Perioperative Care of Patients at High Risk for Stroke during or after Non-Cardiac,Non-Neurologic Surgery:Consensus Statement from the Society for Neuroscience
in Anesthesiology and Critical Care *
George A.Mashour,MD,PhD,*Laurel E.Moore,MD,*Abhijit V.Lele,MD,w
Steven A.Robicsek,MD,PhD,z and Adrian W.Gelb,MBChB y
Abstract:This document is supported by the American Society of Anes-thesiologists.**
Perioperative stroke can be a catastrophic outcome for surgical patients and is associated with increased morbidity and mortality.This consensus statement from the Society for Neu-roscience in Anesthesiology and Critical Care provides evidence-based recommendations and opinions regarding the pre-operative,intraoperative,and postoperative care of patients at high risk for the complication.
Key Words:stroke,ischemic stroke,perioperative stroke,cere-brovascular,cerebrovascular accident,noncardiac surgery,neurologic complication
(J Neurosurg Anesthesiol 2014;26:273–285)
S
troke can be a catastrophic outcome for patients un-dergoing noncardiac,nonneurologic surgery and is associated with an adjusted 8-fold increase in mortality.1Unlike stroke in the community setting,the mechanistic cascade leading to perioperative stroke has a discrete and
highly-predictable origin:surgical intervention.Given the fact that surgery and anesthesia are associated with an increased risk of stroke compared to nonsurgical con-trols,2establishing perioperative recommendations to minimize risk could be impactful.Stroke after non-cardiac,nonneurologic surgery is relatively understudied and there is a need for clarifying the clinical management of surgical patients at high risk for the complication.
METHODOLOGY
Definition of Perioperative Stroke
Sacco et al have developed a consensus statement regarding the broad definition of stroke.3However,for the purposes of this consensus statement,“perioperative stroke”is defined as a brain infarction of ischemic or hemorrhagic etiology that occurs during surgery or within 30days after s
urgery.We recommend that such a standardized definition be adopted for future reports.It is important to note that this clinical situation is distinct from that of a patient presenting for acute therapy after a stroke has occurred in a nonoperative setting.Peri-procedural care of patients presenting for endovascular interventions to treat stroke is described elsewhere.4
Purpose of the Consensus Statement
The purpose of this consensus statement is to pro-vide evidence-based recommendations regarding (1)pre-operative identification of patients at high risk of stroke during or after noncardiac,noncarotid,nonneurologic surgery,(2)preoperative considerations to mitigate risk,(3)intraoperative management to mitigate risk,and (4)appropriate steps for clinical care if stroke is identified in the postoperative period.
Focus
Patients undergoing carotid endarterectomy and a variety of cardiac surgeries are known to be at high risk for perioperative stroke,with fairly clear etiologies (e.g.,em-bolic event).5As such,there has been considerable attention to the prevention of stroke in these populations.The focus of the current consensus statement is the prevention and management of ischemic stroke in adult patients under-goi
ng noncardiac,noncarotid,and nonneurologic surgery.
*Society Consensus Statements published in the Journal of Neuro-surgical Anesthesiology have been reviewed by the JNA Affiliate So-cieties that submit them for publication.They have not undergone review by the Editorial Board of the Journal of Neurosurgical Anesthesiology.
From the *Departments of Anesthesiology and Neurosurgery,Uni-versity of Michigan,Ann Arbor,MI;w Departments of Anesthesi-ology,Neurology,and Neurosurgery,University of Kansas,Kansas City,KS;z Department of Anesthesiology,University of Florida,Gainesville,FL;and y Department of Anesthesia and Perioperative Care,University of California,San Francisco,CA.
**This document has not been approved by ASA’s House of Delegates
or Board of Directors and does not represent an official or approved statement or policy of ASA.
Disclosures:The authors have no conflicts of interest to declare.Funding:Departmental and institutional sources.
Reprints:George A.Mashour,MD,PhD,Division of Neuro-anesthesiology,Department of Anesthesiology,University of Michigan Medical School,1500East Medical Center Drive,1H247UH/SPC-
5048,Ann Arbor,MI 48109-5048(e-mail:gmashour@umich.edu).
