Reviews21 July 2015

Intermediate- and Long-Term Cognitive Outcomes After Cardiovascular Procedures in Older Adults

A Systematic Review
    Author, Article, and Disclosure Information



    Risks for intermediate- and long-term cognitive impairment after cardiovascular procedures in older adults are poorly understood.


    To summarize evidence about cognitive outcomes in adults aged 65 years or older at least 3 months after coronary or carotid revascularization, cardiac valve procedures, or ablation for atrial fibrillation.

    Data Sources:

    MEDLINE, Cochrane, and Scopus databases from 1990 to January 2015;; and bibliographies of reviews and eligible studies.

    Study Selection:

    English-language trials and prospective cohort studies.

    Data Extraction:

    One reviewer extracted data, a second checked accuracy, and 2 independently rated quality and strength of evidence (SOE).

    Data Synthesis:

    17 trials and 4 cohort studies were included; 80% of patients were men, and mean age was 68 years. Cognitive function did not differ after the procedure between on- and off-pump coronary artery bypass grafting (CABG) (n = 6; low SOE), hypothermic and normothermic CABG (n = 3; moderate to low SOE), or CABG and medical management (n = 1; insufficient SOE). One trial reported lower risk for incident cognitive impairment with minimal versus conventional extracorporeal CABG (risk ratio, 0.34 [95% CI, 0.16 to 0.73]; low SOE). Two trials found no difference between surgical carotid revascularization and carotid stenting or angioplasty (low and insufficient SOE, respectively). One cohort study reported increased cognitive decline after transcatheter versus surgical aortic valve replacement but had large selection and outcome measurement biases (insufficient SOE).


    Mostly low to insufficient SOE; no pertinent data for ablation; limited generalizability to the most elderly patients, women, and persons with substantial baseline cognitive impairment; and possible selective reporting and publication bias.


    Intermediate- and long-term cognitive impairment in older adults attributable to the studied cardiovascular procedures may be uncommon. Nevertheless, clinicians counseling patients before these procedures should discuss the uncertainty in their risk for adverse cognitive outcomes.

    Primary Funding Source:

    Agency for Healthcare Research and Quality.

    Approximately 200 000 coronary artery bypass graft (CABG) surgeries, 50 000 carotid revascularizations, 50 000 cardiac valve replacements or repairs, and 10 000 catheter ablations for atrial fibrillation are performed annually in U.S. adults aged 65 years or older (1–6). The older U.S. population also has a high rate of cognitive impairment and dementia, with a combined incidence of 77.5 cases per 1000 person-years in adults aged 72 years or older with normal cognition at baseline (7). However, great uncertainty surrounds the relationship between these cardiovascular interventions and subsequent cognitive outcomes in older patients.

    Most attention has addressed the possible relationship between CABG and cognitive impairment. Early studies reported a high prevalence of cognitive impairment after CABG, possibly attributable to surgical factors, such as anesthesia and cardiopulmonary bypass (8, 9). Later studies suggested that much cognitive impairment after CABG predated the procedure (10) and was related to patient factors, such as age, education, and vascular disease (8, 11–14). Studies of carotid revascularization have reported mixed cognitive outcomes, with some suggesting early improvement (15). Studies of cardiac valve procedures and catheter ablation for atrial fibrillation commonly report imaging-detected cerebral emboli, but cognitive outcomes have been less clear (16, 17).

    However, understanding the longer-term cognitive outcomes attributable to these cardiovascular procedures in older adults has been limited because many studies had no control group, had short follow-up, had a predominance of middle-aged patients, or used surrogate outcomes for cognitive function. Meanwhile, older adults, who have the highest risk for intermediate- and long-term cognitive impairment, increasingly are having these procedures. Therefore, improved understanding of any relationship between these cardiovascular procedures and cognitive outcomes in older patients could enhance physician–patient clinical decision making.

    This systematic review evaluates the evidence from randomized, controlled trials (RCTs) and prospective cohort studies on the association of coronary and carotid revascularization, cardiac valve replacement and repair, and ablation for atrial fibrillation on intermediate- and long-term cognitive outcomes in adults aged 65 years or older. We further sought to evaluate whether these associations were modified by procedural and patient characteristics and by procedure-related stroke or transient ischemic attack (TIA).


    We followed a protocol developed with input from the Agency for Healthcare Research Quality (AHRQ) and the Coverage and Analysis Group at the Centers for Medicare & Medicaid Services (CMS) (Supplement). The full technical report, which incorporated AHRQ, CMS, peer, and public review, is available on the CMS Web site (

    Data Sources

    We searched the MEDLINE, Cochrane Library, Scopus, and databases using a combination of search terms for cognitive function, cardiovascular procedures of interest, and study design (Supplement). We also reviewed reference lists of eligible studies, relevant review articles, and articles suggested by experts. We included only studies published from 1990 to January 2015 to reasonably reflect current clinical practice.

    Study Selection

    We included English-language RCTs, controlled clinical trials, and prospective cohort studies. Eligible studies predominately consisted of adults aged 65 years or older. They also reported clinically diagnosed cognitive impairment (such as dementia or mild cognitive impairment) or results of neuropsychological tests (such as Trail Making Test B) or global cognitive screening tests (such as the Mini-Mental State Examination) at least 3 months after coronary or carotid artery revascularization, cardiac valve replacement or repair, or ablation for atrial fibrillation. Clinically diagnosed cognitive impairment must have been based on an assessment of symptoms, function, and formal neuropsychological testing. Treatment group cognitive outcomes must have been compared with those in a control group. The Supplement shows the detailed eligibility criteria. Two reviewers independently examined titles, abstracts, and full articles for eligibility and resolved discrepancies by discussion and consensus.

