Secondary Logo

Perioperative Low Arterial Oxygenation Is Associated With Increased Stroke Risk in Cardiac Surgery

Dunham, Alexandra M., MD*; Grega, Maura A., MSN; Brown, Charles H. IV, MD; McKhann, Guy M., MD; Baumgartner, William A., MD§; Gottesman, Rebecca F., MD, PhD

doi: 10.1213/ANE.0000000000002157
Hemostasis: Original Clinical Research Report

BACKGROUND: Both patient characteristics and intraoperative factors have been associated with a higher risk of stroke after cardiac surgery. We hypothesized that poor systemic oxygenation in the perioperative period is associated with increased risk of stroke following cardiopulmonary bypass.

METHODS: In this study of 251 adult patients who underwent cardiopulmonary bypass procedures at a single center from 2003 to 2006, cases (patients with a postoperative stroke at least 24 hours after surgery) were matched 1:2 to controls without stroke. Minimum and average partial pressure of oxygen in arterial blood (PaO2) values, from arterial blood gas values during and up to 24 hours after surgery, were evaluated as continuous and categorical predictors. Conditional logistic regression models adjusted for potential confounders (demographics, comorbidities, and intraoperative variables) were used to evaluate associations between PaO2 variables and stroke status.

RESULTS: Lower nadir PaO2 values were associated with postoperative stroke, with estimated odds of stroke increasing over 20% (adjusted odds ratio [OR], 1.23; 95% confidence interval [CI], 1.07–1.41) per 10 mm Hg lower nadir PaO2, and similarly increased odds of stroke per lower quartile of nadir PaO2 (OR, 1.60; 95% CI, 1.19–2.16). When average PaO2 was considered, odds of stroke was also increased (adjusted OR, 1.39 per lower quartile of mean PaO2; 95% CI, 1.05–1.83). Having a nadir PaO2 value in the lowest versus any other quartile was associated with an estimated 2.41-fold increased odds of stroke (95% CI, 1.22–4.78). Quartile of nadir but not average PaO2 results remained significant after adjustment for multiple comparisons.

CONCLUSIONS: Odds of stroke after cardiac surgery are increased in patients with a low minimum PaO2 within 24 hours of surgery. Results should be validated in an independent cohort. Further characterizing the underlying etiology of hypoxic episodes will be important to improve patient outcomes.

From the Departments of *Surgery, Neurology, Anesthesiology, and §Cardiac Surgery, the Johns Hopkins University School of Medicine, Baltimore, Maryland.

Accepted for publication March 24, 2017.

Funding: The Johns Hopkins University School of Medicine, Deans Office Summer Research Grant (A.M.D.), McKhann Scholar Award (R.F.G.), Dana Foundation (G.M.M.).

Conflict of Interest: See Disclosures at the end of the article.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website.

IRB Information: The Johns Hopkins University, School of Medicine, Institutional Review Board; East Baltimore Campus (Central Office), 1620 McElderry St, Reed Hall–B130, Baltimore, MD 21205; 410-955-3008;

Reprints will not be available from the authors.

Address correspondence to Rebecca F. Gottesman, MD, PhD, Phipps 446D, 600 North Wolfe St, Baltimore, MD 21287. Address e-mail to

Between 3% and 9% of people undergoing surgery with cardiopulmonary bypass (CPB) experience postoperative clinical strokes,1 which largely consists of watershed2 and embolic types. Several factors are likely at play including systemic inflammation, embolization, and hypotension. In addition, the intraoperative setting may lead to other alterations in cerebral perfusion and oxygenation that might impact stroke risk; we and others have previously reported that anemia is associated with post-CPB stroke.3,4 Patients undergoing CPB cardiac surgery are also exposed to a wide range of oxygenation levels throughout the intraoperative and perioperative period. Hypoxia in particular might represent another form of injury to the brain that might lead to ischemia.

Hypoxic–ischemic neurologic injury is caused by any event that compromises the ability of the brain to be properly oxygenated5,6 and has been described after cardiac arrest, asphyxia, carbon monoxide toxicity, and more recently in sleep apnea.6–8 Systemic, objective measurements of oxygenation can be used as an estimate of the likely neuronal environment at the time surrounding cerebral hypoxia events. Neurologic complications of hypoxia during cardiac surgery have been studied in neonates with hypoplastic left heart syndrome, demonstrating worse neurocognitive outcomes but no increased risk of stroke in neonates with profound hypoxia,9 but to our knowledge, hypoxia during adult cardiac surgery has not been studied as a risk factor for adverse neurologic events.

In this single-center study, we evaluated whether hypoxia or other fluctuations in oxygenation in the intraoperative to early postoperative period were associated with an increased risk of stroke following cardiac surgery. In this case–control study of CPB cardiac surgery patients, we compared partial pressure of oxygen in arterial blood (PaO2) values from the start of anesthesia to 24 hours after that time point, in individuals with and without stroke. We chose this 24-hour perioperative period to capture intraoperative and early postoperative changes in blood oxygenation occurring before a diagnosis of stroke. We hypothesized that individuals having postoperative stroke would experience lower PaO2 values than their nonstroke controls and that low PaO2 values would be a predictor of postoperative stroke.

Back to Top | Article Outline


The study was approved by the appropriate institutional review board (IRB), and the requirement for written informed consent was waived by the IRB. This article adheres to the applicable EQUATOR (Enhancing the QUAlity and Transparency Of health Research) guidelines. All patients undergoing cardiac surgery at Johns Hopkins Hospital (Baltimore, MD) have been followed prospectively for the development of postoperative neurologic deficits since 1992. For patients experiencing neurologic deficits, neurologic consultations are requested, and if the consultant diagnoses the patient with a stroke (using a combination of clinical information and neuroimaging), it is entered into the Cardiac Surgery Stroke Database. In addition, all Cardiac Surgery cases are maintained in a prospective database. “Cases” are those patients with a postoperative clinically defined stroke, and “Controls” are patients in the Cardiac Surgery Database but not in the Cardiac Surgery Stroke database. Controls were matched 2:1 to cases on the basis of sex, 20-year age block, year of procedure, and type of procedure (isolated coronary artery bypass graft [CABG]/redo CABG, combined CABG with valve repair, combined CABG with carotid endarterectomy, combined CABG with other, aortic procedure, isolated valve repair, or other cardiac surgery). An individual may serve as control for more than one case.10

Records were reviewed for a total of 389 patients, cases and controls, who had undergone cardiac surgery between 2003 and 2006. There was no known large-scale, institutional change in anesthesia practices during this time period. Information on basic demographics, intraoperative factors (CPB time, blood pressure2), and vascular history (including history of hypertension, diabetes, hypercholesterolemia, smoking, myocardial infarction, obesity, prior stroke and ischemic attack) was entered into each of the databases prospectively, at the time of the patient’s hospitalization. For entries with missing data in these categories, full electronic medical records were reviewed. Lack of information confirming obesity, current smoking, or history of myocardial infarction was recorded as absence of each of these conditions. Arterial blood gas value (specifically, PaO2) levels were recorded during surgery (from the start of anesthesia) and up to 24 hours after the start of surgery. Information was recorded from electronic records, hospital records, and perfusion charts. Electronic medical record access allowed the data collector to investigate records extending back to 2003. The reviewer was blinded to the case status of patients at the time of data collection. Both cases and controls were reviewed in conjunction, in random order. After matching was completed, we further excluded all strokes (and their controls) that occurred in the first 24 hours after surgery to assure no overlap between the exposure and outcome time periods. All analyses described below used this population.

Given the final sample size after these exclusions (163 controls and 88 cases), with 19% of controls having minimum PaO2 values in the lowest quartile, we had 90% power (assuming α = .05) to detect a univariate odds ratio (OR) of 2.61 or greater.