Copyright r 2014by Lippincott Williams &Wilkins
S PECIAL A RTICLE
J Neurosurg Anesthesiol
Volume 26,Number 4,October 2014
www.jnsa |
273
Application
This consensus statement is intended for use by anes-thesiologists,anesthesia providers,surgeons and other peri-operative care providers.It may also serve health care professionals such as internists or neurologists who evaluate patients in the perioperative period.For the purposes of this article,anesthesia care refers to general anesthesia,regional anesthesia,or monitored anesthesia care.
Task Force Members and Consultants
The initial recommendations were developed byfive anesthesiologists with expertise in clinical neuroscience and neuroanesthesiology,practicing in academic departments across the U.S.These individuals were chosen from mem-bership of the Society for Neuroscience in Anesthesiology and Critical Care(SNACC),an international organization. Applicants were required to have published peer-reviewed research on the subject of perioperative stroke or have documented experience in the care of patients with stroke. An outline of the proposed consensus statement was devel-oped and approved by the Executive Committee and Board of Directors of SNACC.The task force members agreed on criteria for evidence and then evaluated peer-reviewed stud-ies pertaining to perioperative stroke.Recommendations were developed by the task force and then assessed by the Executive Committee of SNACC.After incorporating input from the Executive Committee,the Task Force presented the draft guidelines to the international membership of SNACC through its website;thirty days were allowed for suggested revisions.This input and all further available information were incorporated into the consensus statement.The con-sensus statement was then reviewed and,after further revision, was supported by the American Society of Anesthesiologists. AVAILABILITY AND STRENGTH OF EVIDENCE This consensus statement was informed by published literature obtained t
hrough PubMed and other health-care databases,direct internet searches,task force members,and manual searches of references found in review articles.Peer-reviewed articles were considered that provided either sci-entific ,randomized controlled trial,RCT)or opinion-based evidence.
a)Scientific Evidence
(i)Category A:Supportive Literature
1.Level1:The literature contains multiple RCTs and
findings are supported by meta-analysis.
2.Level2:The literature contains multiple RCTs,but no
meta-analysis is possible.
3.Level3:Only one RCT exists in the literature.
(ii)Category B:Suggestive Literature
1.Level1:The literature contains observational compar-
isons of ,case-control study)that indicate statistically significant differences with respect to the outcome of perioperative stroke.
2.Level2:The literature contains noncomparative obser-
vational studies with associative or descriptive statistics.
3.Level3:The literature contains case reports.(iii)Category C:Equivocal Literature
1.Level1:Meta-analysis did notfind significant differ-
ences.
2.Level2:No meta-analysis is possible;RCTs found
inconsistent evidence.
3.Level3:Observational studies report inconsistent
findings that do not permit inference.
(iv)Category D:Insufficient Evidence from the Literature 1.Silent:No identified studies exist on the relationship
between intervention and perioperative stroke.
2.Inadequate:The literature does not permit clear
interpretation offindings due to methodological considerations.
Opinion-based Evidence
Category A:Expert opinion from task-force consultants. Category B:Membership opinion obtained from survey. Category C:Informal opinion from open-forum testimony, internet-based comments,and other communications.
PREOPERATIVE RECOMMENDATIONS Identifying patients at high risk of stroke In order to prevent perioperative stroke,it is critical to identify those at high risk for the complication.The in-cidence of stroke in a broad surgical population(excluding cardiac,carotid,major vascular and neurologic surgery)is approximately1per1000cases1and approximately6per 1000cases after major vascular surgery below the dia-phragm6;perioperative stroke increased length of hospital stay and risk of death.
These data are derived from more than550,000patients across two studies of the American College of Surgeons National Surgical Quality Improvement Program(ACS-NSQIP)database,which is currently the highest quality dataset available for the epidemiology of perioperative stroke incidence and outcome.The incidence of perioperative stroke stratified by case type can be found in Table1.The significant majority of perioperative strokes are ischemic rather than hemorrhagic7,8;the ACS-NSQIP database does not distinguish between the two.Preliminary data of the NeuroVision trial,conducted in noncardiac surgery patients with cardiovascular risk factors,suggest that the incidence of covert ,without obvious deficit)is 10%,as identified by magnetic resonance imaging in the postoperative period.9If confirmed by the larger trial,this finding could have important implications for the study and prevention of perioperative stroke after noncardiac surgery.
