The use of antithrombotic therapies has expanded substantially in the general population, particularly among older adults. In Spain, approximately 1 million people receive anticoagulant therapy, mainly for atrial fibrillation (AF). The use of antiplatelet therapy is also growing, owing to the high prevalence of cardiovascular disease and the rising number of percutaneous coronary interventions (PCI) and endovascular procedures.1–5 With increasing life expectancy, it is highly likely that patients receiving antithrombotic therapy will, at some point, require an invasive procedure. In Europe, it is estimated that more than 6 million patients receiving antithrombotic therapy undergo some type of intervention each year.1

In recent years, several clinical practice guidelines have been developed to guide decisions on whether to interrupt antithrombotic therapy and to define the timing and duration of interruption during the periprocedural period. A notable example is the national consensus document coordinated by the Cardiovascular Thrombosis Working Group of the Spanish Society of Cardiology and endorsed by more than 20 Spanish scientific societies.1,6–9 Although this document had considerable impact when first released in 2018, its uptake has been limited in real-world practice. Importantly, inadequate management in this setting is associated with higher rates of both thrombotic and hemorrhagic adverse events.10–12 In light of new evidence on optimal periprocedural management, the consensus document has now been updated to incorporate advances since 2018. The aim of this review is to simplify decision-making, broaden endorsement by scientific societies, and facilitate dissemination to improve clinical practice in this area (table 1).

In addition to thromboembolic risk, this document also addresses bleeding risk assessment in patients receiving antithrombotic therapy who are scheduled to undergo a procedure or surgery. Bleeding risk stratification considers 3 factors: a) the volume of bleeding and the need for intraoperative or postoperative transfusion; b) the consequences of bleeding in an anemic or frail patient, in a confined anatomical space, in a critical organ, or when compromising the success of surgery; and c) patient-related factors. Relevant patient factors include age > 65 years, genetic predisposition, renal impairment, liver dysfunction, hypertension, alcohol consumption, prior bleeding or surgery, previous stroke, malignancy, thrombocytopenia or platelet dysfunction, anemia, unstable international normalized ratio (INR), and concomitant use of antiplatelet, anticoagulant, or anti-inflammatory drugs, among others.

The type of procedure remains the main determinant of bleeding risk and guides management of antithrombotic therapy.8,20,21 Operator training and experience also influence bleeding outcomes. In addition, some procedures, such as dental interventions, are often performed outside the hospital setting, where working conditions may complicate hemostasis and increase the risk of bleeding. With respect to timing, a recent study showed that most bleeding episodes occur within the first postoperative week, although risk persists for up to 30 days.22 Procedures most frequently associated with a high likelihood of transfusion include cardiac surgery, aortic aneurysm repair, radical cystectomy, open fracture repair, splenectomy, and hepatic, pancreatic, and colorectal resections.23

A list of common procedures and their bleeding risk by specialty is provided in the table 1 of the supplementary data (table 1 of the supplementary data). Major bleeding is defined according to the International Society on Thrombosis and Haemostasis (ISTH) as: fatal bleeding; bleeding causing symptoms in a critical organ (brain, spinal cord, eye, retroperitoneum, joint space, pericardium, or intramuscular bleeding with compartment syndrome); bleeding resulting in a hemoglobin drop of ≥ 2g/dL; or bleeding requiring transfusion of ≥ 2 units of red blood cells. The following considerations are essential when stratifying bleeding risk:8

• -

  Minimal risk procedures: These carry a near 0% likelihood of major bleeding. Examples include endoscopy without biopsy, phacoemulsification cataract surgery, cardiac device implantation, and minor dermatological or dental procedures. Hemostasis is generally straightforward, and bleeding rarely requires transfusion, compromises the procedure, or threatens the patient's life.

• -

  Low–intermediate risk procedures: These are associated with an approximately 0% to 2% risk of major bleeding within 30 days. They include interventions in which surgical or endoscopic hemostasis may be more difficult, increasing the likelihood of transfusion or reintervention. Examples are endoscopy with biopsy, complex dental procedures, cholecystectomy, abdominal wall hernia repair, breast surgery, lymph node biopsy, and ophthalmic procedures other than phacoemulsification cataract surgery.

• -

  High-risk procedures: These carry an incidence of major postoperative bleeding ≥ 2%. They include most major surgical and oncologic operations, complex endoscopic procedures, and all interventions involving neuraxial anesthesia. In these cases, postoperative bleeding may compromise procedural outcomes and, in some instances, become life-threatening. Also included are procedures in which the severity of bleeding depends on its anatomical location or on the difficulty of achieving effective hemostasis.

