INTRODUCTION — Neuraxial anesthesia techniques are used in many settings: as intraoperative anesthetics, for postoperative pain control, in the peripartum period, and in the management of chronic pain. Patients who are candidates for neuraxial anesthesia techniques may be receiving chronic antiplatelet or anticoagulation therapy, may require anticoagulation during or following surgery, or may receive prophylactic medication for venous thromboembolism in the perioperative period. Patients who receive anticoagulant or antiplatelet therapy are at increased risk for hemorrhagic complications of neuraxial techniques, most notably spinal epidural hematoma (SEH).
This topic discusses the use of neuraxial anesthesia in patients who are receiving medications that interfere with coagulation or platelet function, including the level of risk and the timing of neuraxial procedures relative to the antithrombotic (anticoagulant and antiplatelet) medication. Issues related to neuraxial anesthesia in individuals with impaired hemostasis for reasons other than medications (eg, severe thrombocytopenia) are discussed separately. (See "Adverse effects of neuraxial analgesia and anesthesia for obstetrics", section on 'Neuraxial analgesia and low platelets' and "Overview of neuraxial anesthesia", section on 'Spinal-epidural hematoma (SEH)'.)
For the purpose of this topic, neuraxial anesthesia refers to spinal, epidural, or combined spinal-epidural procedures that may be performed for surgical anesthesia or perioperative analgesia.
Recommendations for interventional spine and pain procedures in patients receiving antiplatelet or anticoagulant medications are discussed separately [1]. (See "Spinal cord stimulation: Placement and management", section on 'Patient screening'.)
The recommendations presented here may be applicable to patients having paravertebral blocks, deep plexus blocks, or deep peripheral nerve blocks (ie, in anatomic locations not amenable to the application of pressure to control hemorrhage). However, there is lack of consensus on the magnitude of risk and on which blocks should be avoided in patients who are anticoagulated. (See "Overview of peripheral nerve blocks", section on 'Patients on antithrombotic medication'.)
SPINAL EPIDURAL HEMATOMA (SEH) — Bleeding is the major complication of antithrombotic therapy. When bleeding occurs in the closed space of the spinal canal, the expanding hematoma can cause pressure on the spinal cord or cauda equina, which in turn may lead to spinal cord ischemia and infarction, with potential for severe neurologic injury or paraplegia. Following neuraxial anesthesia, bleeding is most commonly from vessels in the prominent venous plexus of the epidural space, although it can be in the subdural or subarachnoid spaces. We will refer to this bleeding as spinal epidural hematoma (SEH); the considerations discussed below do not differ for other bleeding locations within the spinal canal.
Not all SEHs are related to anesthetic or other medical procedures; SEH can occur spontaneously or following other predisposing events such as back surgery [2]. In a retrospective review of all SEHs at a single institution from 1986 to 2001, 7 of the 17 cases were spontaneous; none were related to neuraxial anesthesia [3].
Incidence and risk factors for SEH after neuraxial anesthesia — The incidence of spinal epidural hematoma (SEH) following neuraxial anesthesia is unknown, but likely very low, based on retrospective studies with small numbers of hematomas [4-9]. Risk factors consistently identified include hemostatic abnormalities (eg, drug induced coagulopathy, thrombocytopenia, renal failure, preeclampsia), advanced age, female sex, osteoporosis, spine abnormalities, epidural rather than spinal anesthesia (thought to be due to larger needle gauge with epidural), administration of multiple drugs that affect hemostasis, and difficult neuraxial procedures [4-6,10,11]. Some of these specific risks are discussed in the discussions of individual drugs below. Older adult patients are thought to be at increased risk because of the high incidence of spinal abnormalities (eg, spinal and foraminal stenosis) such that a small volume bleed may lead to spinal cord or cauda equina compression.
The risk of SEH related to neuraxial anesthesia is greatest when a patient's hemostatic system is abnormal either at the time a needle is placed in the neuraxial space or at the time of removal of a continuous neuraxial catheter. In all patients on medication affecting hemostasis, the timing of neuraxial anesthesia must be coordinated with discontinuation and resumption of antithrombotic medications to minimize the risk of significant bleeding from disrupted epidural blood vessels. The specific medication and the timing of the last dose are critical pieces of information for the anesthesiologist planning neuraxial procedures. Patients receiving more than one medication affecting hemostasis generally should not receive neuraxial anesthesia or analgesia. (See 'Guidelines for timing of neuraxial anesthesia procedures' below and 'Multiple antithrombotic drugs' below.)