Due to the relative rarity of overt perioperative stroke in the noncardiac population,prospective identification of risk factors has been limited.Data regarding risk profile have therefore been derived from case series,case-control studies,or large database investigations.Table2shows the independent predictors of perioperative stroke found in recent major studies of the complication.Three of the most consistent risk factors for perioperative stroke identified in the literature are advanced age,renal failure,and a history of stroke or transient ischemic attack.1,6,10–14In general terms,all patients presenting for surgery with a history of
Mashour et al J Neurosurg Anesthesiol Volume26,Number4,October2014 274|www.jnsa r2014Lippincott Williams&Wilkins
cerebrovascular compromise should be regarded as high risk for perioperative stroke.
Informed Consent
Perspectives on informed consent from patients in the U.K.suggest that major complications of surgery with an incidence of>1%should be discussed15;a ma-jority of patients surveyed in a U.S.study suggests that rare but serious complications should also be discussed.16 Mashour et al1showed that patients presenting for non-cardiac,non-major-vascular surgery with any three or four of the risk factors listed in Table2have a0.7% incidence of perioperative stroke and withfive or more risk factors the incidence rises to1.9%.It is therefore reasonable to discuss risk of perioperative stroke in pa-tients with a history of stroke and other risk factors. Recommendations
(1)Screen for risk factors of perioperative stroke,most
notably remote or recent history of stroke,and communicate such risk to patients and providers (Category B,Level2).
Timing of Surgery After Stroke
Patients with acute or recent stroke have impaired cerebrovascular autoregulation and chemoregulation for months,17–20rendering them dependent on systemic pressure and passive perfusion.This dependence creates particular risk for cerebral hypoperfusion,especially in the setting of general anesthesia and the physiologic perturbations of surgery(such as hemorrhage,anemia, hypotension).It has been suggested that elective surgery should be delayed from1to3months after a stroke in order to prevent a secondary cerebrovascular event.21,22 To prevent perioperative stroke in patients with a history of recent cerebrovascular insult,it is likely beneficial to identify the cause of the initial stroke with investigations such as carotid imaging,magnetic resonance angio-graphy,or echocardiogram.Known carotid disease should be treated based on current guidelines.23,24 Despite the intuition that delaying surgery after stroke is beneficial,a study of173surgical patients with a history of recent and remote stroke found no relationship between timing of stroke history and incidence of perioperative stroke.21One retrospective study of hip or knee replacement after stroke or acute coronary syndrome found that stroke within six months prior to surgery was not a predictor of postoperative mortality.25Thesefindings were similar to a recent retrospective study of cardiac surgical patients;the time interval between stroke and coronary artery bypass graft surgery was not found to be
a predictor of postoperative stroke or mortality.26Ultimately,the decision to proceed with surgery will always be a balance between the risks of perioperative stroke and the risks of delaying surgery further.
TABLE1.Incidence of Stroke for Noncardiac,Nonvascular,Nonneurologic Surgeries
Stroke–All Age%,(n)Stroke–Age Z65%,(n) Bateman,et al,2009;Nationwide Inpatient Sample
Hip Arthroplasty(N=1,568)0.4(6)0.5(5)
Lung Resection(N=1,484)0.3(5)0.7(5) Colectomy(N=33,426)0.4(130)0.7(100) Mashour et al,2011;American College of Surgeons-National Surgical Quality Improvement Program
Hepatobiliary–Biliary Tree(N=43,289)0.1(36)0.2(23) Excisional breast(N=36,793)0.0(16)0.1(11)
Hernia–Ventral/Umbilical/Incisional/Other(N=32,638)0.1(28)0.3(21)
Hernia–Inguinal/Femoral Incisional Mesh(N=26,448)0.1(17)0.1(10) Colorectal–Appendectomy(N=26,046)0.0(6)0.2(4) Esophagogastric–Bariatric(N=23,766)0.0(5)0.0(0)
Head and Neck–Tumor(N=20,057)0.0(7)0.1(3)
Minor Vascular-Chest/Extremity(N=5,883)0.0(2)0.1(1)
Small Intestine–Resection/Ostomy(N=5,860)0.5(27)0.6(14)