Before performing an invasive procedure in a patient receiving oral anticoagulant therapy, a key step is to evaluate thrombotic risk, bleeding risk, and the bleeding risk associated with the planned procedure. Management depends on the drug's mechanism of action, half-life, and dosage, which are closely linked to individual patient characteristics such as age, body weight, and hepatic and renal function (figure 2).

Figure 2.

Recommendations for discontinuation and resumption of anticoagulant therapy. CrCl, creatinine clearance. The pill-bottle icon indicates drug administration. * If there is high thromboembolic risk, consider bridging therapy.

(0.67MB).

Interrupting antiplatelet therapy requires balancing thrombotic and bleeding risk (figure 3). Understanding the pharmacokinetic properties of these agents is essential when planning both withdrawal and resumption after a procedure. Because thrombotic risk decreases as time passes from the acute vascular event, delaying surgery should be considered whenever feasible. In patients receiving antiplatelet therapy for primary prevention, treatment can be safely discontinued, and permanent discontinuation should be considered.13 In patients at low thrombotic risk, such as those with chronic coronary syndrome and stent implantation >12 months previously without additional risk factors, continuation of antiplatelet monotherapy during the procedure does not significantly affect ischemic or bleeding outcomes.27 In these cases, therapy withdrawal may also be considered, particularly when the procedure carries an intermediate-to-high bleeding risk. For patients continuing aspirin, doses <200 mg/d are recommended; however, in those receiving a higher dose at the time of the procedure (eg, 300 mg/d), there is no justification for postponing the intervention. In patients receiving P2Y12 receptor inhibitor monotherapy, the following preprocedure interruption periods are recommended: ticagrelor 5 days, clopidogrel 5 days, and prasugrel 7 days. For some interventions, such as coronary revascularization surgery, withdrawal of ticagrelor 3 days before the procedure can be considered. In exceptional cases where other antiplatelet agents are used, discontinuation schedules are as follows: triflusal 3 to 5 days before the procedure, and cilostazol or dipyridamole 48hours before.

Figure 3.

Recommendations for discontinuation and resumption of antiplatelet therapy. *In exceptional cases (eg, in neurosurgery), interruption of aspirin may also be equired.

(0.45MB).

When dual antiplatelet therapy is required due to thrombotic risk and surgery cannot be delayed, individualized management using a multidisciplinary approach is essential. In general, aspirin and a P2Y12 inhibitor should be continued during the first 30 days after a vascular event for procedures with low bleeding risk, whereas discontinuation of the P2Y12 inhibitor should be considered if bleeding risk is intermediate to high. As with anticoagulant therapy, bridging with intravenous antiplatelet agents (tirofiban, eptifibatide, or cangrelor) should be reserved for exceptional cases with high risk of both thrombosis and bleeding (figure 1 of the supplementary data).

In urgent situations, it is fundamental to recognize different clinical scenarios. Emergent procedures are those that must be performed within 6hours of clinical presentation. In these cases, anticoagulation should be stopped and supportive measures initiated. When available and indicated, administration of a reversal agent should be considered, or alternatively, the use of prothrombin complex concentrate (PCC), tailored to the specific anticoagulant. Supportive measures include local bleeding control with surgical hemostasis if feasible, transfusion of blood products, and general supportive care (eg, oxygen therapy, fluid resuscitation, vasoactive drugs). It is also important to identify comorbidities that increase bleeding risk, such as thrombocytopenia, uremia, or concurrent antiplatelet therapy.

Urgent (but nonemergent) procedures are those that can be performed between 6 and 24hours after clinical presentation. These cases require a management algorithm based on the specific anticoagulant the patient is receiving. It is essential to know the dose and timing of the last administration, as well as the patient's anticoagulant status, assessed using laboratory tests such as INR, prothrombin time, and activated partial thromboplastin time.

In patients receiving VKA therapy, INR should be determined (figure 4). If INR is <1.5, the surgery can proceed without delay. If INR is >1.5, management depends on the urgency of the procedure. When surgery can be delayed for 6 to 12hours, 10mg intravenous vitamin K should be administered and the INR reassessed after 6 to 8hours. Because the effect of vitamin K is not immediate, reassessment avoids unnecessary repeat dosing. If surgery cannot be delayed, 10mg of intravenous vitamin K should be administered together with a procoagulant agent, preferably PCC (25-50 IU/kg) adjusted according to current and target INR. Fresh frozen plasma is considered a second-line hemostatic agent, and is only effective (at a dose of 15-30 mg/kg) in the setting of VKA therapy; it is not recommended when PCC is available.28,29

Figure 4.