The incidence of SEH after neuraxial anesthesia among obstetric patients appears to be lower than in non-obstetric patients, based on retrospective database and literature reviews [8,10]. (See "Deep vein thrombosis in pregnancy: Epidemiology, pathogenesis, and diagnosis", section on 'Risk factors'.)
Examples of studies of incidence of SEH include the following, although these may not reflect the true incidence due to under-reporting:
●In a 2004 retrospective study of neuraxial anesthetics over a decade in Sweden, 25 SEHs occurred after approximately 450,000 epidural anesthetics (1:18,000), and eight occurred after 1,260,000 spinal anesthetics (1:158,000) [4]. The incidence of hemorrhagic complications varied significantly depending on the population, from 1 in 200,000 for obstetric patients to 1 in 3600 for female knee arthroplasty patients who received epidural or combined spinal-epidural anesthesia. Of the 33 hematomas, 11 (33 percent) were in patients with coagulopathy or who had an antithrombotic drug administered in temporal proximity to the block, 10 (30 percent) had difficult placement, and 6 (18 percent) had spine pathology.
●In a 2013 review of outcomes of epidural catheterization from the Multicenter Perioperative Outcomes Group, SEH requiring laminectomy for SEH evacuation occurred in 7 out of 62, 450 (1:8900) surgical patients, and in none of 79,837 obstetric patients [5]. Four patients who developed SEH had received anticoagulant or antiplatelet medications in a way that deviated from the existing American Society of Regional Anesthesia and Pain Medicine (ASRA) guidelines, two had difficult neuraxial placement, and one patient was undergoing dialysis and thus may have been at increased risk because of platelet dysfunction.
●In a single center retrospective database review of SEH after surgical or obstetric epidural catheterization from 2001 to 2009, SEH occurred in 6 of 43,200 patients (1:7200), none of whom were obstetric patients [10]. All patients with SEH were over the age of 70, and all received perioperative anticoagulation.
●There are a number of case reports of SEH in obstetric patients who were coagulopathic at the time of epidural catheter placement or removal [4,12,13]. However, a systematic review of the literature from 1952 to 2016 found no cases of SEH after neuraxial anesthesia in obstetric patients that were thought to be related to thromboprophylaxis [8].
Typical presentation — SEH is not clinically apparent until pain or neurologic deficits appear. The most common presenting symptoms of a neurologically-significant SEH are a progressive motor and sensory block (68 percent of patients) or bowel/bladder dysfunction (8 percent); unlike classical disc herniation, radicular pain is an uncommon presenting symptom in patients with SEH [14]. In the American Society of Anesthesiologists (ASA) Closed Claims analysis, 83 percent of patients with SEH presented with increased motor block, and 25 percent with back pain [15].
DECISION TO USE NEURAXIAL ANESTHESIA IN PATIENTS ON ANTITHROMBOTIC MEDICATION — The decision to use neuraxial anesthesia in patients who either have been or will be receiving antithrombotic medication must weigh the benefit of the neuraxial anesthetic against the risk of spinal epidural hematoma (SEH).
Surgical versus obstetric patients — The decision to use neuraxial anesthesia techniques in patients who receive antithrombotic therapy is somewhat different between surgical and obstetric patients.
●Surgical patients – Surgical patients on chronic antithrombotic medication usually have the medication discontinued, medication changed (eg, bridging), or dosing modified prior to elective procedures to minimize surgical bleeding; if antithrombotic medication must be continued without interruption, or for urgent procedures in patients who have received such medications, neuraxial anesthesia may not be an option.
Generally, the decision regarding whether or not to continue the anticoagulant up to the time of the procedure is made by weighing the risk of thrombosis if the anticoagulant is stopped versus the bleeding risk of the procedure, as discussed in more detail separately. (See "Perioperative management of patients receiving anticoagulants", section on 'Deciding whether to interrupt anticoagulation'.)
●Obstetric patients – Decisions regarding the use of neuraxial anesthesia or analgesia in obstetric patients who receive antithrombotic medications are more complex. The small risk of SEH must always be balanced against the patient specific risks of general anesthesia, which may be higher in obstetric patients (algorithm 1 and algorithm 2) [16]. (See "Anesthesia for cesarean delivery", section on 'General versus neuraxial anesthesia'.)