Back to Top | Article Outline

Statistical Analysis

Stata version 13.1 for Macintosh (StataCorp, College Station, TX) was used for all statistical analyses. The primary independent variable, selected a priori, was minimum PaO2 value, as measured from arterial blood gases; average PaO2 was considered as a secondary outcome. Individuals with 4 or more PaO2 values only were included for further analysis. χ2 statistics for categorical values and t tests for continuous variables (accounting for unequal variance when appropriate) were used to explore relationships between variables of interest and stroke status. Among these variables include patient demographics and history, intraoperative data, and arterial blood gas measurements. Oxygenation variables were evaluated as continuous and categorical (in quartiles) predictors of stroke.

Given the concern for potential confounding in evaluation of the association between oxygen levels and postoperative stroke, we considered as covariates those variables with evidence of a univariate association with stroke at P < .2, plus those demographic and medical variables felt to be particularly clinically important (age and CPB time).

Conditional logistic regression models (conditional on case: control pairing) were used to evaluate associations between average PaO2 and minimum PaO2, each, in separate models, and stroke status, with adjustment for the covariates defined as described above. Linear and nonlinear models of fit were explored to further describe the association between nadir oxygen levels and stroke. Linear, quadratic, cubic spline, and fractional polynomial models were all tested to compare goodness of fit. Akaike and Bayesian information criterion were used to select the best fitting model. In addition, a Bonferroni correction was applied to P values given multiple comparisons (of multiple definitions of PaO2 levels, including average and nadir levels, and continuous as well as categorical definitions). Finally, in a sensitivity analysis, we present the results using propensity scores. This is not our primary analysis given the limitations of use of propensity scores in case–control studies,11,12 and uses a generalized propensity score13 to predict the continuous exposure of log-transformed nadir PaO2, our primary exposure of interest, using a maximum likelihood regression. This score included adjustment for age, sex, history of hypertension, hypercholesterolemia, diabetes, prior stroke, peripheral vascular disease, current smoking status, history of myocardial infarction, obesity, CPB time, presence of any circulatory arrest, and procedure type. The propensity score was checked for balance (see Supplemental Digital Content 1, Table, and was included as a covariate in the conditional logistic regression models.

Back to Top | Article Outline


Table 1

Table 1

The original cohort of 389 patients was reduced to 251 for final analysis (88 cases and 163 controls), after exclusions (see Supplemental Digital Content 2, Figure, Five individuals were excluded for having fewer than 4 recorded PaO2 values, 40 were excluded because they were missing data in demographic fields felt important for covariate adjustment (hypertension, hyperlipidemia, diabetes, prior stroke, and bypass time), and 21 excluded for having had a stroke within the first 24 hours (and thus during a partially overlapping time period with PaO2 evaluation). Individuals with less than 4 recorded PaO2 values were excluded for further analysis because we anticipated that those with fewer values had only minor surgical procedures (ie, wound repair) and therefore would have not been on CPB. Review of medical records confirmed that the 5 patients excluded for this reason did not undergo CPB for their procedure. For conditional logistic regression, only cases with at least one control and only controls with a corresponding case were included, accounting for an additional exclusion of 72 patients. These exclusions left 251 patients who are included in this study. Demographic information is reported in Table 1.

Back to Top | Article Outline

Univariate Analyses for Oxygenation and Stroke Status

Figure 1

Figure 1

Figure 2

Figure 2

In univariate analysis, mean average PaO2 was significantly lower in cases with postoperative stroke than in nonstroke controls (189.6 vs 202.2 mm Hg, P = .03). Similarly, mean nadir PaO2 was also lower in persons with stroke than in controls (81.3 vs 94.7 mm Hg, P = .0001). Figures 1 and 2 show the quartile distributions of nadir and average PaO2 values in cases and controls, further supporting the observation of lower values in stroke patients (Figure 1—average PaO2P = .08; Figure 2—nadir PaO2P = .0002).

