Study Suggests Routine Bridging Harmful for A-fib Patients Interrupting Anticoagulation

Although relatively common, use of “bridging” medication to protect A-fib patients during the interruption of anticoagulation for procedures is associated with a greater risk of adverse events, especially bleeding, according to a registry study published online December 12, 2014, ahead of print in Circulation.

Take Home: Study Suggests Routine Bridging Harmful for A-fib Patients Interrupting Anticoagulation“These findings fly somewhat in the face of conventional dogma and may begin a paradigm shift away from the routine use of a bridging strategy,” says an editorial accompanying the paper.

Investigators led by Benjamin A. Steinberg, MD, MHS, of Duke University Medical Center (Durham, NC), analyzed data on 7,372 patients treated with oral anticoagulants who were enrolled in ORBIT-AF, a national outpatient A-fib registry.

Over a median follow-up of 2 years, 30% of patients had at least 1 therapy interruption. The most common reasons for the 2,803 total interruptions were to undergo noncardiac surgery (27%), other procedures (25%), or endoscopy (18%). About one-quarter of the interruptions (24%) were bridged with other medications, including low molecular weight heparin (73%), unfractionated heparin (15%), fondaparinux (1.1%; Arixtra, GlaxoSmithKline), or another anticoagulant (11%). Although the vast majority of interruptions involving bridging occurred in patients treated with warfarin at baseline, 23 involved patients taking the novel oral anticoagulant dabigatran (Pradaxa; Boehringer Ingelheim).

Compared with patients who did not interrupt anticoagulation, those with at least 1 interruption were slightly younger, more likely to be white, and less likely to have new-onset A-fib, and had higher creatinine clearance. Rates of prior coronary vascular and cerebrovascular disease and mean CHADS2 scores were similar between the groups.

Bridged Patients at Higher Thrombotic Risk

Compared with patients with no bridging, those who received any bridging were younger and more likely to have congestive heart failure, prior cerebrovascular events, any valve disease, and a prior mechanical valve. In addition, dabigatran was less likely to be the baseline anticoagulant in those who received bridging. Though the mean CHADS2 and CHA2DS2-VASc scores were higher in bridged patients, there were no differences in rates of CHADS2 or CHA2DS2-VASc scores of at least 2. Use of additional antiplatelet therapy did not differ by bridging status.

Among warfarin-treated patients who had at least 1 follow-up INR assessment following the procedure, the median time to achievement of therapeutic range (first INR 2) after the procedure was shorter for interruptions using than for those forgoing bridging (17 vs 23 days; P < .001). 

More Complications With Bridging During, After Interruption

Overall, complications during anticoagulation interruption were relatively uncommon, but rates of any adverse event were higher when bridging was used (5.3% vs 2.8%; P = .01), as were those of major bleeding (3.6% vs 1.2%; P = .0007), bleeding hospitalization (2.2% vs 0.7%; P = .006), and cardiovascular hospitalization (4.2% vs 2.2%; P = .02). 

At 30 days, the association between bridging and bleeding events (major bleeding or bleeding hospitalization) and overall composite events (stroke or systemic embolism, MI, major bleeding, hospitalization, or death) persisted after adjustment. In addition, there was a trend toward increased cardiovascular events (stroke or systemic embolism, MI, cardiovascular hospitalization; table 1).

Table 1. Adjusted 30-Day Outcomes for Bridging vs No Bridging 

In the adjusted model, the type of baseline anticoagulant used (warfarin vs dabigatran) did not affect outcomes following temporary interruption.

Sensitivity analysis showed that the link between bridging and adverse outcomes was consistent across various baseline concomitant antiplatelet regimens (none, single, or double).

According to the authors, anticoagulation bridging appears to be used more commonly than would be suggested by the most recent US national guidelines, which “highlight the dearth of evidence for the practice.” Furthermore, they observe, there is “mounting evidence” that certain procedures may be performed more safely without interruption of anticoagulation.

Risk Prediction Challenging

In their editorial,Amir Y. Shaikh, MD, and David D. McManus, MD, ScM, of the University of Massachusetts Medical School (Worcester, MA), note that management of these patients is usually based on assessment of their individual risks of thromboembolism and perioperative bleeding and the type of procedure they need. But because there are no validated risk stratification schemes specific to periprocedural anticoagulation decision-making, management varies widely.

Complicating the situation, they add, is the fact that the timing of discontinuation and resumption of anticoagulation and differences in the type of bridging agent are “areas where considerable uncertainty and practice variation remains.”

Interestingly, the editorialists observe, even patients with higher stroke risk based on CHADS2 or CHA2DS2-VASc scores or presence of a mechanical heart valve did not have more thromboembolic events, suggesting that the risk prediction tools validated for ambulatory A-fib populations may not apply to short-term periprocedural risk.

Will the Novel Oral Anticoagulants Affect Management?

An important limitation of the current study, Drs. Shaikh and McManus point out, is the absence of information on whether agents were used to reverse anticoagulation or on the timing of bridging medications, all factors that may contribute to bleeding complications.

Also, further study is needed to compare bleeding risks for novel oral anticoagulants and warfarin, they write. The new drugs have a short time to onset and are cleared relatively quickly, which probably makes bridging redundant, they note.

Despite the contribution of the current analysis, Drs. Shaikh and McManus say future randomized studies should examine questions such as whether outcomes differ depending on the type of chronic anticoagulant taken, the type of short-acting bridging agent used, or the timing of initiation and discontinuation of bridging anticoagulation. Results of the large, randomized BRIDGE and PERIOP-2 studies should help inform periprocedural decision making, they predict.


1. Steinberg BA, Peterson ED, Kim S, et al. Use and outcomes associated with bridging during anticoagulation interruptions in patients with atrial fibrillation: findings from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF). Circulation. 2014;Epub ahead of print.

2. Shaikh AY, McManus DD. A bridge too far? Findings of bridging anticoagulation use and outcomes in the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF). Circulation. 2014;Epub ahead of print.

Related Stories:


Study Suggests Routine Bridging Harmful for A-fib Patients Interrupting Anticoagulation

  • The ORBIT-AF registry is sponsored by Janssen Scientific Affairs.
  • Drs. Steinberg and Shaikh report no relevant conflicts of interest.
  • Dr. McManus reports receiving a University of Massachusetts Center for Clinical and Translational Science award and grant support from the National Heart, Lung and Blood Institute, Biotronik, Medtronic, Philips Healthcare, and Sanofi Aventis.