    Data Extraction and Quality Assessment

    For each article, 1 reviewer extracted details on study design, patient characteristics, cardiovascular procedure characteristics, and cognitive outcomes, and a second reviewer checked accuracy. Using Cochrane Collaboration (for trials) and AHRQ (for prospective cohort studies) criteria (18, 19), 2 reviewers rated individual-study risk of bias as low, moderate, high, or unknown (Appendix Tables 1 and 2). For trials, ratings were based on adequacy of randomization and allocation concealment, masking, accounting for attrition bias, use of intention-to-treat analyses, and selectiveness of outcome reporting. For prospective cohort studies, ratings were based on similarity of prognostic factors between comparison groups, accounting for attrition bias, masking, and selectiveness of outcome reporting. Following AHRQ Effective Health Care Program methods (20), 2 reviewers graded the strength of evidence (SOE) for the association of each treatment comparison with subsequent cognitive outcomes as high, moderate, low, or insufficient based on risk of bias, consistency, precision, and directness (Appendix Table 3). We resolved discrepancies in risk-of-bias ratings and SOE grades by discussion and consensus.

    Appendix Table 1. Risk of Bias in RCTs

    Appendix Table 1.

    Appendix Table 2. Risk of Bias in Prospective Cohort Studies

    Appendix Table 2.

    Appendix Table 3. Strength of Evidence

    Appendix Table 3.

    Appendix Table 3. —Continued

    Appendix Table 3.

    Data Synthesis and Analysis

    We used Review Manager, version 5.2 (Cochrane Collaboration), to estimate relative risks and 95% CIs for the incidence of dichotomous outcomes and standardized mean differences and 95% CIs for continuous outcomes. We did not pool results between studies because no 2 studies had clinically similar patient populations, cardiovascular procedure and control groups, and definitions of cognitive outcomes.

    Role of the Funding Source

    This review was nominated to AHRQ by the CMS and was funded by AHRQ. Staff at AHRQ and CMS helped to develop and refine the scope of the study and reviewed the draft AHRQ report.


    Twenty-five reports of 21 unique studies (17 RCTs and 4 prospective cohort studies) met eligibility criteria (Figure) (21–45). Sixteen of these reports studied CABG; 2 studied cardiac valve replacement, including 1 study of CABG combined with cardiac valve replacement; and 3 studied carotid revascularization.

    Figure. Summary of evidence search and selection.

    * Some references were identified in several databases. Sixty-five additional references were identified by hand searching; 61 of them were excluded at the title and abstract review stage, and 4 were excluded after full-text review. Ninety studies were identified from Among studies not already included in the review, 2 seemed possibly eligible: 1 (NCT02108093) that was still enrolling patients and the other (NCT01743456) that is listed as completed but for which no results appear reported.

    † Included 21 unique studies.

    We rated risk of bias as high in 3 studies, moderate in 11, and unclear in 7. Eighty percent of study patients were men, and mean age was 68 years (mean age was between 65 and 69 years in all but 4 studies). In the 6 studies that reported education, patients completed a mean of 11 years (range, 7 to 14). Few patients had a history of stroke (study range, 0% to 8%; 10 studies) except in the 3 carotid revascularization studies (range, 50% to 100%).

    All studies reported baseline results for individual neuropsychological tests (such as attention, memory, and executive functioning). Five studies excluded patients with baseline dementia (39, 40), abnormal baseline cognitive screening (25, 26, 30, 39, 40), or a learning disorder (36, 40). Among included patients, mean baseline scores in individual studies were similar to estimated age-based norms for IQ and global cognitive screening tests, but more than half of studies reported mean baseline scores in the impaired range for at least 1 neuropsychological test, most frequently with impairment in timed tests.

    All studies reported follow-up results for individual neuropsychological tests. Twelve studies reported incident cognitive impairment as defined by a composite of neuropsychological test results (definitions differed in every study [Appendix Table 4]), but none reported incidence of mild cognitive impairment or dementia. Only 2 studies reported cognitive outcomes more than 1 year after the procedure.

    Appendix Table 4. Individual Study Definitions of Incident Cognitive Impairment Based on Combining Results of Individual Neuropsychological Tests

    Appendix Table 4.

    Coronary Artery Revascularization

    CABG Versus Medical Management

    One prospective cohort study (n = 326) compared on- or off-pump CABG versus medical management in older adults with catheter-proven coronary artery disease (Table 1) (27–29). It found that both the CABG and medical groups had small to moderate improvements in memory versus baseline at 1 year and no changes versus baseline at 6 years for all neuropsychological tests measured. Further, there were no between-group differences at any follow-up or in change from baseline to any follow-up. Risk of bias for this study was high due to important baseline differences between treatment groups and substantial loss to follow-up. The higher dropout rates in the CABG group could have biased the results if they were related to cognitive decline. We rated SOE as insufficient because of these limitations and the absence of corroborating studies.

    Table 1. Coronary Artery Revascularization: Comparison of Intermediate- and Long-Term Postprocedure Cognitive Outcomes Between Treatment Groups

    Table 1.

    Comparisons Between Different Versions of CABG

    On- Versus Off-Pump.

    Five RCTs (n = 2349) (21–26) and 1 prospective cohort study (n = 227) (27–29) compared on- versus off-pump CABG in older adults. On- and off-pump CABG groups had statistically significant within-group improvements versus baseline in most memory tests at 12 months but not at 6 years and no changes versus baseline for other neuropsychological tests or global cognitive screening tests. However, there were no statistically significant between-group differences in incident cognitive impairment at 3 to 12 months after surgery (4 trials reporting) and no consistent between-group differences at follow-up or in change from baseline at any time point for any cognitive test reported (Table 1). We rated SOE as low because, despite several studies with consistent results, CIs were wide, and all studies had at least moderate risk of bias.

    Hypothermia Versus Normothermia.

    In 3 RCTs that randomly assigned older adults to hypothermic or normothermic CABG, we found no statistically significant between-group differences in risk for incident cognitive impairment or changes from baseline to 3-month to 5-year follow-up in any individual neuropsychological test (n = 610) (Table 1) (23, 30–32). In addition, neuropsychological test results did not consistently differ in either treatment group between baseline and follow-up. We rated SOE as low (incident cognitive impairment) to moderate (most neuropsychological tests) because, despite several studies with consistent and mostly precise results, all studies had at least moderate risk of bias.

