Sanjog Singh MS4, University of New Mexico School of Medicine

Chris Gutjahr, Assistant Professor in Interventional Radiology, University of New Mexico School of Medicine 

Even the newest of IR trainees quickly come to realize that the prevention and management of bleeding is an incredibly important consideration for any image-guided intervention. The SIR published its first consensus guidelines on management of thrombotic and bleeding risk in 2012 and updated them through two important papers published in JVIR earlier this year (links at the bottom of the page.) Part I is a relevant review of the physiology and pharmacology of hemostasis. Part II lays out the most updated recommendations. We will focus on 4 major takeaways from the updated recommendations, and we have structured them in the form of an algorithm that can be used to evaluate periprocedural bleeding and thrombosis for all IR patients:

1.     Evaluate bleeding/thrombosis risk based on patient-specific factors.

2.     Evaluate procedure-associated bleeding risk.

3.     Evaluate periprocedural labs.

4.     Use the recommended hold and re-initiation times for commonly used antithrombotics in the context of the overall clinical picture.

Of note, the data remains limited to retrospective studies due to minimal availability of high-quality, randomized, controlled data. Thus, the recommendations should be considered in the context of the patient’s overall clinical status, and a physician may deviate from them as necessitated by the individual patient, available resources, and established guidelines within the institution.

1.     Evaluate bleeding/thrombosis risk based on patient-specific factors. Specific characteristics and comorbidities unique to a patient may increase their risk of bleeding or forming a clot and warrants pre-procedural evaluation. For instance, an underlying bleeding diathesis increases risk of periprocedural bleeding while an underlying malignancy may predispose a patient to periprocedural clot formation. There exist several scoring systems that can help evaluate an individual patient’s bleeding and thrombosis risk. The challenge is that these scoring systems refer to long-term risk, and therefore extrapolating from long-term risk to periprocedural risk is difficult and has not been validated. Nonetheless, when used as a component of the overall clinical picture, they are often used in clinical practice as a general guide to considering factors that may increase patient-specific bleeding/thrombosis risk. The two scoring systems highlighted in the JVIR article are the CHA2DS2-VASc score used to predict annual stroke risk in patients with nonvalvular atrial fibrillation and the HAS-BLED score in assessment of patient bleeding risk. Consider the components of these risk stratification systems when performing a pre-procedural bleeding/thrombosis risk assessment in all IR patients.   

2.     Evaluate procedure-associated bleeding risk. Periprocedural bleeding risk also depends on the procedure being performed. The risk of a major bleeding event is considerably lower for a nephrostomy tube exchange than for a TIPS, for example. The major update in the guidelines is in the categorization of procedure-related bleeding risk. Previous guidelines divided procedures into 3 tiers: low, moderate, and high bleeding risk. The new guidelines use a two-tier system: low risk and high risk. Low risk procedures are those that are expected to rarely have hemorrhagic complications or occur in areas where bleeding is easily identified and controlled, whereas high risk procedures or those in which hemorrhagic complications may be expected or involve locations where bleeding can have devastating consequences. In an attempt to define risk, high risk procedures can be categorized as having a > 1.5% risk of major bleeding or a 2-day risk of major bleeding of 2-4%. Low risk procedures have < 1.5% risk of major bleeding or a 2-day risk of major bleeding of 0-2%. In addition to considering the procedure-related bleeding risk, the guidelines recommend factoring in location and potential consequences of bleeding. 

3.     Evaluate periprocedural labs. Certain lab values are frequently used in IR to assess periprocedural bleeding risk. When interpreting these lab values, it is helpful to keep these 2 thoughts in mind: 1) Be aware of what a positive or negative lab value does and does not tell you. For instance, the prothrombin time (PT) test evaluates the coagulation cascade and may be used to assess bleeding risk before procedures. It was developed to identify the cause of bleeding in symptomatic patients; however, mild to moderate PT prolongation has not been shown to predict bleeding risk in a nonbleeding patient. 2) Always keep in mind the patient’s full clinical context. A useful example is a chronic liver disease (CLD) patient. These patients have nearly twice the thrombotic risk of the general population, yet many of them have an increased PT/INR. This paradox occurs because routine coagulation studies such as the PT/INR reflect a change only in the pro-coagulation factors and not the anti-coagulation factors. The liver produces the majority of both pro-coagulation and anti-coagulation factors used in hemostasis, and patients with CLD have rebalanced primary and secondary hemostasis due to relative decreases in both. Recall that whether excessive bleeding or clot formation occurs is a product of the relative balance between pro/anti-coagulation factors, and not necessarily the absolute value of the pro-coagulation factors. Therefore, a CLD patient with an INR of 3 may be incorrectly assumed to be anticoagulated when, in fact, they may be rebalanced or tipped towards pro-thrombosis, and the elevated INR is only taking into account the decrease in pro-coagulation factors.

