Summary

Primary Question

In patients with afib without mechanical valves, is no bridging noninferior with respect to thromboembolism risk and superior with respect to bleeding risk when compared to perioperative bridging with dalteparin (LMWH)?

Conclusion

In patients with Afib and a CHADS₂ ≤ 3 (or maybe 4) without a mechanical, cancer, bridging before and after surgery appears to increase the risk of bleeding without modifying the risk of stroke.

PICO

Population

Individuals with Afib, including valvular Afib, but not those with mechanical valves. Populations were skewed towards white males, with CHADS₂ Scores ≤ 3-4. Cancer was not well represented in this trial. Nor were stroke / TIA Hx.

Intervention / Comparison

D/C warfarin on Day (-5), Dalteparin or placebo on Day (-3), then restarting Dalteparin or placebo after the procedure until INR was therapeutic.

Outcome(s)

Primary: Arterial thromboembolism (stroke, TIA, systemic embolism) at 37d post-op

​ Non-inferior, not superior

Secondary: Major bleeding, death, acute MI, DVT, PE, minor bleeding

​ No bridging was superior for Major Bleeding, minor bleeding was a superior exploratory endpoint, others not significantly different.

Randomize Trial Checklist

The study population included or focused on those in the relevant specialty? (i.e. the ER / ICU)

Relevant to surgical patients with afib without cancer or a mechanical valve

Adequate Randomization?

Yes

Randomization was concealed?

Yes

Selection bias / consecutive enrollment?

No mention of convenience sampling etc

Baseline demographics were similar?

Yes, slightly more cancer in non-bridging group

Blinding?

Double blind

Equal treatment except intervention?

Yes, but treating physicians were given discretion in determining whether pts were high or low bleeding risk and were permitted to choose time to restart bridging based upon this

Complete f/u (> 80% f/u in both groups)

Yes

All pt important outcomes are considered?

Yes

Adequate sample size and large enough Tx effect to be notices?

Rates of thromboembolic events were lower than initially expected, so pre-specified ranges for noninferiority were wider than likely appropriate.

Comments / Problems

• Median time to thromboembolism after surgery was 19d, when pts should have been adequately anticoagulated on warfarin alone. Why?
• Most pts underwent a low bleeding risk procedure (89.4%), although having more high-risk pts would only bias towards non-bridging due to increased bleeding events
• Relatively few patients had a history of TIA/stroke