•The need for obstetric neuraxial analgesia or anesthesia is often unpredictable, and delay of neuraxial techniques to allow an appropriate interval from last administration of an antithrombotic medication may be undesirable or harmful. For patients who receive antithrombotic therapy during pregnancy, it may be beneficial to switch from low molecular weight heparin (LMWH) to unfractionated heparin (UFH), or to reduce the antenatal dose of UFH, near the end of pregnancy (eg, at approximately 36 weeks) to facilitate neuraxial procedures [16]. (See 'Unfractionated heparin (UFH)' below and 'Low molecular weight heparin (LMWH)' below and "Use of anticoagulants during pregnancy and postpartum", section on 'Labor and delivery'.)
•Neuraxial anesthesia is the technique of choice for cesarean delivery and for labor analgesia; general anesthesia for cesarean delivery is associated with increased intraoperative and postoperative adverse events for both mother and fetus compared with neuraxial anesthesia. Neuraxial labor analgesia may be initiated early in labor specifically to reduce the need for general anesthesia for an unplanned cesarean delivery in select patients (eg, patients with severe preeclampsia or predicted difficult airway). (See "Anesthesia for cesarean delivery", section on 'General versus neuraxial anesthesia' and "Neuraxial analgesia for labor and delivery (including instrumented delivery)", section on 'Preparation for surgical anesthesia'.)
SPINAL VERSUS EPIDURAL TECHNIQUE — When neuraxial anesthesia is used in patients on antithrombotic medication, use of a single-shot spinal technique rather than an epidural may decrease the risk of spinal epidural hematoma (SEH), as estimates of risk are greater with larger needles (epidural compared with spinal) and in techniques with continuous catheters (continuous epidural or continuous spinal) [4]. (See 'Incidence and risk factors for SEH after neuraxial anesthesia' above.)
MULTIPLE ANTITHROMBOTIC DRUGS — We agree with recommendations from the American Society of Regional Anesthesia and Pain Medicine (ASRA) that neuraxial techniques should not be performed in patients who receive more than one antithrombotic medication simultaneously [17].
This includes any combination of anticoagulants such as unfractionated heparin (UFH), low molecular weight heparin (LMWH), warfarin, a direct oral anticoagulant, and/or antiplatelet agents, such as a nonsteroidal antiinflammatory drug (NSAID; including aspirin, but not COX-2 inhibitors), a P2Y12 receptor blocker, or dipyridamole.
For patients with indwelling epidural catheters, the patient's medical record should be reviewed daily for all medications that affect hemostasis. The decision to maintain an epidural catheter in patients who receive anticoagulant or antiplatelet medication (including combinations of such medications) should always involve a risk/benefit analysis. In some cases, the epidural catheter should be removed prior to administration of a second medication that affects hemostasis.
As an example, although a spinal or epidural technique may be performed in a patient who has received aspirin as the sole antithrombotic medication up until the day of the procedure, an epidural catheter should not be maintained postoperatively if that same patient receives LMWH or warfarin after surgery, while the aspirin effect is still present (typically considered to be approximately five to seven days since the last dose of aspirin) (see 'LMWH thromboprophylaxis' below and 'Postoperative initiation of warfarin' below). Similarly, neuraxial anesthesia should not be performed in patients who have taken aspirin up to the day of the procedure if they will receive intraoperative intravenous heparin.
The risk of spinal epidural hematoma (SEH) in patients who receive low dose subcutaneous heparin (ie, 5000 units subcutaneously two to three times daily) along with antiplatelet medication is unknown, but likely increased. Whereas SEH has not been reported in patients who have received low dose thromboprophylaxis with UFH along with antiplatelet medication, administration of multiple anticoagulant medications is a known risk factor for SEH in patients who receive higher dose UFH. (See 'Incidence and risk factors for SEH after neuraxial anesthesia' above.)
Avoidance of neuraxial anesthesia in patients who have taken more than one medication that affects hemostasis is based on the following types of indirect evidence:
●Increased bleeding from other sites has been reported with combination antithrombotic therapy [18].
●Case series of patients with SEH have included many patients receiving more than one antithrombotic medication [14,15,19,20].
