To summarize current evidence on intermediate- and long-term cognitive outcomes after coronary and carotid revascularization, cardiac valve procedures, and ablation for atrial fibrillation in older adults, and their association with procedure-related stroke, transient ischemic attack (TIA), and other procedure and patient characteristics.
MEDLINE, Cochrane Database of Systematic Reviews, Scopus, and ClinicalTrials.gov electronic databases from 1990 through July 2014; hand searches of references from relevant reviews and eligible studies.
We screened abstracts and full-text articles of identified references for randomized controlled trials (RCTs) and prospective cohort studies in adults aged ≥65 years that reported intermediate (3 to 12 months) and/or long-term (>12 months) cognitive outcomes after one or more of the above selected cardiovascular procedures. Cognitive outcomes of interest were clinical diagnoses (e.g., mild cognitive impairment), neuropsychological test results, and incident cognitive impairment derived from a composite of neuropsychological test results. We extracted data, rated individual study risk of bias, and graded strength of evidence (SOE).
Seventeen RCTs and 4 prospective cohort studies were included. Eighty percent of participants were male and mean age was 68 years. Five studies excluded participants for some measure of abnormal baseline cognition. Nevertheless, more than half of studies reported mean baseline scores in the impaired range for at least one neuropsychological test, most frequently with slowing in timed tests. There was no significant difference in post-procedure cognitive function between on- versus off-pump coronary artery bypass grafting (CABG) (n=6) (low SOE), hypothermic versus normothermic CABG (n=3) (moderate to low SOE), or CABG versus medical management (n=1) (insufficient SOE). One trial reported lower risk of incident cognitive impairment at 3 months with minimal versus conventional extracorporeal bypass CABG (RR=0.34 [95%CI=0.16–0.73]). Two trials found no difference between surgical and endovascular carotid revascularization (low to insufficient SOE). One cohort study reported increased cognitive decline after transcatheter versus surgical aortic valve replacement at 3 months (28% versus 6 %, p=0.04), but results may have been limited by large selection and outcome measurement biases (insufficient SOE). Because study participants had few strokes and transient ischemic attacks, we could not determine whether these events affected post-procedure cognitive outcomes. We found no evidence from eligible studies about whether patient characteristics such as age and baseline cognitive function modify the association between these cardiovascular procedures and intermediate- or long-term post-procedure cognitive outcomes. This review was limited by the small number of eligible studies for each treatment comparison, including no eligible studies that assessed cognitive outcomes after ablation for atrial fibrillation; heterogeneity of cognitive outcomes; and limited individual study quality. Results may have limited generalizability to the elderly, women, or individuals with substantial baseline cognitive impairment.
Persistent cognitive impairment attributable to studied cardiovascular procedures in older adults appeared uncommon and may reflect pre-existing cognitive impairment. Specifically, CABG may have little intermediate to long-term cognitive effect in older adults, including numerous comparisons of different versions of CABG versus each other. Intermediate-term cognitive effects may be similar between those who undergo surgical versus endovascular carotid revascularization. Results suggesting better cognition after minimal versus conventional extracorporeal bypass CABG are promising but need confirmation. Confidence in review findings should be tempered by substantial limitations in primary data quantity and quality. Results may not be generalizable to old-old patients, to women, or to patients with substantial baseline cognitive impairment.