Recommendations for patients receiving vitamin K antagonists who require urgent surgery. INR, international normalized ratio; iv, intravenous; PCC, prothrombin complex concentrate; VKA, vitamin K antagonists. * According to the summary of product characteristics.

(0.33MB).

In patients receiving DOAC therapy who require urgent surgery, tests that measure plasma drug levels or activity can be useful, if available. Although evidence is limited regarding the effect of perioperative DOAC monitoring on reducing bleeding risk, plasma levels <30 ng/mL are recommended to allow surgery with an acceptable bleeding risk. Qualitative tests may also provide guidance, including activated partial thromboplastin time for dabigatran, and prothrombin time for rivaroxaban (but not for apixaban or edoxaban). A ratio ≤ 1.2 can with high probability rule out significant anticoagulant activity. Conversely, if coagulation testing indicates residual activity or if therapeutic adherence is adequate for apixaban or edoxaban, the maximum allowable surgical delay should be determined based on the patient's clinical condition (bleeding status, hemodynamic stability, and risk of life-threatening complications or sequelae). If surgery can be delayed, the timing should be guided by the last drug dose and renal function. When anticoagulant activity is present and surgery cannot be delayed, reversal agents should be considered: idarucizumab should be used for dabigatran, and andexanet al.fa for rivaroxaban and apixaban (but not for edoxaban), if these agents are available. If not, PCC can be used (figure 5). Neuraxial anesthesia is generally contraindicated in these patients.

Figure 5.

Recommendations for patients receiving direct oral anticoagulants who require urgent surgery. aPTT, activated partial thromboplastin time; CrCl, creatinine clearance; DOAC, direct oral anticoagulants; iv; intravenous; PCC, prothrombin complex concentrate; PT, prothrombin time. * Andexanet is not indicated as a reversal agent for edoxaban.

(0.47MB).

Patients receiving antithrombotic therapy who sustain a hip fracture represent a special case, as delaying surgery beyond 24 to 48hours is associated with increased morbidity and mortality. Therefore, despite the inherent bleeding risk of the procedure, surgery is recommended within 24 to 48hours of admission. Peripheral nerve blocks and the choice of anesthesia (neuraxial or general) should be tailored to the patient's hemostatic status.

Despite the potential for increased bleeding, antiplatelet therapy in patients requiring urgent surgery or an invasive procedure generally does not justify delaying the intervention if the patient's clinical status precludes postponement.6 It is therefore essential to classify the surgery as emergent (within 6hours), urgent (6-24hours), or deferred urgency (can be delayed beyond 24hours). The feasibility of implementing measures to reduce or reverse the antiplatelet effect of the drugs—while balancing thrombotic risk—depends on having sufficient time before the procedure.

If urgent surgery can be delayed for a sufficient period, it is possible to evaluate whether to continue or interrupt antiplatelet therapy, including the selective use of bridging therapy. This decision should follow careful assessment of the procedure's bleeding risk, ideally by a multidisciplinary team. Platelet function testing can be used to shorten the discontinuation period for P2Y12 antagonists.30 Several assays are available, but no single gold standard has been established (table 3 of the supplementary data). Although interpretation requires experience, point-of-care tests are generally preferred because they are easy to use and do not require specialized laboratory personnel.

When surgery is emergent or cannot be delayed long enough to allow antiplatelet discontinuation, options to reduce periprocedural bleeding risk are limited. The most effective strategies are meticulous surgical hemostasis and platelet transfusion in the event of intraoperative bleeding; prophylactic platelet transfusion is generally not recommended. Transfusion is minimally effective in patients receiving ticagrelor, due to its pharmacologic characteristics, particularly reversible receptor binding and the half-life of the drug and its active metabolite. Nevertheless, platelet transfusion may be useful in patients treated with aspirin, clopidogrel, or prasugrel. Typically, 2 to 4 units are required for clopidogrel and prasugrel, with an ideal waiting period of 6hours after the last dose. In selected cases, such as cardiac surgery, hemoadsorption via an adsorbent filter connected to the extracorporeal circuit can remove most circulating ticagrelor within 2 to 3hours. Recently, the monoclonal antibody bentracimab has shown promise as a ticagrelor reversal agent in emergent surgery;31 however, it is not yet available for clinical use. Finally, despite limited evidence for desmopressin, hemostatic or procoagulant agents are not routinely recommended to counteract the effects of antiplatelet therapy in patients requiring urgent surgery, given the potential for increased thrombotic risk.