GUIDELINES FOR TIMING OF NEURAXIAL ANESTHESIA PROCEDURES
Overview of timing of neuraxial anesthesia — Guidelines have been developed for the timing of neuraxial procedures and removal of neuraxial catheters relative to administration of antithrombotic medications, and timing of administration of antithrombotic medications after such procedures, to minimize the risk of hemorrhage [16,17,21,22]. The guidelines represent a consensus of expert opinion, and are often based on case reports, case series, and knowledge of the pharmacokinetics, pharmacodynamics, and actions of each drug. Thus, the specific recommendations in these guidelines may vary somewhat (table 1 and table 2).
For newer drugs for which there is less clinical experience, recommendations are often based on elimination half-lives. Some experts and organizations recommend an interval of two drug half-lives between the last dose and performance of neuraxial anesthesia, particularly for patients with a high risk of thrombosis. This strategy would mean that a low level of anticoagulation may be present when neuraxial anesthesia is performed. A more conservative approach would be to wait for five elimination half-lives, which would result in approximately 97 percent elimination; such a strategy may be more appropriate for patients with a low risk of thrombosis. Management decisions should ultimately be individualized based on the patient specific risks of bleeding versus thrombosis.
Unless noted, our recommendations are consistent with the 2018 American Society of Regional Anesthesia and Pain Medicine (ASRA) guidelines for the use of regional anesthesia in patients on various antithrombotic and thrombolytic medications [17], which are shown in a table (table 1).
Guidelines have also been published by anesthesiology societies in Europe (table 2) [21-25] as well as the US Food and Drug Administration (FDA) for specific drugs. (See 'Society guideline links' below.)
Significant differences between the ASRA recommendations and those of other organizations are noted in the text below, and in the table (table 1).
UNFRACTIONATED HEPARIN (UFH) — The effect of UFH can be monitored with the activated partial thromboplastin time (aPTT) or by anti-factor Xa activity. (See "Clinical use of coagulation tests", section on 'Monitoring heparins'.)
If necessary, UFH can be reversed with protamine. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Unfractionated heparin'.)
Due to the risk of heparin-induced thrombocytopenia (HIT) for patients who have received intravenous (IV) or subcutaneous UFH heparin for four or more days, platelet count should be checked prior to performing neuraxial anesthesia and/or prior to neuraxial catheter removal. In addition, periodic platelet count monitoring may be appropriate, depending on the baseline risk of HIT, as discussed separately (see "Heparin and LMW heparin: Dosing and adverse effects", section on 'Platelet count monitoring'). HIT may occur in 0.1 to 5 percent of surgical patients who are exposed to UFH for more than 4 days; the risk in obstetric patients is approximately 0.1 to 1 percent. American Society of Regional Anesthesia and Pain Medicine (ASRA) recommends checking a platelet count for patients who have received heparin for more than four days, and more than five days, respectively (table 1).
Intravenous UFH — Anticoagulation with IV heparin may be required at any point in the perioperative period, including after performance of neuraxial procedures, and with epidural catheters in place. Risk factors for spinal epidural hematoma (SEH) during heparin therapy include traumatic needle placement, coagulation disorders, concomitant administration of other drugs that affect hemostasis, and administration of heparin less than one hour after neuraxial needle placement [19,20].
The anticoagulant effect of IV heparin is immediate. Metabolism of heparin is complex and dose dependent, with a half-life of approximately 45 minutes to one hour [26,27]. When heparin is stopped preoperatively, the effect is expected to be largely eliminated after five half-lives (approximately four to five hours). (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Elective procedure/minor bleeding'.)
●Preoperative IV UFH – Neuraxial anesthesia techniques may be used in patients who have been receiving therapeutic anticoagulation with IV UFH (aPTT >1.5 to 2 times baseline level), if the heparin can be temporarily discontinued, and it is clinically acceptable to return to normal coagulation status for several hours both for epidural or spinal insertion, and for removal of a continuous catheter. Heparin infusion should be stopped for at least 4 to 6 hours and a normal aPTT should be documented prior to neuraxial anesthesia or catheter removal [28].
●IV UFH bolus during surgery – IV UFH should not be administered for at least one hour after a neuraxial anesthesia procedure or neuraxial catheter removal. Vascular surgery often requires IV administration of UFH (typically 5000 to 10,000 units) prior to clamping large vessels. There is extensive experience with the safe use of therapeutic intraoperative UFH in patients with epidural catheters, when the heparin was given at least one hour after the neuraxial catheter placement [29].