Small Intestine–Lysis of adhesions,other(N=5,683)0.3(17)0.7(14) Abdominal–Exploration(N=5,760)0.5(26)0.9(18) Hepatobiliary–Pancreas(N=4,832)0.3(15)0.5(10) Musculoskeletal–Amputation(N=4,800)0.8(37) 1.1(29) Esophagogastric–Gastric(N=4,749)0.3(16)0.7(12) Esophagogastric(N=4,635)0.0(1)0.1(1) Hysterectomy(N=4,454)0.1(3)0.2(1) Musculoskeletal–Arthroscopy(N=4,255)0.0(0)0.0(0) Musculoskeletal–Spine(N=3,480)0.1(4)0.3(3) Colorectal–Abdominoperineal resection(N=3,169)0.2(7)0.5(5) Musculoskeletal–Knee(N=2,970)0.1(4)0.2(4)
Anorectal–Abscess(N=2,508)0.0(0)0.0(0)
Simple skin and soft tissue(N=2,383)0.3(6)0.6(4) Colorectal–Low anastomosis(N=2,293)0.2(4)0.2(2) Hepatobiliary–Liver(N=2,144)0.3(6)0.8(6)
Anorectal–Resection(N=2,103)0.0(1)0.0(0) Musculoskeletal–Fracture repair(N=2,065)0.1(3)0.3(3)
Biopsy skin and soft tissue(N=2,014)0.1(2)0.2(1)
J Neurosurg Anesthesiol Volume26,Number4,October2014Perioperative Stroke Consensus Statement r2014Lippincott Williams&Wilkins www.jnsa|275
Recommendations
1.Discuss surgical timing with a neurologist and consider
delaying elective surgical cases in patients with recent stroke until the etiology is investigated and the peak of autoregulatory disturbances has passed(likely at one month)(Opinion-based evidence,Category A).How-ever,observational studies to date do not suggest a clear relationship between timing of past stroke history and incidence of postoperative stroke(Category B,Level2). Management of Anti-coagulants and Anti-platelet Drugs
There are two common clinical scenarios in which management of anticoagulants or antiplatelet drugs can be linked to risk of perioperative stroke.Thefirst is manage-ment of anticoagulants for patients with atrialfibrillation,a major risk factor for perioperative stroke,and the second is the management of aspirin in patients with cardiovascular or cerebrovascular ,for primary or secondary stroke prevention).The clinical dilemma relates to balancing the risks of excessive perioperative bleeding and the risk of rebound hypercoagulability in the setting of the pro-thrombotic state induced by surgery;a su
mmary of surgical cases with various levels of expected blood loss can be found in Table3.27There are few data to guide management of this situation that pertain specifically to perioperative stroke.The American College of Chest Physicians recommends that heparin therapy be considered for postoperative atrialfi-brillation in patients with a history of stroke or transient ischemic attack28;an approach to the management of anti-coagulation therapy can be found in Table4.27,29,30 There are currently no studies of perioperative stroke and antiplatelet drug therapy for noncardiac,noncarotid surgery.Observational studies of cardiac surgery patients taking aspirin within5days prior to )re-vealed a protective effect of aspirin with respect to the outcome of perioperative stroke.31,32These studies are limited by an observational design but are consistent with a large randomized controlled trialfinding a benefit to postoperative aspirin in preventing stroke after coronary TABLE3.Possible Risk Classification for Bleeding According to Surgery/Procedure Type
Bleeding Risk Category Type of Surgery/Procedure High risk Intracranial or spine surgery
Major vascular surgery(aortic
aneurysm repair,aortofemoral
bypass)
Major urologic surgery(prostatectomy,
bladder tumor resection)
Major orthopedic surgery(hip
replacement)
Lung resection
Intestinal anastomosis surgery
Permanent pacemaker or internal
defibrillator placement
Selected procedures:colonic
polypectomy of large polyp,
endoscopic retrograde
cholangiopancreatography with
sphincterotomy,kidney biopsy Moderate risk Other intra-abdominal surgery
Other intrathoracic surgery
Other orthopedic surgery
Other vascular surgery
Selected procedures:colonic
polypectomy,prostate biopsy,
cervical biopsy
Low risk Laparoscopic cholecystectomy
Laparoscopic inguinal hernia repair
Noncataract ophthalmologic
procedures
Coronary angiography
Gastroscopy or colonoscopy(with/
without biopsy)
Very low risk(anticoagulation
interruption not required)
Single tooth extraction
Skin biopsy or selected skin cancer
removal
Cataract removal
Note that risk in certain cases may also relate to the consequences of bleeding (e.g.,intracranial or spi
ne procedure)rather than merely the volume of bleeding. Table reproduced with permission from Darvish-Kazem and Douketis,Perioper-ative management of patients having noncardiac surgery who are receiving anti-coagulant or antiplatelet therapy:an evidence-based but practical approach.Semin Thromb Hemost2012;38:652-660.