Managing antithrombotic therapy in the periprocedural period remains a clinical challenge, with potentially serious consequences if handled inadequately. This consensus statement, endorsed by the majority of scientific societies involved in perioperative care (table 1), aims to provide clear, practical guidance to support the appropriate use of anticoagulant and antiplatelet therapy. Its goal is to harmonize clinical practice across specialties, ensuring consistent patient care regardless of the treating physician (figure 6).

Figure 6.

Central image. Recommendations for the management of antithrombotic therapy during the perioperative and periprocedural periods. VKA, vitamin K antagonist.

(0.35MB).

None.

The authors declare that artificial intelligence was not used for the preparation of this article.

All authors fulfill the following requisites: 1) they made substantial contributions to the conception and design of the manuscript; 2) they drafted the article or critically reviewed its intellectual content; 3) they approved the final version for publication; and 4) they accept responsibility for all aspects of the article and agree to investigate and resolve any issues regarding the accuracy and integrity of any part of it.

D. Vivas: lecture fees from AstraZeneca, Bayer, Bristol-Myers Squibb, Pfizer, and Ferrer; consultancy fees from AstraZeneca and Daiichi-Sankyo. R. Ferrandis: lecture fees from Octapharma AG, LFB Bioterapias Hispania, and Behring; support for conference attendance from Octapharma AG and LFB Bioterapias Hispania. M. Echeverri: lecture fees from Octapharma and LFB. J.L. Ferreiro: lecture fees from AstraZeneca, Abbott, Bristol-Myers Squibb, Pfizer, Ferrer, Terumo, and Chiesi; consultancy fees from AstraZeneca, Daiichi-Sankyo, and Biotronik. D. González-Casal: grants/support from Abbott, St. Jude Medical, and Boston Scientific; support for conference attendance from Abbott, St. Jude Medical, Boston Scientific, Biosense Webster, and Johnson & Johnson. J. López-Menéndez: lecture fees from Medtronic, Edwards, and Getinge; shares in Johnson & Johnson and Abbott. E. Rodríguez de Santiago: lecture fees from Olympus, Izasa, Boston Scientific, Apollo Endosurgery, Erbe, AstraZeneca, and Norgine; consultancy fees from Olympus; grants/support from 3D Matrix, Spanish Society of Digestive Endoscopy, and AES 2023. J. Mateo-Arranz: grants/support from Daiichi-Sankyo. J. Polo: lecture fees from Boehringer-Ingelheim, AstraZeneca, and BMS-Pfizer; consultancy fees from Incyte and AstraZeneca; support for conference attendance from Incyte, AstraZeneca, and Boehringer-Ingelheim. Á. Castellanos: lecture fees from Semfyc, Almirall, Boehringer, Menarini, Daiichi-Sankyo, and AstraZeneca. O. Madridano: lecture fees from Rovi and Leo Pharma; support for conference attendance from Rovi, Leo Pharma, and Bristol-Myers Squibb. D. Jiménez: lecture fees from Rovi. E. Figuero: lecture fees from Oral B, Colgate, Kenvue, and several periodontology societies; consultancy fees from Kenvue; grants/support from Dentaid, Lacer, Straumann, and Colfeed4print; support for conference attendance from Colgate, Kenvue, and several periodontology societies. F. Marín: lecture fees from Novartis, Bayer, and Pfizer; consultancy fees from Novartis; grants/support from BMS and Daiichi-Sankyo; support for conference attendance from Novartis and Pfizer; payment for expert testimony from Boehringer-Ingelheim. L. Rodríguez-Padial: member of the committee on VKA control in patients with atrial fibrillation of the Spanish Society of Cardiology, supported by Daiichi-Sankyo; president of the Spanish Society of Cardiology. I. Fernández-Lozano: lecture fees from Microport; grants/support from the Interhospital Research Foundation and SEC; participation in the advisory board of Abbott. C. Cassinello: lecture fees from Medtronic. E. Muñoz: lecture fees from the Spanish Society of Periodontology; support for conference attendance from the Spanish Society of Periodontology.

The remaining authors declare no conflicts of interest.