    Minimal Versus Conventional Extracorporeal Bypass.

    One RCT (n = 64) randomly assigned older adults to on-pump CABG with conventional extracorporeal bypass versus on-pump CABG with minimal extracorporeal bypass (33). Patients with minimal extracorporeal bypass performed significantly better at 3 months on all but 1 neuropsychological test and were less likely to have incident cognitive impairment (21% vs. 61%; risk ratio, 0.34 [95% CI, 0.16 to 0.73]). Because the criterion for incident impairment was lenient (≥1 SD decline in ≥1 of 7 neuropsychological tests), some patients could have achieved it by chance; thus, this outcome is of uncertain clinical importance. We rated SOE as low because, although the magnitude of differences in cognitive outcomes between these groups seemed large and results were precise, data were limited to a single small study.


    In 6 additional RCTs that assigned older adults to different approaches to CABG, cognitive outcomes at 3 to 12 months or change from baseline to 3 to 12 months did not differ between treatment groups assigned to CABG using fentanyl and propofol (n = 180) (34), high- and low-dose fentanyl (n = 350) (35), hyperbaric and atmospheric oxygen (n = 64) (36), cell saver and cardiotomy suction (n = 226) (37), high and low mean arterial blood pressure during bypass (n = 248) (38), or preoperative angiotensin-receptor blocker and preoperative placebo (n = 106) (39). Strength-of-evidence ratings for all of these comparisons were insufficient for incident cognitive impairment. For neuropsychological test outcomes, although studies of all comparisons had at least moderate risk of bias and there were no corroboratory studies, SOE ranged from insufficient to low because estimates around the measure of effect were more precise in some studies (such as fentanyl vs. propofol and high- vs. low-dose fentanyl) than others (such as angiotensin-receptor blocker vs. placebo and hyperbaric vs. atmospheric oxygen).

    Carotid Artery Revascularization

    Carotid Endarterectomy Versus Nonrevascularization Surgical Control

    One prospective cohort study compared cognitive outcomes between older adults with symptomatic or asymptomatic carotid stenosis (≥60%) who had carotid endarterectomy (CEA) (n = 159) and age- and sex-matched control patients without carotid stenosis who had laparoscopic cholecystectomy (n = 68) (40). The symptomatic CEA group had greater improvement than control patients in both Montreal Cognitive Assessment and Mini-Mental State Examination scores at 3 months but only on the Montreal Cognitive Assessment for change between baseline and 12 months (1.0 vs. 0.1; P ≤ 0.01, as reported by authors) (Table 2). These cognitive screening test scores did not significantly change between baseline and 3 or 12 months in the asymptomatic CEA or control groups, and change from baseline did not differ between groups at either follow-up time point.

    Table 2. Carotid Artery Revascularization: Comparison of Intermediate- and Long-Term Postprocedure Cognitive Outcomes Between Treatment Groups

    Table 2.

    CEA Versus Endovascular Carotid Revascularization

    One RCT compared CEA versus carotid artery stenting (CAS) in 177 older adults with recent symptomatic carotid stenosis (>50%) (41). Neuroprotective devices were used in most patients with CEA, and patients with CAS received aspirin and clopidogrel. A sum score of neuropsychological test results was significantly worse at 6 months than at baseline in the CAS group but was unchanged from baseline in the CEA group. However, the mean change from baseline to 6 months either in the sum score or in any individual neuropsychological domain score reported did not significantly differ between groups (Table 2). We rated SOE for these outcomes as insufficient because of moderate risk of bias, wide estimates of effect, and no corroboratory studies.

    One RCT compared CEA versus carotid angioplasty in 46 older adults with recent symptomatic carotid stenosis (>70%) (42). At 6 months, the risk for incident cognitive impairment did not significantly differ between treatment groups (18% vs. 38%; risk ratio, 0.47 [CI, 0.16 to 1.35]), and there were no consistent differences between groups in individual neuropsychological test results (Table 2). There were also no significant between- or within-group changes from baseline to 6 months in any neuropsychological test result. We rated SOE for these outcomes as insufficient because of moderate risk of bias, wide estimates of effect, and no corroboratory studies.

    Cardiac Valve Replacement

    Surgical Aortic Valve Replacement Versus Transcatheter Aortic Valve Replacement

    In 1 prospective cohort study (n = 64), older patients with aortic stenosis who had transapical transcatheter aortic valve replacement (TAVR) were more likely than those who had surgical aortic valve replacement (SAVR) to have cognitive decline at 3 months (28% vs. 6%; P = 0.041) (Table 3) (43). However, results may have been biased because patients having TAVR were older, were less educated, and had much higher surgical risk (log European System for Cardiac Operative Risk Evaluation, 36.0% vs. 2.6%). In addition, the groups were administered different batteries of neuropsychological tests and had cognitive decline defined differently. Neither patients having TAVR nor SAVR had statistically significant within-group changes in any neuropsychological test between baseline and 3 months. Study results were further limited by higher loss to follow-up in the TAVR group and small sample sizes. We rated SOE as insufficient because of high risk of bias and the absence of corroboratory studies.

    Table 3. Cardiac Valve Replacement or Repair: Comparison of Intermediate- and Long-Term Postprocedure Cognitive Outcomes Between Treatment Groups

    Table 3.

    Surgical Cardiac Valve Replacement With or Without CABG Versus CABG Alone

    One prospective cohort study compared cognitive outcomes between patients who had CABG alone (n = 59) and those who had surgical aortic or mitral valve replacement alone or valve replacement combined with CABG (n = 50) (Table 3) (44). At 6 months after surgery, the proportion of patients rated as having a deficit did not significantly differ for 13 of 14 neuropsychological tests reported. We rated SOE as insufficient because of moderate risk of bias, wide estimates of effect, and no corroboratory studies.