Back to Top | Article Outline

Multivariable Analyses

Even after adjustment for potential confounders, lower PaO2 values remained consistently and significantly associated with postoperative stroke, particularly for evaluation of nadir values, one of which remained significant after consideration of multiple comparisons. Per 10 mm Hg lower nadir PaO2 value, estimated odds of stroke increased by over 20% (adjusted OR, 1.23; 95% confidence interval [CI], 1.07–1.41). Each 10 mm Hg decrease in mean PaO2 was associated with smaller and not significantly increased estimated odds of stroke (adjusted OR, 1.07; 95% CI, 1.00–1.15; Table 2). Belonging to a lower quartile of minimum oxygenation (having a lower nadir PaO2) was associated with an estimated adjusted 1.60-fold increase of odds (95% CI, 1.19–2.16) of having a stroke when compared to an individual in the next highest quartile (Table 2). A similar, but smaller, pattern of association was found with average PaO2, although this lost statistical significance with consideration of multiple comparisons; belonging to a lower quartile of average PaO2 increased estimated odds of having postoperative stroke by 1.39-fold (adjusted; 95% CI, 1.05–1.83).

Table 2

Table 2

A near-stepwise progression of increasing significance and increased estimated odds of having stroke was seen in comparing each quartile of minimum PaO2 to the lowest quartile. This relationship is shown in Table 2. In addition, having a nadir PaO2 value in the lowest (38–70 mm Hg) versus highest (106–257 mm Hg) quartile was associated with an estimated 6.51-fold increased adjusted odds of stroke (95% CI, 1.38–30.71). Belonging to the lowest quartile of nadir PaO2 versus all other quartiles combined was associated with an estimated 2.41-fold increased odds of stroke (95% CI, 1.22–4.78).

Back to Top | Article Outline

Sensitivity Analysis

In a sensitivity analysis, instead of including potential confounders in our final model, we created a generalized propensity score and adjusted for this in the final model. These results were very similar to the primary results (see Supplemental Digital Content 1, Table,

Back to Top | Article Outline

Models of Fit

Both linear and nonlinear models were explored to further characterize the association between nadir oxygen levels and postoperative stroke. The quality of each model was assessed by weighing goodness of fit and complexity using Akaike and Bayesian information criterion. Models remained statistically similar; however, the linear model had the best statistical fit so we used this for the primary results.

Back to Top | Article Outline


This single-center retrospective study found that lower PaO2 values were associated with increased risk of postoperative stroke, particularly for nadir PaO2, independent of other risk factors that might indicate a complex surgical course or increased risk of an adverse outcome. These data suggest that, after further supportive evidence, avoidance of hypoxia might be a potential target for treatment for reducing the burden of postoperative stroke.

Relatively high pump flow rates are commonly used during CPB surgery to increase delivery of oxygen to tissues and prevent organ injury. Higher pump flow rates aim to raise mean arterial blood pressures, to maintain pre-CPB perfusion pressures, and to compensate for decreased oxygen delivery inherent in hemodilution, which reduces hematocrit and oxygen-carrying capacity.14,15 Hemodilution offsets increases in blood viscosity due to hypothermia. Therapeutic hypothermia has been used for decades to create a favorable balance between oxygen supply and demand. Hypothermia offers cerebral protection by reducing cerebral metabolic requirements; however, hypothermia shifts the oxygen dissociation curve to the left. The leftward shift caused by hypothermia increases CO2 solubility and promotes alkalosis, resulting in a further leftward shift of the oxygen dissociation curve. These changes may potentially lead to impaired unloading of oxygen to organ tissues and organ injury, which may further exacerbate any observed decreases in measured PaO2 level.

