision, humeral broaching, glenoid reaming, glenoid impaction, and humeral impaction) were obtained. The mean oxygen saturation was 77% (95% confidence interval [CI] 72-81) at baseline, which decreased to 59% (95% CI 55-64) by our final event (humeral impaction). Mean oxygen saturation levels steadily declined with each event until the fourth time point, at which point the levels stabilized (Fig. 1).
Next, we compared the mean change in oxygen saturation levels during the 6 transitions (defined as change in oxygen saturation from one event to the next). Only the first transition from baseline to semi–beach chair demonstrated a significant decrease (8%, 95% CI 2-13, P value ¼ .01) (Tables IV and V). The average oxygen saturation level during the baseline period was shown to be rising slightly before patients were shifted into the semi–beach chair position (Fig. 2). A nonlinear decrease in oxygen saturation is demonstrated with the onset of the semi–beach chair position. None of the remaining 5 transitions demonstrated a significant decrease in oxygen saturation levels (P ¼ .07, .18, .68, .90, .64). We also studied mean oxygen saturation in each interval as a percentage change from baseline measurements during transitions. The reason for pursuing an analysis regarding percentage change was based on the observation that raw mean oxygen saturation difference between patients provided large CIs. The trend for percentage change shows an overall decrease in oxygen saturation, with the steepest slope from the baseline to semi–beach chair transition (Fig. 3). When compared to baseline, there was a statistically significant decrease in mean oxygen saturation during the semi–beach chair interval (10%, 95% CI 6-13, P < .05) and the transition from semi–beach chair to incision (7%, 95% CI 2-13, P < .05) (Table VI). The most significant nonlinear change occurred with the transition from baseline to the semi–beach chair position (Fig. 4).
Table III
Summary of demographic data in relation to CDE
Number of patients Sex, male/female, n/N (%) BMI Age Operative time, min
Desaturation group
Nondesaturation group
19 5 (26.3%) / 14 (73.7%) 27.73 6.01 65.7 10.8 93 18
6 4 (66.7%) / 2 (33.3%) 27.75 2.78 69.8 6.9 104 10
CDE, cerebral desaturation event; BMI, body mass index. Values are mean standard deviation unless otherwise noted.
4
J.H. Chan et al.
Figure 1
Average cerebral oxygenation during arthroplasty time points.
There were no specific cerebral complications, such as weakness, nausea, or focal neurologic deficits, recorded in any of our 25 patients.
Discussion Multiple case reports and studies have shown that surgery performed in the semi–beach chair position may affect cerebral oximetry.4,7-9,16,17 Despite this, shoulder arthroplasty surgery is exclusively performed in the beach chair position because it allows for anatomic shoulder positioning and easier shoulder access, improved airway access, reduced bleeding risk, and reduced brachial plexus injury risk.18 In addition, shoulder surgeons will often ask for induced hypotension for hemostasis control and better visualization while in the beach chair position.11 In shoulder arthroplasty, the glenoid and/or humeral component require pressurized cementation during the surgical
Table IV
procedure, which may theoretically further increase their risk for a CDE in the beach chair position. Thus, the aim of this prospective observational study was to not only investigate the incidence of CDEs but also measure oxygen saturation changes at specific time points in a shoulder arthroplasty procedure. Nineteen of our 25 patients (76%) experienced CDEs in our study, with a mean maximal decrease of 25% from baseline. Previous studies measured CDEs in all patients receiving shoulder surgery in the beach chair position and showed that 18% of those patients had CDEs.17 The increase in CDEs in our patient population may be due to the fact that our anesthesia team was blinded to RSO2 results from the monitors, whereas in previous studies they were not blinded and were instructed to intervene when a CDE occurred. Our mean maximal decrease is similar to previous studies.
Table V interval
Time point legend
Time point no.
Time point
Time point abbreviation
0 1 2 3 4 5 6
Baseline Semi–beach chair positioning Incision Humeral broaching Glenoid reaming Glenoid impaction Humeral impaction
BL BC In HB GR GI HI
Average change in cerebral oxygenation during each
Events compared
Value
95% CI
P value
BL-BC* BC-In* In-HB* HB-GR* GR-GI* GI-HI*
8 5 4 1 0 –1
2, 13 0, 11 –2, 10 –5, 7 –6, 6 –7, 5
.01 .07 .18 .68 .90 .64
CI, confidence interval. Positive change values represent a decrease in oxygen saturation.