    Comparisons Between Different Versions of Surgical Cardiac Valve Replacement

    One RCT (n = 60) compared cognitive outcomes between older adults assigned to SAVR performed under hypothermic and under normothermic conditions (Table 3) (45). Authors reported finding no between-group differences in Mini-Mental State Examination or Trail Making Test A results for up to 4 months after surgery, but they provided no numerical data. They did not report results of comparisons of changes in cognitive testing from baseline between groups. We rated SOE as insufficient because of the unclear risk of bias, absence of analyzable data, and the lack of corroboratory studies.

    Ablation for Atrial Fibrillation

    No eligible study compared ablation versus medical management or any other treatment of atrial fibrillation in older adults and reported intermediate- to long-term cognitive outcomes.

    Association of Incident Stroke or TIA With Cognitive Outcomes After Cardiovascular Procedures

    We could not determine whether intermediate- or long-term cognitive outcomes in older adults after the procedures discussed here are attributable to periprocedural stroke or TIA. First, incident stroke and TIA were uncommon among the 13 studies that reported any data on incident stroke or TIA (that is, n ≤ 1 [and/or ≤4%] in each treatment group in all but 1 study) (21, 22, 24, 26, 27, 33–35, 39–41, 43, 45). Second, incidence of stroke, TIA, and intermediate- and long-term cognitive outcomes did not significantly differ between treatment groups in any eligible study that reported both stroke or TIA and these cognitive outcomes.

    Association of Patient Characteristics With Cognitive Outcomes After Cardiovascular Procedures

    No data from eligible studies addressed whether age, baseline cognitive function, prior stroke or TIA, baseline cardiovascular disease severity, hypertension, diabetes, or depression modify the association between the previously mentioned cardiovascular procedures and intermediate- or long-term cognitive outcomes in older adults.


    Few eligible studies compared intermediate- or long-term cognitive outcomes in older adults having several commonly performed invasive cardiovascular procedures, either versus medical management or versus alternative procedures. The few eligible studies we identified reported little to no difference in cognitive outcomes between or within treatment groups, although conclusions were usually limited by scant and heterogeneous data and low individual-study quality.

    For CABG versus medical management, 1 eligible cohort study with substantial methodological limitations suggested that CABG may be associated with a transient cognitive benefit in older adults and little to no intermediate- to long-term cognitive impairment (27–29). Results also showed no differences in intermediate- or long-term cognitive outcomes between on- versus off-pump CABG treatment groups. These results are generally consistent with those from previous systematic reviews comparing on- versus off-pump CABG that also included younger patients and shorter follow-up (9, 46–48), all but 1 of which reported no between-group difference in cognitive outcomes (48). Results from several trials that compared hypothermic versus normothermic CABG, and 6 more comparisons of 1 version of CABG with another, also found no difference between treatment groups. With all of these studies suggesting that procedure-related factors have little association with cognitive outcomes after CABG in older adults, the results of 1 small RCT that reported significantly better cognitive function after minimal versus conventional extracorporeal bypass CABG may be notable (33). Although the clinical importance of the cognitive outcomes reported by this study is uncertain, the differences between treatment groups seemed large, and a follow-up trial is ongoing (49).

    In older patients with predominately symptomatic carotid stenosis, the scant available data suggested that CEA was not associated with worsened intermediate-term cognitive function and that postprocedure cognitive function did not differ between CEA and CAS or between CEA and carotid angioplasty. These results are consistent with previous systematic reviews that compared cognitive outcomes after and among CEA, CAS, and carotid angioplasty and included younger patients, as well as additional smaller and shorter-term studies (50, 51). Results in some earlier uncontrolled studies that reported cognitive benefits after CEA may in part have reflected testing practice effects (52, 53). Although well-established RCT evidence in older symptomatic adults shows that CEA reduces stroke risk compared with medical management (54) and CAS (55), too few strokes occurred in the studies included in the current review to determine whether any association between carotid revascularization and subsequent cognitive function in older adults was attributable to stroke.

    The single study that compared transapical TAVR and SAVR reported that TAVR was associated with significantly worse intermediate-term cognitive function in older adults. However, this prospective cohort study was biased by substantial differences between treatment groups, including in cognitive risk factors, loss to follow-up, and definition of cognitive impairment. Although previous RCTs of TAVR versus SAVR in high-risk surgical candidates (56) and versus medical management in nonsurgical candidates (57) found that TAVR was associated with a borderline increase in risk for major stroke, it is not clear whether any adverse intermediate-term cognitive effects from TAVR, if present, are due to stroke. There were almost no strokes in the single eligible prospective cohort study, and neither of these earlier RCTs reported cognitive outcomes.

    No eligible studies addressed whether older patients with atrial fibrillation treated with catheter ablation differed in intermediate- or long-term cognitive outcomes compared with no ablation. Results from several recent studies not included in our review (because they were small [58], included younger populations [58, 59], or had a retrospective cohort design [60]), suggest a mixed picture. In 1 study, persons who had ablation for atrial fibrillation had scores similar to those of untreated healthy control participants in most neuropsychological tests (58), whereas 2 studies that used patients with medically treated atrial fibrillation as control participants reported that those who had ablation had higher risk for postprocedure cognitive decline (59) but lower risk for subsequent dementia (60). It is uncertain whether the former results were attributable to chance or the latter findings were affected by misclassification and treatment selection bias. Additional studies are needed to clarify these conflicting results and their applicability to older adults.

    Our findings have several implications for clinical practice. Results suggest that persistent cognitive impairment after the studied cardiovascular procedures may be uncommon or reflect cognitive impairment that was present before the procedure. Although preprocedure cognitive testing could possibly predict subsequent cognitive function, determining whether such screening would improve outcomes was outside the scope of our review. Physicians counseling older patients having the studied cardiovascular procedures should advise that although current data suggest cognitive risks may be small, there is substantial uncertainty about these estimates.