The recognized clinical consequences of cerebral hypoxia range from transient ischemic attacks with no durable consequence to the more frequent outcome of devastating functional impairment and death. A brief hypoxic event can cause reversible damage to neuronal tissue.5,9 Downstream, secondary injuries can continue through an excitotoxicity cascade well after the primary insult and potentially lead to permanent tissue damage. The extent of ultimate damage is often more widespread than the primary event, secondary to edema and inflammation. It is likely that the trajectory of secondary brain injury can be modified by systemic parameters including hypotension, electrolytes, fever, and continued hypoxia.5 Hypoxia during cardiac surgery is certainly not acting alone in its association with postoperative stroke, given the large number of other fluctuations in blood pressure, inflammation, and other systemic changes, but our data suggest that it may not only be an innocent bystander in increasing risk of stroke.

These data support other increasing evidence that stroke in the setting after cardiac surgery is not entirely due to embolization, as previously believed.16 Our own work has demonstrated increased frequency of watershed-type strokes, which are often felt to be due to low flow states, and which, in our prior studies, were associated with larger drops in blood pressure during surgery.2 We have also identified decreases in hemoglobin as a risk factor for stroke after cardiac surgery,3 and impaired autoregulation, likely representing an increased risk of ischemia in the setting of impaired flow, perfusion, or perhaps even oxygenation, is also another identified risk factor for postcardiac surgery stroke.17 A hemodynamic mechanism of injury—whether due to low blood pressure, low relative blood pressure, low oxygen-carrying capacity of red blood cells due to anemia, or due to systemic hypoxia—needs to be recognized as a potential important factor contributing to postoperative neurologic complications, including stroke. The role of hypoxia or hemodynamic compromise is widely recognized as a contributor to acute kidney injury,18 and increasing data suggest that mechanisms of injury to the brain may share some of these mechanisms. An acute drop in oxygen level, even if for a brief period of time (as indicated by a low nadir value), has been implicated as an important risk factor for postoperative kidney injury,19 and our data suggest a similar role for brain injury.

Our single-center, case–control study is not without limitations. Potential unmeasured confounders may have influenced the results. Despite our consideration of multiple potential confounders of the relationship between oxygen level and postoperative stroke, it is still plausible that residual confounders exist, thus leading to an overestimation of a true association. We further attempted to account for this confounding in our sensitivity analysis, which includes a generalized propensity score, although this method has limitations in case–control studies given the fact that cases and controls are selected so exposure probabilities are not representative of the true population.12 In addition, medical history information from admission history and physical forms may have been incomplete, which may have inadequately captured comorbidity and thus confounder information (although these were entered before surgery, so should not be differential with respect to case status). Furthermore, it is possible that lower nadir oxygen or lower mean oxygen levels are simply markers of a complex surgical course not otherwise captured by CPB time nor other vascular risk factors, which might increase stroke risk. Our models only include a few indicators of the intraoperative course (circulatory arrest and bypass time). In addition, our sample represented a single institution’s experience. Patients undergoing CPB are generally maintained at very high flow rates of oxygen, as confirmed by the high mean PaO2 values noted in our study, but our data do suggest that some patients experience significant drops in oxygenation that might be dangerous to the brain. This practice did not change in any substantive way during the course of the study; patients were generally ventilated with 100% oxygen during the operative period, and PaO2 targets during bypass were at least >200 and were usually higher. In the postoperative period, patients were generally started on 100% fraction of inspired oxygen (FIO2) in the intensive care unit, and this was weaned after the first arterial blood gas (ABG) toward a goal of 40% FIO2. Low PaO2 may reflect more significant pulmonary or systemic vascular disease, which serves as a marker for sicker, more complex patients. In addition, PaO2 may be an insufficient measurement of oxygenation that may be better described by oxygenation saturation. However, only a small minority of patients in our databases had consistent oxygenation saturation measurements on record.