●Postoperative IV UFH – When epidural analgesia is used in conjunction with IV UFH, use of the smallest effective concentrations of local anesthetic will allow for earlier recognition, imaging, and management should motor or sensory loss be caused by SEH. (See 'Prevention of neurologic damage from spinal hematoma' below.)
The epidural catheter should be removed four to six hours after the last dose of IV UFH and after a normal aPTT is documented. IV UFH should not be administered for at least one hour after removal of an epidural catheter.
●Traumatic neuraxial needle or catheter placement – After a traumatic (bloody) neuraxial procedure, therapeutic heparinization probably increases the risk of SEH compared with the use of lower dose prophylactic heparin, although the magnitude of risk is unknown [28]. Intraoperative heparinization is mandatory for some surgical procedures. When neuraxial needle or catheter placement is traumatic (bloody) or requires multiple attempts, concerns about subsequent anticoagulation should be discussed with the surgeon, including the possibility of delaying the case for 24 hours. In the case of an urgent surgical procedure that requires heparinization, sensory and motor neurologic function should be closely monitored around the time of expected recovery from anesthesia. Using the lowest effective concentrations of local anesthetic in the postoperative period facilitates monitoring of motor blockade and earlier detection of SEH. (See 'Neurologic monitoring' below.)
●Cardiopulmonary bypass – The use of neuraxial techniques for postoperative analgesia (eg, epidural analgesia or neuraxial opioids) for cardiac surgery that requires cardiopulmonary bypass is controversial. Higher dose IV UFH is required for these procedures, compared with other surgery, which may increase the risk of SEH. Spontaneous SEH has been reported after heparinization for cardiopulmonary bypass [30], but SEH related to neuraxial anesthesia in these patients has not been reported, and the incidence of this complication is unknown [31-33].
Subcutaneous UFH — The effect of prophylactic dose subcutaneous unfractionated heparin (UFH) on the aPTT (and therefore the risk of bleeding) is variable. The onset of effect of subcutaneous heparin occurs in 20 to 30 minutes and peaks at two to four hours, with considerable individual variation [27]. Subcutaneous heparin is administered two or three times per day. Up to 7 percent of patients develop an aPTT >1.5 times the normal level, especially those who are frail and/or older adults, and small number of patients become fully anticoagulated after a dose of 5000 units subcutaneously [34,35].
Recommendations for the timing of neuraxial procedures and catheter removal depend on the dose of subcutaneous UFH administered. Higher doses are often used for pregnant patients.
●Low dose thromboprophylaxis (5000 units subcutaneously two or three times daily) – Neuraxial anesthesia techniques or catheter removal should be performed at least four to six hours after the last dose, or after confirmation of normal aPTT or anti-factor Xa level prior to the procedure.
●Higher dose thromboprophylaxis (7500 to 10,000 units subcutaneously twice daily, total daily dose ≤20,000 units) – Neuraxial anesthesia or catheter removal should be performed at least 12 hours after last dose and after confirmation of normal aPTT or anti-factor Xa level.
●Therapeutic dose UFH (individual dose >10,000 units subcutaneously or total daily dose >20,000 units) – Neuraxial anesthesia or catheter removal should be performed at least 24 hours after last dose and after confirmation of a normal aPTT or anti-factor Xa level.
Low dose (5000 units) subcutaneous UFH may be administered immediately after performance of a neuraxial procedure or removal of a neuraxial catheter.
LOW MOLECULAR WEIGHT HEPARIN (LMWH) — Low molecular weight heparin (LMWH) is increasingly used rather than unfractionated heparin (UFH) for thromboprophylaxis and for therapeutic anticoagulation. Compared with UFH, LMWHs have longer half-lives, are associated with a lower incidence of heparin-induced thrombocytopenia (HIT), cannot be fully reversed with protamine, and cannot be assessed with an aPTT [36]. Anti-factor Xa levels where available can be measured to assess LMWH effect, but safe levels for performance of neuraxial anesthesia have not been determined. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'LMW heparin'.)