    This review was limited by the available literature. First, although many published studies compare these cardiovascular procedures, for some treatments no eligible studies have reported cognitive outcomes (such as CABG vs. percutaneous coronary intervention; or percutaneous coronary intervention, carotid revascularization, or cardiac valve procedure vs. medical management). For other treatment comparisons, there were few eligible studies, sample sizes were often small, and cognitive outcomes were usually reported in only a subset of persons in both treatment groups, increasing the risk for results to be biased by selective reporting. Second, no studies reported clinically diagnosed cognitive outcomes (such as mild cognitive impairment or dementia) or used the same definition for incident cognitive impairment. Because findings may be sensitive to the manner in which cognitive function is measured and cognitive impairment is defined (61, 62), the clinical importance of any cognitive differences between or within treatment groups was hard to interpret, especially across studies. Third, applicability of the findings of this review to women and patients who may be at highest risk for worsened cognitive impairment is uncertain, including the most elderly patients and those with prior stroke or substantial cognitive impairment, because few of these persons were included in eligible studies (other than for stroke in the carotid revascularization studies) and no studies reported results in these subgroups. We also found no data from eligible studies that addressed whether education level, social support, cardiovascular disease severity, hypertension, diabetes, or depression affect the association between our selected cardiovascular procedures and cognitive outcomes, as had been suggested in predominately uncontrolled studies (8, 11–14).

    Overall, we found that CABG may have little persistent adverse cognitive effect in older adults, and cognitive outcomes seem similar among the CABG approaches and between surgical and endovascular carotid revascularization. Results from 1 study suggesting that transapical TAVR was associated with greater risk for cognitive decline than SAVR were prone to several important sources of bias and were essentially uninterpretable. No eligible studies reported data on cognitive outcomes after percutaneous coronary intervention or catheter ablation of atrial fibrillation in older adults. Physicians discussing risks for persistent cognitive impairment associated with these procedures will need to temper any reassurance with the caveat that the SOE behind this conclusion is mostly low to insufficient.

    Future RCTs and well-controlled prospective cohort studies should compare intermediate- and long-term cognitive outcomes between more versus less invasive or alternative cardiovascular procedures (including medical management) in older adults. Treatment comparisons for which there were the fewest or no eligible studies may be higher priority. Studies should include the most elderly patients and those with prior stroke or baseline cognitive impairment and should report results stratified by or adjusted for these and other potentially important predictors of postprocedural cognitive outcomes. Studies should formally evaluate baseline and postprocedure cognitive function in all randomly assigned patients. Cognitive assessments should include a standardized battery of neuropsychological and functional tests that evaluate a broad set of evidence-based domains, including sensitivity to lateralized deficits and performance validity (63). Calculated outcomes should include a standardized definition of incident cognitive impairment that represents a clinically meaningful change. A consensus process and accumulating intervention data may help further define a cognitive battery that is sensitive to change over time, minimizes learning effects, and will be broadly accepted and implemented. Long-term incidence of clinical outcomes, such as mild cognitive impairment and dementia, should also be reported.