TABLE2.Independent Predictors of Perioperative Stroke Identified in Large Epidemiologic Studies
Predictors Odds
Ratio
Confidence
Intervals
Independent Predictors found in Bateman et al,2009.Nationwide
Inpatient Sample;hip,colon and lung surgery
Renal disease 2.98 2.52to3.54
Atrialfibrillation 1.95 1.69to2.26
History of stroke 1.64 1.25to2.14
Valvular disease 1.54 1.25to1.90
Congestive heart failure 1.44 1.21to1.70
Age(per10years) 1.43 1.35to1.51
Diabetes mellitus 1.18 1.01to1.39
Female(vs.Male) 1.21 1.07to1.36
Independent predictors found in Mashour et al,2011.American College of
Surgeons-National Surgical Quality Improvement Program;broad
population of noncardiac,nonvascular,nonneurologic surgery
Age Z62years 3.9 3.0to5.0
Myocardial infarction within6months 3.8 2.4to6.0
Acute renal failure 3.6 2.3to5.8
History of stroke 2.9 2.3to3.8
Pre-existing dialysis 2.3 1.6to3.4
Hypertension requiring medication 2.0 1.6to2.6
History of transient ischemic attack 1.9 1.3to2.6
Chronic obstructive pulmonary disease 1.8 1.4to2.4
Current smoker 1.5 1.1to1.9
Body mass index35-40kg/m2(protective)0.60.4to0.9
Independent predictors found in Sharifpour,Moore et al,2013.American
College of Surgeons-National Surgical Quality Improvement Program;
noncarotid vascular surgery
Acute renal failure 2.03 1.39to2.97
History of stroke,transient ischemic attack,or
hemiplegia
1.72 1.29to
2.30
Female(vs.Male) 1.47 1.12to1.93
History of cardiac disease(myocardial
infarction,congestive heart failure,angina,
prior cardiac intervention)
1.42 1.07to1.87
Age(each additional year of life) 1.02 1.01to1.04
Note that certain variables relevant to stroke,such as atrialfibrillation and
valvular disease,are not collected in the National Surgical Quality Improvement
Program database.
Mashour et al J Neurosurg Anesthesiol Volume26,Number4,October2014 276|www.jnsa r2014Lippincott Williams&Wilkins
artery bypass surgery.33It is as yet unclear how these data apply to noncardiac surgery,although recent studies in patients undergoing hip arthroplasty suggest that aspirin reduces perioperative stroke.34In a nonoperative pop-ulation,withdrawal of antiplatelet and antithrombotic medications was associated with a5.2%incidence of stroke within60days of drug cessation.35Furthermore,non-operative patients having strokes while offantiplatelet and antithrombotic drugs had greater morbidity and mortality compared to patients who continued taking them.35How-ever,failing to stop these agents preoperatively may place patients at increased risk of intraoperative hemorrhage, which also increases the risk for perioperative stroke(see section on Intraoperative Recommendations).
Recently,the PeriOperative Ischemic Evaluation (POISE)-2trial demonstrated that perioperative aspirin did not reduce the incidence of death or nonfatal my-ocardial infarction after noncardiac surgery,but did in-crease the risk of major bleeding.36Of note,patients who had aspirin therapy that was initiated in the course of the study had a reduced incidence of stroke compared to placebo;patients who were continuing aspirin therapy showed no reduction in stroke incidence.Stroke was not a primary outcome of the study and the authors acknowl-edged that thefindings in the initiation group could be spurious.However,the results suggest that some patients at risk of stroke may benefit from preoperative initiation of aspirin therapy,but this must be balanced against the now well-documented risk of significant increases in bleeding and must be demonstrated in a larger trial. Recommendations
(1)Medically manage atrialfibrillation and continue
anticoagulation in patients with atrialfibrillation for
minor surgeries or those in which high blood loss is unlikely.Discontinue anticoagulation in surgical patients at high risk of bleeding(with appropriate bridging strategies as indicated),but resume as soon as the risk of surgical bleeding is considered to be low (Opinion-based evidence,Category A).