Another limitation of the study is in the database definition of strokes. Strokes were all first defined clinically and postoperative neurologic assessment was only sought if patient symptoms suggested neurologic complication. It is possible that some of the control patients actually had a neurologic injury, including but not limited to undiagnosed stroke. Brain imaging, which would expand the definition of stroke to radiographic strokes,20 is not a standard procedure for all patients before and after cardiac surgery.

Despite these limitations, we believe our study provides valuable information about the possible role of hypoxia in the development of postoperative stroke. Although traditional emphasis has been on the role of embolism in post-CPB stroke, accumulating evidence supports additional mechanisms, and these observed associations with nadir PaO2 level, combined with our prior data describing associations with anemia (possibly leading to poor oxygen delivery)3 and with decreases in blood pressure (associated with watershed strokes, in particular)2 suggest that there may be an important additional component of inadequate blood/oxygen delivery to regions of the brain during CPB.

In conclusion, lower PaO2 values were consistently associated with increased risk of postoperative stroke for nadir PaO2. Our findings suggest that patients with lower oxygenation, as measured by PaO2, have a greater risk of developing postoperative stroke. Characterizing the underlying etiology of low oxygenation, whether it be because of pulmonary complications, unstable hemodynamics, vascular disease, or systemic inflammation, will be important to improve patient outcomes from cardiac surgery. Determining factors that predispose patients to hypoxic episodes may allow for prescreening and practical adjustments in the plan of care.21 Further studies are needed to address the value of using PaO2 measurements as a marker for postoperative stroke22 and to validate these findings in a separate cohort. In addition, further study is needed to evaluate the clinical impact of stabilizing measures of oxygenation after CPB and to determine if such a practice were to reduce postoperative stroke rates.

Back to Top | Article Outline


Name: Alexandra M. Dunham, MD.

Contribution: This author helped with drafting of the manuscript and analysis of data.

Conflicts of Interest: None.

Name: Maura A. Grega, MSN.

Contribution: This author helped with collection of data, interpretation of data, and critical review of the manuscript.

Conflicts of Interest: None.

Name: Charles H. Brown IV, MD.

Contribution: This author helped with interpretation of data and critical review of the manuscript.

Conflicts of Interest: None.

Name: Guy M. McKhann, MD.

Contribution: This author helped with collection of data, interpretation of data, and critical review of the manuscript.

Conflicts of Interest: None.

Name: William A. Baumgartner, MD.

Contribution: This author helped with collection of data, interpretation of data, and critical review of the manuscript.

Conflicts of Interest: None.

Name: Rebecca F. Gottesman, MD, PhD.

Contribution: This author helped with analysis and interpretation of data and drafting of the manuscript.

Conflicts of Interest: Dr Gottesman is an editor for the journal Neurology. No other conflicts.

This manuscript was handled by: Roman M. Sniecinski, MD.

