When Is Bridging Anticoagulation Needed Before Surgery?

Bridging anticoagulation is the temporary use of a short-acting injectable blood thinner, typically a heparin-based drug, to cover the gap when a patient’s long-term oral anticoagulant is paused for surgery or a procedure. The practice was once considered routine for anyone on chronic anticoagulation, but large clinical trials over the past decade have turned that thinking on its head. For most patients, skipping bridging turns out to be just as safe for preventing blood clots and significantly safer for avoiding bleeding complications. The question has shifted from “how do we bridge?” to “should we bridge at all?” and the answer, for most people, is no.

Why Bridging Existed in the First Place

When you take a drug like warfarin for a condition such as atrial fibrillation or a mechanical heart valve, stopping the medication before surgery leaves your blood unprotected against clots for several days. Warfarin takes roughly five days to clear from your system before surgery and another several days to reach a therapeutic level afterward. That window of vulnerability worried clinicians, so the traditional approach was to inject a faster-acting anticoagulant, usually low-molecular-weight heparin (LMWH), during those gaps. The idea made intuitive sense: keep the blood thin at all times, and you prevent strokes and other clot-related events.

The problem was that this logic was never tested rigorously for a long time. Clinicians bridged patients based on theoretical risk, and the practice became widespread before anyone checked whether it actually improved outcomes. When researchers finally did check, the results were surprising.

The Trial That Changed Practice

The BRIDGE trial, published in the New England Journal of Medicine, enrolled over 1,800 patients with atrial fibrillation who needed to stop warfarin for elective surgery. Half received bridging with LMWH; the other half received placebo injections. The primary outcome was arterial thromboembolism, meaning strokes and similar clotting events. The rate was 0.4% in the group that did not receive bridging and 0.3% in the group that did, a difference so small it confirmed that skipping bridging was not inferior.

The bleeding results, however, were dramatic. Major bleeding occurred in 1.3% of the no-bridging group versus 3.2% of the bridging group. That means bridging more than doubled the risk of serious bleeding without providing any meaningful clot protection in return. The trial had a key limitation: it mostly included patients at low to moderate clot risk, with scores below the highest risk tier. Whether the results apply to the highest-risk patients remained an open question.

What About Higher-Risk Patients

The PERIOP2 trial tried to fill that gap. It enrolled patients considered at higher risk for clots, including those with mechanical heart valves, and randomized them to postoperative bridging with dalteparin (an LMWH) or placebo. The findings were remarkably consistent with the BRIDGE trial: major thromboembolism occurred at similar rates in both groups (about 1% with bridging versus 1.2% without), and the difference was not statistically meaningful. Major bleeding trended lower with bridging in this trial but again did not reach statistical significance.

The results held true for both the atrial fibrillation and the mechanical heart valve subgroups. This was a meaningful extension of the evidence, because mechanical valve patients had always been considered the population most likely to benefit from bridging. Even in that group, the data did not support a clear advantage.

⁴

What Current Guidelines Recommend

The American College of Chest Physicians (ACCP) updated their perioperative guidelines to reflect this evidence, and the recommendations are strikingly consistent across indications. For patients on warfarin for atrial fibrillation who need elective surgery, the guideline makes a strong recommendation against bridging. For patients on warfarin for a mechanical heart valve or for venous thromboembolism (VTE), the guideline suggests against bridging, using somewhat softer language that reflects the thinner evidence base in those populations. For patients on direct oral anticoagulants (DOACs) who need surgery, the guideline also suggests against bridging.

⁵

The distinction between “recommend against” and “suggest against” matters clinically. A strong recommendation means the evidence is solid enough that most patients should follow it. A conditional suggestion means the evidence points in one direction but is less certain, leaving more room for individualized decisions. For mechanical valve patients specifically, clinicians still weigh patient-specific and valve-specific factors when deciding whether a period without anticoagulation is safe.

⁶

How Direct Oral Anticoagulants Changed the Equation

Much of the original bridging rationale was built around warfarin’s slow on-and-off pharmacology. DOACs like apixaban, rivarelbaan, edoxaban, and dabigatran work differently. They reach full effect within hours and clear the body within one to two days, depending on kidney function. That short timeline means the vulnerable window around surgery is much narrower than with warfarin, making the case for bridging even weaker.

A large prospective study of nearly 3,000 patients on DOACs tested a simple interruption-and-resumption strategy: skip the DOAC for one day before low-bleeding-risk procedures and two days before high-bleeding-risk procedures, then restart one day after low-risk procedures and two to three days after high-risk ones. No bridging was used. The rates of major bleeding and thromboembolism during the 30-day follow-up were low.

⁷

This straightforward approach has become the de facto standard for DOAC patients. Because the drugs wear off and kick back in quickly, there is no extended gap to “bridge” across. Adding injectable heparin on top of a DOAC’s already tight timeline only introduces bleeding risk without a corresponding benefit in clot prevention.

How Procedure Risk Factors In

The decision to interrupt anticoagulation at all, and how long to interrupt, depends heavily on the procedure itself. Some operations carry a high inherent bleeding risk (cardiac surgery, major abdominal procedures, spinal surgery), while others carry low risk (cataract removal, many dental procedures, some endoscopies). An international subcommittee of thrombosis specialists has published guidance emphasizing that procedural bleed risk stratification should drive the timeline of anticoagulation interruption and resumption.

⁸

For low-bleeding-risk procedures, many patients can continue their anticoagulant without interruption, eliminating the bridging question entirely. When anticoagulation does need to be paused, the duration of the interruption and the timing of resumption depend on whether the procedure is classified as high or low risk for bleeding. Interestingly, one analysis of data from the BRIDGE trial found that the surgical bleed risk classification predicted bleeding only in a simple analysis and lost significance once other patient factors were accounted for, suggesting that individual characteristics matter as much as procedure type.

⁹

LMWH Versus Unfractionated Heparin

When bridging is used, the two main agents are low-molecular-weight heparin (LMWH) and unfractionated heparin (UFH). LMWH, given as a subcutaneous injection once or twice daily, is far more convenient than UFH, which traditionally requires continuous intravenous infusion in a hospital. A meta-analysis comparing the two agents found that LMWH was associated with significantly fewer clotting events than UFH, while major bleeding rates were similar between the two.

¹⁰

A more recent systematic review focused specifically on mechanical heart valve patients undergoing non-cardiac surgery found no significant differences between the two agents across any outcome, including clotting events, mortality, major and minor bleeding, and low platelet counts.

¹¹

One small pilot study in patients undergoing cardiac surgery noted that those receiving unfractionated heparin had significantly more hemorrhagic pericardial effusions requiring drainage compared to those on LMWH (about 6% versus 2%).

¹²

In practice, LMWH dominates bridging when it is used, largely because patients can inject it themselves at home rather than spending extra days in the hospital.

Bleeding Complications When Bridging Is Used

The bleeding complications from bridging are not abstract statistics. In patients receiving cardiac implantable devices like pacemakers and defibrillators, bridging has been linked to significantly higher rates of pocket hematoma, the painful accumulation of blood at the implant site. One study of 161 device procedures found that heparin bridging was significantly associated with complications, while warfarin therapy itself was not.

¹³

The timing of the bridging injections matters. A study of patients receiving pacemakers and defibrillators found that giving LMWH after surgery was the main driver of hematomas. Postoperative LMWH raised the hematoma rate from 8% to 23%. The dose given the night before surgery, by contrast, did not significantly affect hematoma rates. A multivariate analysis confirmed postoperative LMWH as an independent predictor of significant wound hematoma.

¹⁴

Even in a carefully managed standardized protocol, excessive wound-related bleeding after the procedure precluded giving the planned postoperative dose in nearly 6% of patients undergoing non-high-bleeding-risk procedures. The body simply was not ready for anticoagulation yet.

¹⁵

Rebound Hypercoagulability After Stopping Anticoagulants

One of the theoretical justifications for bridging was the idea that stopping an anticoagulant creates a brief “rebound” state where the blood becomes even more prone to clotting than it was before treatment. Research suggests this phenomenon is real but modest. A pooled analysis of 20 trials found that clot recurrences were about 2.6 times more frequent in the two months after stopping oral anticoagulants than in the months that followed, working out to roughly 2% of patients experiencing a rebound-related clot event.

¹⁶

Lab studies have confirmed the biochemical basis. When researchers measured thrombin-antithrombin complex levels (a marker of clot formation activity) after stopping anticoagulants, they found a transient spike peaking around day four. Almost all patients showed levels above normal for a stretch of days three through eleven, before values returned to baseline after about a month.