    • 1. Roger VLGo ASLloyd-Jones DMAdams RJBerry JDBrown TMet alAmerican Heart Association Statistics Committee and Stroke Statistics SubcommitteeHeart disease and stroke statistics—2011 update: a report from the American Heart Association. Circulation2011;123:e18-e209. [PMID: 21160056] doi:10.1161/CIR.0b013e3182009701 CrossrefMedlineGoogle Scholar
    • 2. Go ASMozaffarian DRoger VLBenjamin EJBerry JDBorden WBet alAmerican Heart Association Statistics Committee and Stroke Statistics SubcommitteeHeart disease and stroke statistics—2013 update: a report from the American Heart Association. Circulation2013;127:e6-e245. [PMID: 23239837] doi:10.1161/CIR.0b013e31828124ad CrossrefMedlineGoogle Scholar
    • 3. Piccini JPSinner MFGreiner MAHammill BGFontes JDDaubert JPet alOutcomes of Medicare beneficiaries undergoing catheter ablation for atrial fibrillation. Circulation2012;126:2200-7. [PMID: 23019293] doi:10.1161/CIRCULATIONAHA.112.109330 CrossrefMedlineGoogle Scholar
    • 4. Medicare Payment Advisory Commission. The Medicare Advantage program: status report. In: Report to the Congress: Medicare Payment Policy. Washington, DC: Medicare Payment Advisory Commission; 2012:311-35. Accessed at on 17 March 2015. Google Scholar
    • 5. National Center for Health StatisticsHealth, United States, 2012: With Special Feature on Emergency Care. Hyattsville, MD: National Center for Health Statistics; 2013. Google Scholar
    • 6. Coyle RF. Release of County Fee-for-Service Expenditure Data. 7 April 2014. Accessed at on 17 March 2015. Google Scholar
    • 7. Plassman BLLanga KMMcCammon RJFisher GGPotter GGBurke JRet alIncidence of dementia and cognitive impairment, not dementia in the United States. Ann Neurol2011;70:418-26. [PMID: 21425187] doi:10.1002/ana.22362 CrossrefMedlineGoogle Scholar
    • 8. Selnes OAGottesman RFGrega MABaumgartner WAZeger SLMcKhann GMCognitive and neurologic outcomes after coronary-artery bypass surgery. N Engl J Med2012;366:250-7. [PMID: 22256807] doi:10.1056/NEJMra1100109 CrossrefMedlineGoogle Scholar
    • 9. Cormack FShipolini AAwad WIRichardson CMcCormack DJColleoni Let alA meta-analysis of cognitive outcome following coronary artery bypass graft surgery. Neurosci Biobehav Rev2012;36:2118-29. [PMID: 22732162] doi:10.1016/j.neubiorev.2012.06.002 CrossrefMedlineGoogle Scholar
    • 10. Ernest CSMurphy BMWorcester MUHiggins ROElliott PCGoble AJet alCognitive function in candidates for coronary artery bypass graft surgery. Ann Thorac Surg2006;82:812-8. [PMID: 16928490] CrossrefMedlineGoogle Scholar
    • 11. Lombard FWMathew JPNeurocognitive dysfunction following cardiac surgery. Semin Cardiothorac Vasc Anesth2010;14:102-10. [PMID: 20478950] doi:10.1177/1089253210371519 CrossrefMedlineGoogle Scholar
    • 12. Dieleman JSauër AMKlijn CNathoe HMoons KKalkman Cet alPresence of coronary collaterals is associated with a decreased incidence of cognitive decline after coronary artery bypass surgery. Eur J Cardiothorac Surg2009;35:48-53. [PMID: 19027312] doi:10.1016/j.ejcts.2008.10.004 CrossrefMedlineGoogle Scholar
    • 13. Ho PMArciniegas DBGrigsby JMcCarthy MMcDonald GOMoritz TEet alPredictors of cognitive decline following coronary artery bypass graft surgery. Ann Thorac Surg2004;77:597-603. [PMID: 14759444] CrossrefMedlineGoogle Scholar
    • 14. Kadoi YKawauchi CIde MKuroda MTakahashi KSaito Set alPreoperative depression is a risk factor for postoperative short-term and long-term cognitive dysfunction in patients with diabetes mellitus. J Anesth2011;25:10-7. [PMID: 21161290] doi:10.1007/s00540-010-1072-5 CrossrefMedlineGoogle Scholar
    • 15. Goldberg JBGoodney PPKumbhani SRRoth RMPowell RJLikosky DSBrain injury after carotid revascularization: outcomes, mechanisms, and opportunities for improvement. Ann Vasc Surg2011;25:270-86. [PMID: 20926244] doi:10.1016/j.avsg.2010.07.013 CrossrefMedlineGoogle Scholar
    • 16. Miller DCBlackstone EHMack MJSvensson LGKodali SKKapadia Set alPARTNER Trial Investigators and PatientsTranscatheter (TAVR) versus surgical (AVR) aortic valve replacement: occurrence, hazard, risk factors, and consequences of neurologic events in the PARTNER trial. J Thorac Cardiovasc Surg2012;143:832-843. [PMID: 22424519] doi:10.1016/j.jtcvs.2012.01.055 CrossrefMedlineGoogle Scholar
    • 17. Healey JSNair GMDoes catheter ablation for atrial fibrillation increase or reduce neurological insult? Curr Opin Cardiol2012;27:36-40. [PMID: 22123602] doi:10.1097/HCO.0b013e32834d846a CrossrefMedlineGoogle Scholar
    • 18. Higgins JPAltman DGGøtzsche PCJüni PMoher DOxman ADet alCochrane Bias Methods GroupThe Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ2011;343:d5928. [PMID: 22008217] doi:10.1136/bmj.d5928 CrossrefMedlineGoogle Scholar
    • 19. Viswanathan MAnsari MTBerkman NDChang SHartling LMcPheeters Met alAssessing the Risk of Bias of Individual Studies in Systematic Reviews of Health Care Interventions. Methods Guide for Comparative Effectiveness Reviews. Rockville, MD: Agency for Healthcare Research and Quality; 2012. Google Scholar
    • 20. Owens DKLohr KNAtkins DTreadwell JRReston JTBass EBet alAHRQ series paper 5: grading the strength of a body of evidence when comparing medical interventions—Agency for Healthcare Research and Quality and the Effective Health Care Program. J Clin Epidemiol2010;63:513-23. [PMID: 19595577] doi:10.1016/j.jclinepi.2009.03.009 CrossrefMedlineGoogle Scholar
    • 21. Lamy ADevereaux PJPrabhakaran DTaggart DPHu SPaolasso Eet alCORONARY InvestigatorsEffects of off-pump and on-pump coronary-artery bypass grafting at 1 year. N Engl J Med2013;368:1179-88. [PMID: 23477676] doi:10.1056/NEJMoa1301228 CrossrefMedlineGoogle Scholar