Evaluation of cerebral perfusion during TSA
Figure 2
5
Average change in cerebral oxygenation during semi–beach chair positioning. CI, confidence interval.
We saw stabilization of cerebral oxygenation after humeral broaching. Although there were no statistically significant changes to cerebral oxygenation after beach chair positioning, it is worth pointing out that there were changes to cerebral oxygenation throughout the duration of implant placement. Our study used several different methods to study change in cerebral oximetry over time during the shoulder
Figure 3
arthroplasty procedure. We hypothesized that parts of the shoulder arthroplasty procedure, along with cement pressurization, would cause decreases in cerebral oximetry over time. In every method, the largest overall and relative decrease was seen during the transition from baseline to semi–beach chair position. This is consistent with previous literature that showed most CDEs occurred an average of 8 minutes 9 seconds after transition to the semi–beach chair
Percentage change in cerebral oxygenation during arthroplasty time points.
6
J.H. Chan et al. Table VI Percentage change in cerebral oxygenation during each interval Events compared
% O2 change
95% CI
P value
BL-BC BC-In In-HB HB-GR GR-GI GI-HI
10 7 5 2 1 –2
6, 13 2, 13 0, 10 –4, 7 –4, 7 –8, 3
<.001 .01 .07 .55 .64 .43
CI, confidence interval. Positive change values represent a decrease in oxygen saturation.
position.17 We can only speculate that this may be due to the lack of a sympathetic nervous response from a general anesthetic during supine to upright positioning. This could be an interesting future follow-up study. The main limitation to our study is the sample size. Our results are underpowered for small cerebral oxygen saturation changes seen during the shoulder arthroplasty procedure. Our results are also underpowered for significant differences in incidence of CDEs in different age, sex, or body mass index. In addition, our study was a prospective observational study; thus, we do not have a control group with patients undergoing shoulder arthroplasty without cerebral oximetry recordings. We would have liked to limit our study to a single total shoulder design; however, we do not think the differences in the prosthesis design would make a large difference in our cerebral oximetry data.
Figure 4
Finally, our study is very specific in its scope and is by design not generalizable. It is worth mentioning that despite several case reports with catastrophic neurologic deficits after CDEs, this did not occur in any of our patients. The clinical relevance of intraoperative CDEs thus is still not well understood, and the incidence of CDEs did not correlate with clinical outcomes in our study.
Conclusion The largest decline in cerebral oxygen saturation during shoulder arthroplasty performed in the beach chair position is during semi–beach chair positioning itself. Shoulder arthroplasty, cement, and implant impaction during the procedure do not cause any additional risk for cerebral hypoperfusion. Noninvasive blood pressure monitoring does not guarantee normal cerebral perfusion, and in fact a majority (76%) of our patients experienced an intraoperative CDE. Though this did not correlate with any clinical findings in our study, it is clear that oxygen saturation is compromised in the semi–beach chair position, and additional monitoring such as our near-infrared spectroscopy system allows for prompt identification of cerebral hypoperfusion in cases where the beach chair position is necessary. In cases such as shoulder arthroplasty where the semi–beach chair position is used exclusively, protocols aimed at detecting and reversing CDE especially at the time of positioning may improve patient safety.
Percentage change in cerebral oxygenation during semi–beach chair positioning. CI, confidence interval.
Evaluation of cerebral perfusion during TSA
7
Disclaimer 8.
Guido Marra reports that he has received royalties and consultant fees from Zimmer; is on the editorial/governing board for Clinical Orthopaedics & Related Research; and is a board member/committee appointments for the American Academy of Orthopaedic Surgeons, American Shoulder and Elbow Surgeons, and the Association of Bone & Joint Surgeons. Matthew Saltzman reports that he has received royalties from Wright Medical Technology and consultant fees from Medacta and Wright Medical Technology. The other authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
9.
10. 11.
12.
13.