(2)There is no evidence to suggest that continuation of
aspirin in patients at risk for vascular complications reduces the risk of stroke after noncardiac surgery (Category A,Level3).
Role of Preoperative Beta Blockers and Statins in Perioperative Stroke
The original POISE trial,which evaluated the car-dioprotective effects of metoprolol in8351noncardiac surgery patients,demonstrated that patients receiving me-toprolol had a significantly higher risk of stroke(hazard ratio 2.17,p=0.005)and death(hazard ratio 1.33, p=0.032).37Data from the POISE trial and other inves-tigations contributed to a meta-analysis suggesting that beta blockers increase risk of nonfatal stroke after non-cardiac surgery38;patients from the POISE trial constituted the majority in this study.A retrospective case-control study subsequent to the POISE trial suggested no increased risk of perioperative stroke with clinically routine doses of beta blockers39;a study of low-dose bisoprolol also con-cluded that there was no increased risk of perioperative stroke.40Based on these studies it is unclear if there is a drug-specific effect of metoprolol that increases risk of stroke or whether increased risk in the POISE trial was due to bradycardia and hypotension in the treatment group. Based on retrospective data from a single-center ob-servational study,Mashour et al8demonstrated that rou-tinely prolol conferred a higher risk of postoperative stroke compared to a matched cohort taking atenolol.These data are consistent with the results of a U.
S.-wide Veterans Administration hospital study by London et al showing a higher risk of stroke after non-cardiac surgery in patients taking metoprolol compared to atenolol.41In a single-center observational study,Ashes et al demonstrated that bisoprolol is associated with a lower stroke risk than either metoprolol or atenolol.42Current guidelines endorse the perioperative continuation of beta blockers in surgical patients who are already taking this class of drug.Large prospective studies are required to confirm that surgical patients who continue perioperative beta blockers are at increased risk of stroke if metoprolol is administered vs.another beta blocker.In surgical patients who are beta-blocker-naı¨ve,high-dose beta blockers should not be administered without dose titration.43
As with beta blockers,discontinuation of statins in the perioperative period may have adverse consequences. With respect to nonoperative stroke,discontinuation of statins in individuals with acute ischemic stroke was asso-ciated with a high risk of early neurologic deterioration.44A recent,preliminary,retrospective study of asymptomatic surgical patients presenting for carotid endarterectomy suggested that statins could reduce neurologic injury,as defined by both stroke and cognitive dysfunction.45Statins
TABLE4.Approach to Perioperative Bridging of Anticoagulation Management
Patient
Group Preoperative management Postoperative management
Low-to-moderate bleeding risk Stop therapeutic-dose
LMWH bridging on
morning(20-24hours)
before surgery
Omit evening dose with
twice-daily regimen
Give50%total dose with
once-daily regimen
Resume therapeutic-dose
LMWH bridging24
hours after surgery and
when there is adequate
hemostasis
High bleeding risk Stop therapeutic-dose
LMWH bridging on
morning(20-24hours)
before surgery
Omit evening dose with
twice-daily regimen
Give50%total dose with
once-daily regimen
Resume therapeutic-dose
LMWH bridging48-72
hours after surgery and
when there is adequate
hemostasis
Administer low-dose
LMWH or avoid
postoperative bridging
LMWH=low-molecular-weight heparin
Table reproduced with permission from Darvish-Kazem and Douketis,Peri-
operative management of patients having noncardiac surgery who are receiving
anticoagulant or antiplatelet therapy:an evidence-based but practical approach.
Semin Thromb Hemost2012;38:652–660.
J Neurosurg Anesthesiol Volume26,Number4,October2014Perioperative Stroke Consensus Statement r2014Lippincott Williams&Wilkins www.jnsa|277
have also been shown to reduce the incidence of atrialfib-rillation and other adverse outcomes that may be associated with postoperative stroke.46,47However,there are no data to suggest that starting statins in the preoperative period can prevent stroke around the time of noncardiac,non-neurologic and noncarotid surgery. Recommendations
(1)Metoprolol or other beta blockers should only be
started in the preoperative period with careful titration(Category A,Level3).
(2)Continue beta blockers and statins throughout the
perioperative period in patients already taking them (Opinion-based evidence,Category A with respect to stroke risk).