¹⁷

This rebound effect was one of the stronger theoretical arguments for bridging. But the clinical trial data showed that even with this transient hypercoagulable window, actual clotting events remained rare whether or not bridging was used. The rebound appears to be a real physiological phenomenon that, for most patients, does not translate into a clinical problem large enough to justify the bleeding costs of bridging.

Special Considerations for Obesity and Kidney Disease

When bridging is used, dosing LMWH in patients with obesity or impaired kidneys requires extra thought. LMWH is typically dosed by body weight, which raises questions when a patient weighs significantly more than the populations in the original dosing studies. A pharmacokinetic analysis found that dosing by total body weight in obese patients (with BMIs up to about 32) produced drug levels comparable to non-obese patients. For more extreme body weights, dose capping may be needed because the drug’s distribution volume increases in proportion to lean body mass rather than total body weight.

¹⁸

Some clinicians order anti-Xa blood level monitoring to check whether the LMWH dose is appropriate in these patients. However, a critical review of the evidence found that anti-Xa levels are not strongly linked to clinical outcomes such as bleeding or clotting. Clinical studies across surgical and medical settings found no difference in outcomes between obese and non-obese patients receiving LMWH. The review concluded that routine monitoring of anti-Xa levels in obese patients is not supported by current evidence, though it may be useful when kidney function is impaired, when a patient responds unexpectedly, or when there is a question about whether injections are actually being taken as directed.

¹⁹

Kidney disease is the more consequential variable. LMWH is cleared primarily through the kidneys, so impaired kidney function causes the drug to accumulate, raising the risk of dangerous bleeding. This is where dose adjustments and occasionally anti-Xa monitoring become genuinely important clinical tools.

The Cost and Practical Burden

Bridging is not cheap, and the strategy chosen can swing costs dramatically. A cost analysis estimated that for a surgery requiring an overnight stay, bridging with self-administered LMWH at home costs roughly $672, LMWH given by a visiting nurse costs about $933, and continuous intravenous unfractionated heparin in the hospital runs around $3,816.

²⁰

The hospital-based UFH approach costs more than five times as much as home-administered LMWH. This financial difference, combined with the clinical equivalence of the two agents in most studies, is a major reason LMWH became the dominant bridging agent. But the cost comparison also highlights a broader point: if most patients do not benefit from bridging at all, the entire expense is avoidable.

Self-Injection Adherence Is Worse Than Clinicians Assume

When bridging with LMWH is prescribed for outpatient use, patients are typically asked to inject themselves at home. Clinicians often assume this happens reliably, but the reality is messier. A study of outpatients prescribed self-administered heparin injections found that about 8% refused to do it at all, requiring a nursing service instead. Among those who did self-inject, about a third had initial difficulty, and about 6% struggled with the injections throughout the entire course. Roughly 28% had uncertain compliance, meaning they either forgot individual doses or stopped the therapy early. Every patient under age 20 in the study fell into the uncertain-compliance group.

²¹

This has real implications. If a clinician prescribes bridging on the assumption it will reliably protect the patient, but the patient misses doses or stops early, the anticipated benefit evaporates while the prescription still complicates the perioperative plan and adds cost. Adherence concerns are another reason the move away from routine bridging is pragmatically sensible: a strategy that does not depend on patients faithfully injecting themselves at home is inherently more reliable.

Neuraxial Anesthesia and Spinal Bleeding

One uncommon but devastating complication relevant to bridging is spinal epidural hematoma, which can occur when anticoagulant drugs are present around the time of spinal or epidural anesthesia. Case reports have documented this complication in patients receiving LMWH in the perioperative period. The accumulation of blood in the spinal canal can compress the spinal cord, and without emergency surgery, permanent paralysis can result. Guidelines for timing neuraxial procedures around anticoagulant use exist precisely because of this risk. When a patient is receiving bridging LMWH and their procedure requires spinal or epidural anesthesia, the timing of injections relative to needle insertion and catheter removal must be carefully coordinated, typically requiring at least 12 hours between a prophylactic LMWH dose and the procedure, and longer for therapeutic doses.

This concern applies whether the LMWH is being used for bridging or for any other indication, but bridging introduces it into surgical situations where it might not otherwise be a factor. Patients on warfarin alone, without bridging, would simply stop the warfarin and present for surgery with a normal or near-normal clotting profile, posing minimal concern for neuraxial procedures.