With these two concepts in mind, the most commonly used periprocedural labs for evaluating bleeding risk are the PT/INR, platelet count, and hemoglobin. Keep in mind that there are no high-quality data to guide whether preprocedural lab testing reduces periprocedural bleeding risk. The panel did not recommend routine hemoglobin or platelet counts for procedures with low bleeding risks. However, they may be indicated if a patient inherently has a higher bleeding risk such as those with hematologic disorders and patients receiving certain chemotherapies. For procedures with high bleeding risk, the panel recommends routine preprocedural tests which may include hemoglobin, platelet count and PT/INR. PTT or anti-Xa can be considered for patients receiving heparin. The suggested laboratory thresholds are to transfuse platelets if <50 x 109/L and correct INR to 1.5-1.8. Although of low quality, available literature suggests low bleeding risk procedures can be safely performed at INR >1.5 or platelet count > 20 x 109/L. Thus, the recommendation for low bleeding risk procedures is to correct INR to within range of 2.0-3.0 and transfuse if platelets < 20 x 109/L. If arterial access is required for low risk procedures, the recommended INR threshold is < 1.8 for femoral access and <2.2 for radial access. 

Low Risk Procedures

First Definition

Expected to rarely have hemorrhagic complications


Occur in areas where bleeding is easily identified or controlled


-Catheter exchanges (gastrostomy, abscess, nephrostomy)

-Dialysis access interventions


-IVC filter placement

-PT/INR, hemoglobin, and platelet count not routinely recommended


INR: correct to within range of  2.0–3.0‡

-Platelets: transfuse if < 20×109/L

Second Definition

< 1.5% risk of major bleeding


2-day risk of major bleeding of 0-2%

High Risk Procedures

First Definition

Hemorrhagic complications may be expected


Involve locations where bleeding is difficult to identify or can have devastating consequences


-Solid organ ablations

-Catheter-directed thrombolysis

– Portal vein interventions

– Solid organ biopsies

-Routine labs including PT/INR, hemoglobin, and platelet count recommended


INR: correct to within range of 1.5–1.8

-Platelets: transfuse if < 50x109/L

Second Definition

> 1.5% risk of major bleeding


2-day risk of major bleeding of 2-4%


Additionally, PT/INR and PTT can be used to monitor the presence and level of specific anti-coagulants such as warfarin and heparin respectively. Tests such as thrombin time and Xa activity test can be used to measure the presence and level of direct oral anticoagulants such as apixaban. You may see thromboelastography (TEG) used at some point during your IR rotations. TEG works by assessing the viscoelastic properties of clot formation in whole blood from the start of primary hemostasis through clot lysis. In contrast to PT and PTT which only analyze a part of the clotting cascade, TEG measures the entire process of hemostasis including platelet function, clot formation, fibrin cross-linking, etc. The value of TEG in the preprocedural workup is unknown, and these tests have not been validated to assess bleeding risk in nonbleeding patients. Some authors have advocated that TEG has distinct benefits for patients with CLD in whom traditional coagulation studies are less accurate.

4.     Use the recommended hold and re-initiation times for commonly used antithrombotics in the context of the overall clinical picture. Antithrombotic medications are among the most commonly prescribed medications in the US and consist of antiplatelets and anticoagulants. Antiplatelets work by inhibiting platelet activation and aggregation, while anticoagulants work by either blocking the synthesis or inhibiting the action of pro-thrombotic enzymes in the coagulation pathway. Part I of the consensus guidelines article goes into some detail and has some useful tables regarding each class of medication.

One of the most common questions IR physicians get from consulting providers is whether or not to hold anticoagulation and/or antiplatelets. Three factors must be considered when answering this question: 1) the overall patient status including thrombotic and bleeding risks 2) procedural bleeding risk 3) pharmacological factors of the medication being held. The goal, of course, is to minimize risk of medication-related bleeding during the procedure while avoiding thrombosis since the patient has some propensity to form clots.

Essentially, the decision must be made whether the procedure can be performed while the patient is on the anticoagulant.  If not, then decisions on how long to hold the medication before the procedure, and when to resume the patient’s anticoagulant after the procedure must be made.  Stepwise algorithms to assist in decision making are provided in the article (see algorithm.) Additionally, medication specific recommendations are included (see table). For some patients with a high thrombotic risk, ‘bridging’ with a different anticoagulant is required to minimize the periprocedural risk of a thrombotic event.  Bridging is done with a parenteral agent that has a short half-life, typically heparin, giving the ability to minimize the amount of time a patient is not anticoagulated.     

Take away points

·       The clinical decision-making for periprocedural management of thrombotic and bleeding risk is complex and should be specific to each patient with variables including the provider, the practice, available resources, the patient’s specific risks, and comorbid conditions affecting the decision.

·       The new consensus guidelines provide useful timetables and algorithms for management but should be used in the context of the previous point.

Part I:

Part 2: