Author + information
- William F. McIntyre, MD∗ (, )
- Jeff S. Healey, MD, MSc,
- P.J. Devereaux, MD, PhD and
- David Conen, MD, MPH
- ↵∗Population Health Research Institute, McMaster University, C3-13A DBCVSRI, 20 Copeland Avenue, Hamilton, Ontario L8L 2X2, Canada
There is uncertainty regarding the long-term prognosis and management of perioperative atrial fibrillation (POAF) that is first detected during hospitalization for noncardiac surgery (1,2). We commend Butt et al. (1) for their recent paper undertaking an important analysis in this field.
The incidence of POAF (0.4%) in the paper by Butt et al. (1) was lower than in other POAF studies, including studies using administrative data (2,3). This could mean that the authors may have missed a significant number of shorter, potentially asymptomatic POAF episodes. On the other hand, the long-term risk of thromboembolism in POAF patients (31.7 events per 1,000 person-years) was higher than in previous studies (3). This could be explained by the selection of higher-risk patients, or by the outcome choice of thromboembolic events—other similar studies have assessed stroke. Finally, although it is tempting to assume that the risk of thromboembolism in the POAF group was secondary to ongoing or recurrent AF, this is unclear. For example, the POISE (Perioperative Ischemic Evaluation) trial demonstrated that POAF was independently associated with an increased stroke risk at 30 days after surgery, and metoprolol compared with placebo decreased the risk of POAF. Despite this, metoprolol significantly increased the risk of stroke (4).
There is also some concern about the author’s findings of similar risks of thromboembolism between POAF and nonvalvular atrial fibrillation (NVAF). The authors defined NVAF as new-onset AF occurring during hospitalization, and 34.6% of patients had AF as a “secondary” diagnosis, with pneumonia being the most common “primary” diagnosis. Thus, the NVAF comparator group may also have included patients with lower-risk AF, and this may have diluted the risk of thromboembolism in this group. The fact that rates of thromboembolism were similar in the POAF and NVAF groups despite nearly double the use of oral anticoagulation (OAC) in the NVAF group also suggests that the POAF group may have had a lower baseline risk of stroke.
Some physicians may assume that the effects of OAC in patients with NVAF apply to patients with POAF. However, in the absence of randomized controlled trials, we caution clinicians against the routine use of OAC for patients with new-onset, potentially transient POAF. History provides many examples where assumptions regarding extending treatment efficacy have failed in what were considered similar patient populations (5). We strongly agree with the authors that patients with POAF should be seen as a unique population in whom randomized controlled trials should evaluate the risks and benefits of OAC.
Please note: Dr. Healey has received research grants from Bristol-Myers Squibb and Pfizer. Dr. Devereaux has received research grants from Abbott Diagnostics, Boehringer Ingelheim, Philips Healthcare, and Roche Diagnostics. Drs. McIntyre and Conen have received speaker fees from Servier Canada.
- 2019 American College of Cardiology Foundation
- Butt J.H.,
- Olesen J.B.,
- Havers-Borgersen E.,
- et al.
- McIntyre W.F.,
- Connolly S.J.,
- Healey J.S.
- The TIMI IIIB Investigators