    • 22. Vedin JNyman HEricsson AHylander SVaage JCognitive function after on or off pump coronary artery bypass grafting. Eur J Cardiothorac Surg2006;30:305-10. [PMID: 16828295] CrossrefMedlineGoogle Scholar
    • 23. Lund CSundet KTennøe BHol PKRein KAFosse Eet alCerebral ischemic injury and cognitive impairment after off-pump and on-pump coronary artery bypass grafting surgery. Ann Thorac Surg2005;80:2126-31. [PMID: 16305857] CrossrefMedlineGoogle Scholar
    • 24. Lee JDLee SJTsushima WTYamauchi HLau WTPopper Jet alBenefits of off-pump bypass on neurologic and clinical morbidity: a prospective randomized trial. Ann Thorac Surg2003;76:18-25. [PMID: 12842506] CrossrefMedlineGoogle Scholar
    • 25. Jensen BOHughes PRasmussen LSPedersen PUSteinbrüchel DACognitive outcomes in elderly high-risk patients after off-pump versus conventional coronary artery bypass grafting: a randomized trial. Circulation2006;113:2790-5. [PMID: 16769915] CrossrefMedlineGoogle Scholar
    • 26. Jensen BØRasmussen LSSteinbrüchel DACognitive outcomes in elderly high-risk patients 1 year after off-pump versus on-pump coronary artery bypass grafting.A randomized trial. Eur J Cardiothorac Surg2008;34:1016-21. [PMID: 18778948] doi:10.1016/j.ejcts.2008.07.053 CrossrefMedlineGoogle Scholar
    • 27. McKhann GMSelnes OAGrega MABailey MMPham LDBaumgartner WAet alSubjective memory symptoms in surgical and nonsurgical coronary artery patients: 6-year follow-up. Ann Thorac Surg2009;87:27-34. [PMID: 19101263] doi:10.1016/j.athoracsur.2008.09.023 CrossrefMedlineGoogle Scholar
    • 28. Selnes OAGrega MABorowicz LMBarry SZeger SBaumgartner WAet alCognitive outcomes three years after coronary artery bypass surgery: a comparison of on-pump coronary artery bypass graft surgery and nonsurgical controls. Ann Thorac Surg2005;79:1201-9. [PMID: 15797050] CrossrefMedlineGoogle Scholar
    • 29. Selnes OAZeger SLCoronary artery bypass grafting baseline cognitive assessment: essential not optional [Editorial]. Ann Thorac Surg2007;83:374-6. [PMID: 17257951] CrossrefMedlineGoogle Scholar
    • 30. Boodhwani MRubens FWozny DRodriguez RNathan HJEffects of sustained mild hypothermia on neurocognitive function after coronary artery bypass surgery: a randomized, double-blind study. J Thorac Cardiovasc Surg2007;134:1443-50. [PMID: 18023662] CrossrefMedlineGoogle Scholar
    • 31. Nathan HJRodriguez RWozny DDupuis JYRubens FDBryson GLet alNeuroprotective effect of mild hypothermia in patients undergoing coronary artery surgery with cardiopulmonary bypass: five-year follow-up of a randomized trial. J Thorac Cardiovasc Surg2007;133:1206-11. [PMID: 17467430] CrossrefMedlineGoogle Scholar
    • 32. Nathan HJWells GAMunson JLWozny DNeuroprotective effect of mild hypothermia in patients undergoing coronary artery surgery with cardiopulmonary bypass: a randomized trial. Circulation2001;104:I85-91. [PMID: 11568036] CrossrefMedlineGoogle Scholar
    • 33. Anastasiadis KAntonitsis PPapakonstantinou CWhen is the optimal time to perform neurocognitive assessment after coronary artery bypass surgery? [Letter]. Ann Thorac Surg2011;92:1933. [PMID: 22051305] doi:10.1016/j.athoracsur.2011.03.074 CrossrefMedlineGoogle Scholar
    • 34. Kadoi YSaito SKunimoto FGoto FFujita NComparative effects of propofol versus fentanyl on cerebral oxygenation state during normothermic cardiopulmonary bypass and postoperative cognitive dysfunction. Ann Thorac Surg2003;75:840-6. [PMID: 12645704] CrossrefMedlineGoogle Scholar
    • 35. Silbert BSScott DAEvered LALewis MSKalpokas MMaruff Pet alA comparison of the effect of high- and low-dose fentanyl on the incidence of postoperative cognitive dysfunction after coronary artery bypass surgery in the elderly. Anesthesiology2006;104:1137-45. [PMID: 16732083] CrossrefMedlineGoogle Scholar
    • 36. Alex JLaden GCale ARBennett SFlowers KMadden Let alPretreatment with hyperbaric oxygen and its effect on neuropsychometric dysfunction and systemic inflammatory response after cardiopulmonary bypass: a prospective randomized double-blind trial. J Thorac Cardiovasc Surg2005;130:1623-30. [PMID: 16308008] CrossrefMedlineGoogle Scholar
    • 37. Djaiani GKatznelson RFedorko LRao VGreen RCarroll Jet alEarly benefit of preserved cognitive function is not sustained at one-year after cardiac surgery: a longitudinal follow-up of the randomized controlled trial. Can J Anaesth2012;59:449-55. [PMID: 22290354] doi:10.1007/s12630-012-9675-y CrossrefMedlineGoogle Scholar
    • 38. Gold JPCharlson MEWilliams-Russo PSzatrowski TPPeterson JCPirraglia PAet alImprovement of outcomes after coronary artery bypass. A randomized trial comparing intraoperative high versus low mean arterial pressure. J Thorac Cardiovasc Surg1995;110:1302-11. [PMID: 7475182] CrossrefMedlineGoogle Scholar
    • 39. Flesch MKnipp SKessler GGeissler HJMassoudy PWilhelm Het alARTA: AT1-receptor blocker therapy in patients undergoing coronary artery bypass grafting. Clin Res Cardiol2009;98:33-43. [PMID: 18853093] doi:10.1007/s00392-008-0719-9 CrossrefMedlineGoogle Scholar
    • 40. Baracchini CMazzalai FGruppo MLorenzetti RErmani MBallotta ECarotid endarterectomy protects elderly patients from cognitive decline: a prospective study. Surgery2012;151:99-106. [PMID: 21943640] doi:10.1016/j.surg.2011.06.031 CrossrefMedlineGoogle Scholar
    • 41. Altinbas AvanZandvoort MJvan den Berg EJongen LMAlgra AMoll FLet alCognition after carotid endarterectomy or stenting: a randomized comparison. Neurology2011;77:1084-90. [PMID: 21880992] doi:10.1212/WNL.0b013e31822e55b9 CrossrefMedlineGoogle Scholar
