Early observations and pilot data that first suggested a new direction
For patients on warfarin requiring surgery, 'bridging anticoagulation' — substituting low-molecular-weight heparin (LMWH) during the perioperative warfarin interruption — became standard practice in the early 2000s based on theoretical reasoning rather than clinical trial evidence. The logic seemed sound: the 4-5 day gap while warfarin was held exposed patients to thromboembolic risk, and LMWH could fill that gap. Observational studies and expert consensus guidelines recommended bridging for patients at moderate-to-high thromboembolic risk, and the practice became nearly universal for patients with atrial fibrillation, mechanical heart valves, or recent VTE. However, early registry data began suggesting that bridging might cause more bleeding than it prevented in thromboembolism, planting seeds of doubt.
Landmark RCTs and pivotal trials that established the evidence base
The BRIDGE trial (2015) was a practice-changing double-blind RCT that definitively answered the bridging question for atrial fibrillation. In 1,884 patients with atrial fibrillation on warfarin requiring an elective procedure, perioperative bridging with dalteparin was compared to placebo. The results were unequivocal: no bridging was non-inferior to bridging for thromboembolic events (0.4% vs 0.3%) and was associated with significantly less major bleeding (1.3% vs 3.2%). The trial demonstrated that for the vast majority of AF patients, bridging anticoagulation caused net harm — the excess bleeding far outweighed the minimal thromboembolic risk during a brief warfarin interruption. This was one of the clearest demonstrations that a deeply entrenched practice, supported by physiological reasoning and expert consensus, was actually harmful.
Follow-up studies, subgroup analyses, and real-world validation
The BRIDGE trial resolved the question for most AF patients, but important populations remained underrepresented: patients with mechanical heart valves, recent VTE, and high CHADS2-VASc scores. The PERIOP-2 trial (2019) evaluated bridging versus placebo in a broader population including patients with mechanical valves and VTE, though the trial was underpowered after slow enrolment. Post-hoc analyses confirmed the BRIDGE findings in higher-risk subgroups. The emergence of DOACs (direct oral anticoagulants) further simplified perioperative management — their short half-lives (12-24 hours) eliminated the need for bridging entirely in most cases, as simply holding the drug for 1-2 half-lives provided adequate clearance. The PAUSE study (2019) prospectively validated a standardised DOAC interruption protocol without bridging in 3,007 patients, demonstrating low rates of both thromboembolism (0.16-0.60%) and major bleeding (0.90-1.85%).
Integration into clinical practice guidelines and recommendations
Post-BRIDGE, guidelines uniformly shifted against routine bridging. The AHA/ACC 2023 guidelines for perioperative management recommend against bridging anticoagulation for most patients with atrial fibrillation (Class III: Harm). The ASH 2018 guidelines suggest forgoing bridging in most patients with VTE beyond the first month. For DOAC patients, guidelines recommend simple interruption without bridging based on the PAUSE protocol. Bridging is now reserved for the highest-risk patients: mechanical mitral valves, recent stroke (<3 months), and recent VTE (<3 months), where the thromboembolic risk during interruption may justify the bleeding cost.
Recommend against bridging anticoagulation for patients with atrial fibrillation without mechanical heart valves (Class III: Harm)
ASH Guidelines on VTE Management 2018
Suggest forgoing bridging in most patients with VTE requiring warfarin interruption (>1 month from acute event)
Now
Current standard of care and ongoing research directions
The paradigm has shifted from 'bridge everyone' to 'bridge almost no one.' For the majority of AF patients on warfarin, bridging is no longer recommended. For DOAC patients, simple drug interruption based on procedure bleeding risk and renal function is the standard approach — no bridging needed. The remaining clinical dilemma centres on the truly high-risk patient: those with mechanical mitral valves (where thromboembolic risk is highest) and those with very recent thromboembolism. Even in these populations, the evidence increasingly favours shorter warfarin interruptions and limited bridging. The broader trend is the replacement of warfarin by DOACs for most indications, which inherently simplifies perioperative management. Current research focuses on perioperative DOAC management for high-bleeding-risk procedures, the role of DOAC reversal agents (idarucizumab, andexanet alfa) in urgent surgery, and point-of-care coagulation testing to guide timing of procedures after anticoagulant interruption.
Which patients still need bridging anticoagulation?+
Bridging is now reserved for a small subset of high-risk patients: those with mechanical mitral valves (higher thromboembolic risk than aortic valves), patients with recent stroke or systemic embolism (<3 months), and patients with recent VTE (<3 months). Even in these populations, the decision is individualised, weighing the procedure-specific bleeding risk against the thromboembolic risk. The BRIDGE trial excluded mechanical valve patients, so this population remains an evidence gap.
Do DOAC patients need bridging?+
No. The pharmacokinetics of DOACs (short half-lives of 12-24 hours) make bridging unnecessary. The PAUSE study validated a simple protocol: hold the DOAC for 1 day before low-bleeding-risk procedures and 2 days before high-bleeding-risk procedures (longer for renal impairment with dabigatran). Resume 1-3 days postoperatively depending on procedure bleeding risk. No LMWH bridging is needed.
How should I manage warfarin perioperatively?+
For most patients: stop warfarin 5 days before surgery, allow INR to normalise, proceed with surgery, and restart warfarin on the evening of surgery or day 1 postoperatively. No bridging for AF patients (unless mechanical mitral valve or very high-risk features). Check INR the day before surgery — if >1.5, consider vitamin K 1-2 mg orally. Resume warfarin as soon as adequate haemostasis is achieved.
What about urgent surgery on anticoagulants?+
For warfarin: IV vitamin K (5-10 mg) plus 4-factor prothrombin complex concentrate (PCC) for immediate reversal. For dabigatran: idarucizumab 5g IV provides complete reversal within minutes. For rivaroxaban/apixaban: andexanet alfa is approved for life-threatening bleeding; 4-factor PCC is used off-label when andexanet is unavailable. For all DOACs, the short half-life means that delaying surgery by 12-24 hours, when feasible, often provides sufficient drug clearance.
