Implications of new-onset atrial fibrillation for in-hospital and long-term prognosis in patients with acute myocardial infarction: the CBD Bank study

Abstract Background Atrial fibrillation (AF) often coexists with coronary artery disease. Data on the incidence and prognostic impact of new-onset AF following acute myocardial infarction (AMI) with current optimal therapy are insufficient, especially in Asian populations. Purpose To investigate the incidence of new-onset AF following AMI and to assess its impact on in-hospital and long-term prognosis. Methods We included consecutive AMI patients between December 2012 and July 2019, and excluded those with prior known AF on presentation. New-onset AF was defined as newly detected AF during the index hospitalization following AMI. The primary outcomes comprised of all-cause death and cardiovascular death occurred during hospitalization; and all-cause death and cardiovascular death during long-term follow-up among those AMI survivors. Follow-up visits were routinely scheduled after discharge, at 1 month, 3 months, 6 months, 12 months and every 12 months thereafter. Results Of 3686 patients enrolled, new-onset AF was documented in 138 (3.7%) patients during a mean duration of hospitalization of 8.8±5.8 days. Independent risk factors of new-onset AF were age ≥75 years, left atrial diameter ≥40mm, high levels of cardiac troponin-I or high sensitive C reactive protein. During hospitalization, all-cause death occurred in 22 (15.9%) new-onset AF patients and 67 (1.9%) non-AF patients (p<0.001); cardiovascular death occurred in 19 (13.8%) new-onset AF patients and 58 (1.6%) non-AF patients (p<0.001). On multivariable logistic analysis, new-onset AF was an independent predictor of in-hospital all-cause death (OR 5.85, 95% CI: 3.24–10.55) and cardiovascular death (OR 5.44, 95% CI: 2.90–10.20). Apart from the in-hospital deaths, another 265 (7.7%) were lost to follow-up; thus, 3332 patients were included in the long-term follow-up analysis: 106 new-onset AF and 3226 non-AF patients. After a mean follow-up period of 1096.7±682.0 days, all-cause death occurred in 19 new-onset AF patients and 249 non-AF patients; corresponding rates were 8.08 (95% CI: 5.15–12.67) vs. 2.55 (95% CI: 2.25, 2.88) per 100 person-years, respectively (p<0.001). Cardiovascular death occurred in 11 new-onset AF patients and 150 non-AF patients; corresponding rates were 4.68 (95% CI: 2.59–8.45) vs. 1.53 (95% CI: 1.31–1.80) per 100 person-years, respectively (p=0.002). After multivariable Cox adjustment, there was no significant association between new-onset AF and long-term all-cause death (HR 1.45, 95% CI: 0.90–2.35) or cardiovascular death (HR 1.21, 95% CI: 0.65–2.26). Conclusion New-onset AF following AMI was an independent predictor of increased risk of in-hospital mortality, but had no independent association with long-term death. Funding Acknowledgement Type of funding source: None