References 1. Bhatti MT, Enneking FK. Visual loss and ophthalmoplegia after shoulder surgery. Anesth Analg 2003;96:899-902. https://doi.org/10. 1213/01.ANE.0000047272.31849.F9 2. Carlin RE, McGraw DJ, Calimlim JR, Mascia MF. The use of nearinfrared cerebral oximetry in awake carotid endarterectomy. J Clin Anesth 1998;10:109-13. 3. Casati A, Spreafico E, Putzu M, Fanelli G. New technology for noninvasive brain monitoring: continuous cerebral oximetry. Minerva Anestesiol 2006;72:605-25 [in English, Italian]. 4. Dippmann C, Winge S, Nielsen HB. Severe cerebral desaturation during shoulder arthroscopy in the beach-chair position. Arthroscopy 2010;26(9 Suppl):S148-50. https://doi.org/10.1016/j.arthro.2010.03.012 5. 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 2009;9:304-7. https://doi.org/10.1111/j.1533-2500.2009.00282.x 6. Friedman DJ, Parnes NZ, Zimmer Z, Higgins LD, Warner JJ. Prevalence of cerebrovascular events during shoulder surgery and association with patient position. Orthopedics 2009;32. https://doi.org/10. 1213/ane.0b013e3181a9d8b5 7. Gillespie R, Shishani Y, Streit J, Wanner JP, McCrum C, Syed T, et al. The safety of controlled hypotension for shoulder arthroscopy in the
14.
15.
16.
17.
18.
19.
beach-chair position. J Bone Joint Surg Am 2012;94:1284-90. https:// doi.org/10.2106/JBJS.J.01550 Jeong H, Lee SH, Jang EA, Chung SS, Lee J, Yoo KY. Haemodynamics and cerebral oxygenation during arthroscopic shoulder surgery in beach chair position under general anaesthesia. Acta Anaesthesiol Scand 2012;56:872-9. https://doi.org/10.1111/j.1399-6576.2012. 02716.x Lee JH, Min KT, Chun YM, Kim EJ, Choi SH. Effects of beach-chair position and induced hypotension on cerebral oxygen saturation in patients undergoing arthroscopic shoulder surgery. Arthroscopy 2011; 27:889-94. https://doi.org/10.1016/j.arthro.2011.02.027 Madsen PL, Secher NH. Near-infrared oximetry of the brain. Prog Neurobiol 1999;58:541-60. Morrison DS, Schaefer RK, Friedman RL. The relationship between subacromial space pressure, blood pressure, and visual clarity during arthroscopic subacromial decompression. Arthroscopy 1995;11: 557-60. Murphy GS, Szokol JW, Marymont JH, Greenberg SB, Avram MJ, Vender JS, et al. Cerebral oxygen desaturation events assessed by near-infrared spectroscopy during shoulder arthroscopy in the beach chair and lateral decubitus positions. Anesth Analg 2010;111:496-505. https://doi.org/10.1213/ANE.0b013e3181e33bd9 Orsini EC, Byrick RJ, Mullen JB, Kay JC, Waddell JP. Cardiopulmonary function and pulmonary microemboli during arthroplasty using cemented or non-cemented components. The role of intramedullary pressure. J Bone Joint Surg Am 1987;69:822-32. Papadonikolakis A, Wiesler ER, Olympio MA, Poehling GG. Avoiding catastrophic complications of stroke and death related to shoulder surgery in the sitting position. Arthroscopy 2008;24:481-2. https://doi. org/10.1016/j.arthro.2008.02.005 Parvizi J, Holiday AD, Ereth MH, Lewallen DG. The Frank Stinchfield Award. Sudden death during primary hip arthroplasty. Clin Orthop Relat Res 1999;369:39-48. Salazar D, Sears B, Acosta A, Aghdasi B, Francois A, Tonino P, et al. Effect of head and neck positioning on cerebral perfusion during shoulder arthroscopy in beach chair position. J Surg Orthop Adv 2014; 23:83-9. Salazar D, Sears BW, Andre J, Tonino P, Marra G. Cerebral desaturation during shoulder arthroscopy: a prospective observational study. Clin Orthop Relat Res 2013;471:4027-34. https://doi.org/10.1007/ s11999-013-2987-6 Skyhar MJ, Altchek DW, Warren RF, Wickiewicz TL, O’Brien SJ. Shoulder arthroscopy with the patient in the beach-chair position. Arthroscopy 1988;4:256-9. Tange K, Kinoshita H, Minonishi T, Hatakeyama N, Matsuda N, Yamazaki M, et al. Cerebral oxygenation in the beach chair position before and during general anesthesia. Minerva Anestesiol 2010;76: 485-90.