Back to Top | Article Outline


1. McKhann GM, Grega MA, Borowicz LM Jr, Baumgartner WA, Selnes OA. Stroke and encephalopathy after cardiac surgery: an update. Stroke. 2006;37:562–571.
2. Gottesman RF, Sherman PM, Grega MA, et al. Watershed strokes after cardiac surgery: diagnosis, etiology, and outcome. Stroke. 2006;37:2306–2311.
3. Bahrainwala ZS, Grega MA, Hogue CW, et al. Intraoperative hemoglobin levels and transfusion independently predict stroke after cardiac operations. Ann Thorac Surg. 2011;91:1113–1118.
4. Kulier A, Levin J, Moser R, et al; Investigators of the Multicenter Study of Perioperative Ischemia Research Group; Ischemia Research and Education Foundation. Impact of preoperative anemia on outcome in patients undergoing coronary artery bypass graft surgery. Circulation. 2007;116:471–479.
5. Busl KM, Greer DM. Hypoxic-ischemic brain injury: pathophysiology, neuropathology and mechanisms. NeuroRehabilitation. 2010;26:5–13.
6. Huang BY, Castillo M. Hypoxic-ischemic brain injury: imaging findings from birth to adulthood. Radiographics. 2008;28:417–439.
7. Pamenter ME, Powell FL. Signalling mechanisms of long term facilitation of breathing with intermittent hypoxia. F1000Prime Rep. 2013;5:23.
8. Dumitrascu R, Heitmann J, Seeger W, Weissmann N, Schulz R. Obstructive sleep apnea, oxidative stress and cardiovascular disease: lessons from animal studies. Oxid Med Cell Longev. 2013;2013:234631.
9. Hoffman GM, Brosig CL, Mussatto KA, Tweddell JS, Ghanayem NS. Perioperative cerebral oxygen saturation in neonates with hypoplastic left heart syndrome and childhood neurodevelopmental outcome. J Thorac Cardiovasc Surg. 2013;146:1153–1164.
10. Wacholder S, Silverman DT, McLaughlin JK, Mandel JS. Selection of controls in case-control studies. III. Design options. Am J Epidemiol. 1992;135:1042–1050.
11. Månsson R, Joffe MM, Sun W, Hennessy S. On the estimation and use of propensity scores in case-control and case-cohort studies. Am J Epidemiol. 2007;166:332–339.
12. Allen AS, Satten GA. Control for confounding in case-control studies using the stratification score, a retrospective balancing score. Am J Epidemiol. 2011;173:752–760.
13. Hirano K, Imbens GW. Gelman A, Meng X-L. The propensity score with continuous treatments. In: Applied Bayesian Modeling and Causal Inference from Incomplete-Data Perspectives: An Essential Journal with Donald Rubin’s Statistical Family. 2004. Chichester, UK: John Wiley & Sons, Ltd; doi: 10.1002/0470090456.ch7.
14. Karkouti K, Beattie WS, Wijeysundera DN, et al. Hemodilution during cardiopulmonary bypass is an independent risk factor for acute renal failure in adult cardiac surgery. J Thorac Cardiovasc Surg. 2005;129:391–400.
15. Habib RH, Zacharias A, Schwann TA, et al. Role of hemodilutional anemia and transfusion during cardiopulmonary bypass in renal injury after coronary revascularization: implications on operative outcome. Crit Care Med. 2005;33:1749–1756.
16. Selnes OA, Gottesman RF, Grega MA, Baumgartner WA, Zeger SL, McKhann GM. Cognitive and neurologic outcomes after coronary-artery bypass surgery. N Engl J Med. 2012;366:250–257.
17. Ono M, Joshi B, Brady K, et al. Risks for impaired cerebral autoregulation during cardiopulmonary bypass and postoperative stroke. Br J Anaesth. 2012;109:391–398.
18. Thiele RH, Isbell JM, Rosner MH. AKI associated with cardiac surgery. Clin J Am Soc Nephrol. 2015;10:500–514.
19. Magruder JT, Dungan SP, Grimm JC, et al. Nadir oxygen delivery on bypass and hypotension increase acute kidney injury risk after cardiac operations. Ann Thorac Surg. 2015;100:1697–1703.
20. Sacco RL, Kasner SE, Broderick JP, et al; American Heart Association Stroke Council, Council on Cardiovascular Surgery and Anesthesia; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular and Stroke Nursing; Council on Epidemiology and Prevention; Council on Peripheral Vascular Disease; Council on Nutrition, Physical Activity and Metabolism. An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2013;44:2064–2089.
21. Carrascal Y, Guerrero AL, Blanco M, Valenzuela H, Pareja P, Laguna G. Postoperative stroke related to cardiac surgery in octogenarians. Interact Cardiovasc Thorac Surg. 2014;18:596–601.
22. Hill MD. Diagnostic biomarkers for stroke: a stroke neurologist’s perspective. Clin Chem. 2005;51:2001–2002.

Supplemental Digital Content

Back to Top | Article Outline
© 2017 International Anesthesia Research Society