    • 42. Crawley FStygall JLunn SHarrison MBrown MMNewman SComparison of microembolism detected by transcranial Doppler and neuropsychological sequelae of carotid surgery and percutaneous transluminal angioplasty. Stroke2000;31:1329-34. [PMID: 10835452] CrossrefMedlineGoogle Scholar
    • 43. Knipp SCKahlert PJokisch DSchlamann MWendt DWeimar Cet alCognitive function after transapical aortic valve implantation: a single-centre study with 3-month follow-up. Interact Cardiovasc Thorac Surg2013;16:116-22. [PMID: 23148084] doi:10.1093/icvts/ivs461 CrossrefMedlineGoogle Scholar
    • 44. Andrew MJBaker RABennetts JKneebone ACKnight JLA comparison of neuropsychologic deficits after extracardiac and intracardiac surgery. J Cardiothorac Vasc Anesth2001;15:9-14. [PMID: 11254832] CrossrefMedlineGoogle Scholar
    • 45. Fakin RZimpfer DSodeck GHRajek AMora BDumfarth Jet alInfluence of temperature management on neurocognitive function in biological aortic valve replacement. A prospective randomized trial. J Cardiovasc Surg (Torino)2012;53:107-12. [PMID: 22231536] MedlineGoogle Scholar
    • 46. Marasco SFSharwood LNAbramson MJNo improvement in neurocognitive outcomes after off-pump versus on-pump coronary revascularisation: a meta-analysis. Eur J Cardiothorac Surg2008;33:961-70. [PMID: 18424064] doi:10.1016/j.ejcts.2008.03.022 CrossrefMedlineGoogle Scholar
    • 47. Kennedy EDChoy KCAlston RPChen SFarhan-Alanie MMAnderson Jet alCognitive outcome after on- and off-pump coronary artery bypass grafting surgery: a systematic review and meta-analysis. J Cardiothorac Vasc Anesth2013;27:253-65. [PMID: 23507014] doi:10.1053/j.jvca.2012.11.008 CrossrefMedlineGoogle Scholar
    • 48. Sun JHWu XYWang WJJin LLCognitive dysfunction after off-pump versus on-pump coronary artery bypass surgery: a meta-analysis. J Int Med Res2012;40:852-8. [PMID: 22906257] CrossrefMedlineGoogle Scholar
    • 49. Anastasiadis K. Comparison of minimal versus conventional extracorporeal circulation in coronary surgery. Accessed at on 26 May 2015. Google Scholar
    • 50. Paraskevas KILazaridis CAndrews CMVeith FJGiannoukas ADComparison of cognitive function after carotid artery stenting versus carotid endarterectomy. Eur J Vasc Endovasc Surg2014;47:221-31. [PMID: 24393665] doi:10.1016/j.ejvs.2013.11.006 CrossrefMedlineGoogle Scholar
    • 51. De Rango PCaso VLeys DPaciaroni MLenti MCao PThe role of carotid artery stenting and carotid endarterectomy in cognitive performance: a systematic review. Stroke2008;39:3116-27. [PMID: 18723423] doi:10.1161/STROKEAHA.108.518357 CrossrefMedlineGoogle Scholar
    • 52. Lunn SCrawley FHarrison MJBrown MMNewman SPImpact of carotid endarterectomy upon cognitive functioning. A systematic review of the literature. Cerebrovasc Dis1999;9:74-81. [PMID: 9973649] CrossrefMedlineGoogle Scholar
    • 53. Irvine CDGardner FVDavies AHLamont PMCognitive testing in patients undergoing carotid endarterectomy. Eur J Vasc Endovasc Surg1998;15:195-204. [PMID: 9587331] CrossrefMedlineGoogle Scholar
    • 54. Rerkasem KRothwell PMCarotid endarterectomy for symptomatic carotid stenosis. Cochrane Database Syst Rev2011:CD001081. [PMID: 21491381] doi:10.1002/14651858.CD001081.pub2 CrossrefMedlineGoogle Scholar
    • 55. Bonati LHLyrer PEderle JFeatherstone RBrown MMPercutaneous transluminal balloon angioplasty and stenting for carotid artery stenosis. Cochrane Database Syst Rev2012;9:CD000515. [PMID: 22972047] doi:10.1002/14651858.CD000515.pub4 CrossrefMedlineGoogle Scholar
    • 56. Smith CRLeon MBMack MJMiller DCMoses JWSvensson LGet alPARTNER Trial InvestigatorsTranscatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med2011;364:2187-98. [PMID: 21639811] doi:10.1056/NEJMoa1103510 CrossrefMedlineGoogle Scholar
    • 57. Leon MBSmith CRMack MMiller DCMoses JWSvensson LGet alPARTNER Trial InvestigatorsTranscatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med2010;363:1597-607. [PMID: 20961243] doi:10.1056/NEJMoa1008232 CrossrefMedlineGoogle Scholar
    • 58. Schwarz NKuniss MNedelmann MKaps MBachmann GNeumann Tet alNeuropsychological decline after catheter ablation of atrial fibrillation. Heart Rhythm2010;7:1761-7. [PMID: 20691284] doi:10.1016/j.hrthm.2010.07.035 CrossrefMedlineGoogle Scholar
    • 59. Medi CEvered LSilbert BTeh AHalloran KMorton Jet alSubtle post-procedural cognitive dysfunction after atrial fibrillation ablation. J Am Coll Cardiol2013;62:531-9. [PMID: 23684686] doi:10.1016/j.jacc.2013.03.073 CrossrefMedlineGoogle Scholar
    • 60. Bunch TJCrandall BGWeiss JPMay HTBair TLOsborn JSet alPatients treated with catheter ablation for atrial fibrillation have long-term rates of death, stroke, and dementia similar to patients without atrial fibrillation. J Cardiovasc Electrophysiol2011;22:839-45. [PMID: 21410581] doi:10.1111/j.1540-8167.2011.02035.x CrossrefMedlineGoogle Scholar
    • 61. Lewis MSMaruff PSilbert BSEvered LAScott DADetection of postoperative cognitive decline after coronary artery bypass graft surgery is affected by the number of neuropsychological tests in the assessment battery. Ann Thorac Surg2006;81:2097-104. [PMID: 16731137] CrossrefMedlineGoogle Scholar
    • 62. Keizer AMHijman RKalkman CJKahn RSvan Dijk DOctopus Study GroupThe incidence of cognitive decline after (not) undergoing coronary artery bypass grafting: the impact of a controlled definition. Acta Anaesthesiol Scand2005;49:1232-5. [PMID: 16146457] CrossrefMedlineGoogle Scholar
    • 63. Larrabee GJTest validity and performance validity: considerations in providing a framework for development of an ability-focused neuropsychological test battery. Arch Clin Neuropsychol2014;29:695-714. [PMID: 25280794] doi:10.1093/arclin/acu049 CrossrefMedlineGoogle Scholar


    Sign In to Submit A Comment