Page 79 - Read Online

Page 4 of 9 Rademacher et al. Vessel Plus 2022;6:39 https://dx.doi.org/10.20517/2574-1209.2021.138

Table 1. Unadjusted estimates: Reported unadjusted estimates of association between POAF and stroke in eight separate meta-
analyses. Characteristics of pooled primary studies used to obtain these estimates are also described
Included Total Pooled estimate & Average time to follow-up for
Year & Name Population
studies subjects CI creating estimate
[11]
Hernandez et al. , CABG, Valve, CABG + 10 17,127 OR 1.77 (95%CI: not reported
2013 Valve 1.36-2.31)
[12]
Megens et al. , 2017 CABG, CABG + Valve 13 108,711 ES 1.36 (95%CI: 2.05 years
1.12-1.65)
Lin et al. [13] , 2019 CABG, Valve, CABG + 6 33,319 OR 1.62 (95%CI: 30 days
Valve 1.47-1.80)
Kerwin et al. [14] , 2020 CABG 12 89,498 OR 2.15 (95%CI: not reported
1.82-2.53)
Woldendorp et al. [15] , CABG, Valve, Other 61 239,018 OR 2.29 (95%CI: 30 days
2020 Cardiac 1.97-2.66)
Caldonazo et al. [18] , Any cardiac surgery 57 246,340 OR 2.17 (95%CI: 30 days
2021 1.90-2.49)
Chau et al. [17] , 2021 CABG 3 28,003 RR 1.57 (95%CI: 4.79 years
1.27-1.95)
[16]
Eikelboom et al. , CABG, CABG + Valve 4 9,072 OR 4.09 (95%CI: 6.1 years
2021 2.49-6.72)

the arguably most comprehensive analysis of the association of POAF and stroke in patients undergoing
cardiac surgery. The unadjusted, pooled OR for stroke during follow-up in this study is 2.29 (1.97-2.66) and
derived from 61 studies. No confounder-adjusted estimate was reported in this meta-analysis. A 2021 study
by Eikelboom et al. reported an unadjusted, pooled OR estimate from 4 studies on patients undergoing
[16]
CABG and/or valve procedures, that is significantly higher [4.09 (2.49-6.72)] compared to the other meta-
analyses. And Chau et al. (2021) included 3 studies of patients undergoing CABG and reported a pooled
[17]
OR estimate of 1.57 (1.27-1.95). Finally, Caldonazo et al. reported their pooled estimate for the unadjusted
[18]
OR for stroke in patients with POAF at 2.17 (1.90-2.49), including 57 studies in patients undergoing cardiac
surgery.

Several published studies on POAF and stroke were not included in the meta-analyses cited above. A
notable example is a report from the SWEDHEART registry on this topic. SWEDHEART is an
observational, nationwide, population-based, longitudinal registry-based cohort study including all Swedish
residents > 18 years receiving cardiovascular care. This study included 24,523 patients with no previous
history of atrial fibrillation, no taking anticoagulation preoperatively, undergoing first-time CABG. Thirty
percent developed POAF (within 30 days of the index procedure) . At 5 years follow-up, POAF was
[19]
associated with an increased risk of ischemic stroke [HR 1.18 (1.05-1.32)] after adjustment for variables
differentiating them from the group without POAF.

Evidence on anticoagulation in POAF to prevent stroke
The evidence for anticoagulation in the setting of POAF is relatively sparse. We identified 5 studies
evaluating the impact of anticoagulation on stroke in the setting of POAF [Table 2] [20-24] . Of these, 3 studies
had results that supported the use of anticoagulants [20-22] , and two studies had inconclusive evidence or noted
no adverse effects of anticoagulation [23,24] .

Current guidelines suggest using anticoagulation in patients with POAF are based primarily on one study
[9]
(new-onset atrial fibrillation predicts long-term mortality after coronary artery bypass graft) and the
[25]
experience in patients with atrial fibrillation in general.