INTRAOPERATIVE RECOMMENDATIONS
Intraoperative events are frequently cited as a cause of postoperative stroke,despite controversial evidence. Stroke presenting in the postoperative period on the same day of surgery(which would suggest a clear intra-operative etiology)is relatively infrequent.8In this section we review the available data guiding intraoperative management of anesthetic technique,ventilation strategy,fluid and blood transfusion,glycemic control and blood pressure.
Anesthetic Technique
There has been extensive interest in the possible neuroprotective effects of anesthetic agents.Most studies have been in patients for whom cerebral ischemia is pre-dictable;examples of predictable cerebral ischemia include carotid endarterectomy,cerebral aneurysm surgery and procedures requiring deep hypothermic circulatory arrest. For these procedures,anesthetic technique may be adjusted prospectively in an attempt to minimize neurologic injury. However,data supporting anesthetic neuroprotection even for these procedures is limited or absent.Given the low incidence of perioperative stroke in noncardiac and non-vascular surgical patients(making RCT difficult),inves-tigators have studied neurologic outcome after cardiac surgery,in which the stroke risk is higher.Bilotta
et al performed an extensive literature review of randomized trials but could not make definitive conclusions due to the small number and heterogeneity of studies.48
It is also important to consider anesthetics that might potentially increase stroke risk.For example,ni-trous oxide is associated with an acute increase in plasma homocysteine concentrations,which could impair endo-thelial function and increase adverse cardiovascular events.49However,no association has been demonstrated in several large studies between intraoperative admin-istration of nitrous oxide and postoperative stroke.8,49–51
A recent retrospective review of57,000patients8 revealed no difference in stroke risk between regional and general anesthesia in noncardiac patients.However,a large database study(>200,000patients)focusing on knee and hip arthroplasty found that neuraxial anesthesia was associated with a lower incidence of stroke(0.07%) compared to combined neuraxial/general anesthesia (0.12%)and general anesthesia(0.13%,p=0.006).52 Overall30-day mortality was also reduced in the neu-raxial(0.10;n=40,036)and combined neuraxial-general (0.10;n=49,396)groups compared to the general anes-thesia group(0.18;n=292,804;p<0.001).Similarly,in a2010single center observational study of18,745con-secutive joint replacements,general anesthesia was an independent predictor of postoperative stroke(OR3.54, 95%CI1.01-12.39).53In another patient population from the
GALA trial,which studied carotid endarterectomy patients,there was no difference in stroke rates between general and regional anesthesia.54Thus,while there are no data in a broad,representative surgical population to support ional anesthetic technique,there are now at least two large outcomes studies suggesting that regional anesthesia for hip and knee arthroplasty may be associated with a lower risk of perioperative stroke.
Recommendations
1.Despite characteristics that would appear to predis-
pose to perioperative stroke,nitrous oxide use has not been associated with an increased incidence of perioperative stroke(Category B,Level1).
2.Recent retrospective data suggest that neuraxial
techniques may be associated with a lower incidence of perioperative stroke for hip and knee arthroplasty (Category B,Level1).No such data exist for other surgical populations.
Intraoperative Use of Beta Blockers
The most important intraoperative pharmacologic association with postoperative stroke is the admin-istration of beta blockers.Mashour et al8retrospectively studied57,218noncardiac patients,of whom55had perioperative strokes.Intraoperative metoprolol admin-istration was associated with a3.3-fold increased risk of perioperative stroke(p=0.003;95%CI1.4-7.8).No such association was found for intraoperative esmolol or la-betolol.While the investigators also found intraoperative hypotension to be associated with perioperative stroke, no co-linearity existed between intraoperative metoprolol administration and hypotension.
documented evidenceThe mechanism of metoprolol’s apparent role in perioperative stroke is unclear.There is wide diversity in beta1-adrenergic selectivity among beta antagonists,with bisoprolol having the greatest selectivity,atenolol inter-mediate and metoprolol the least selectivity among clin-ically used“cardioselective”beta blockers.55Recent animal data suggest that metoprolol,as a relatively nonselective beta1antagonist,may reduce brain tissue oxygenation by impairing beta2mediated cerebral vaso-dilation in mice.56Furthermore,in rats,metoprolol im-pairs the compensatory increase in cardiac output that occurs in response to anemia,reducing cerebral tissue oxygenation.57
Mashour et al J Neurosurg Anesthesiol Volume26,Number4,October2014 278|www.jnsa r2014Lippincott Williams&Wilkins
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论