A number of cases of spinal epidural hematoma (SEH) have been reported in patients who had neuraxial anesthesia while receiving LMWH at various doses. The labels of all anticoagulants, including LMWHs, include a boxed warning from the US Food and Drug Administration (FDA) regarding this risk. In addition, in 2016 the FDA issued a safety warning for the timing of neuraxial anesthesia in patients receiving LMWHs to decrease the risk of bleeding [37]. The recommendations were prompted by report of 100 confirmed spinal hematomas between 1992 and 2013 in patients who had spinal or epidural anesthesia and were receiving prophylactic dose enoxaparin. Twice daily dosing of enoxaparin was identified among half of the reported cases. (See 'Incidence and risk factors for SEH after neuraxial anesthesia' above.)
The available LMWHs differ with respect to their chemical properties, doses, and half-lives. Thus, recommendations for the timing of neuraxial anesthesia are based on the specific drug, dose, and frequency of administration. Perioperative bridging anticoagulation with LMWH may be prescribed at prophylactic, intermediate, or therapeutic dosing levels. (See "Perioperative management of patients receiving anticoagulants", section on 'Heparin product and dose'.)
The risk of HIT is lower with LMWH than with unfractionated heparin. HIT occurs in 0.1 to 1 percent of postoperative patients who receive thromboprophylaxis with LMWH, whereas medical or obstetric patients who receive LMWH have a risk of HIT of <0.1 percent. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Platelet count monitoring'.)
To evaluate for HIT, a platelet count should be checked prior to neuraxial anesthesia in patients who have received LMWH for four or more days. In addition, subsequent periodic platelet count monitoring may be appropriate, depending on the baseline risk of HIT, as discussed separately (see "Heparin and LMW heparin: Dosing and adverse effects", section on 'Platelet count monitoring'). Similarly, American Society of Regional Anesthesia and Pain Medicine (ASRA) recommends checking a platelet count for patients who have received LMWH for more than four days. However, practice varies, and other organizations make different recommendations because of the low risk of HIT with LMWH. As an example, the 2017 Society for Obstetric Anesthesia and Perinatology consensus statement on anesthesia for patients who receive anticoagulants does not recommend routine platelet count monitoring for obstetric patients who receive LMWH [16]. The American College of Chest Physicians (ACCP) Practice Guidelines do not recommend routine platelet count monitoring for patients with a risk of HIT <1 percent. Criteria for evaluating the likelihood of HIT and platelet count thresholds for neuraxial anesthesia are discussed in separate topic reviews. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia" and "Adverse effects of neuraxial analgesia and anesthesia for obstetrics", section on 'Neuraxial analgesia and low platelets' and "Anesthesia for the patient with preeclampsia", section on 'Coagulation'.)
LMWHs are cleared by the kidneys, and the FDA recommends doubling the intervals between LMWH dosing and neuraxial procedures, and between neuraxial procedures and subsequent dosing, for patients with creatinine clearance (CrCl) <30 mL per minute.
When neuraxial needle or catheter placement is traumatic (bloody) or requires multiple attempts, LMWH may be dosed no sooner than 24 hours after the placement or attempt, and these concerns should be discussed with the surgeon.
Therapeutic LMWH — Patients who are receiving therapeutic doses of low molecular weight heparin (LMWH) are at greater risk of SEH with neuraxial needle or catheter placement than those treated with prophylactic doses. Additional caution is required in patients on long-term LMWH and in patients with renal failure, which may lead to accumulation of anticoagulant effect and fibrinolysis [38].
Therapeutic doses of LMWH include:
●Enoxaparin (Lovenox) 1 mg/kg every 12 hours
●Enoxaparin 1.5 mg/kg daily
●Dalteparin (Fragmin) 100 to 120 units/kg every 12 hours
●Dalteparin 200 units/kg daily
●Tinzaparin (Innohep, Logiparin) 175 units/kg daily
●Nadroparin (Fraxiparine) 86 units/kg every 12 hours
●Nadroparin 171 units/kg daily
ASRA guidelines recommend performing a neuraxial anesthetic no sooner than 24 hours after the last dose of LMWH (therapeutic) [17]. Guidelines from the European Society of Anesthesiology and Intensive Care/European Society of Regional Anesthesia (ESAIC/ESRA) also recommend an interval of 24 hours between the last dose of high dose LMWH, except in patients with CrCl < 30 mL/min, for whom they recommend waiting 48 hours or until the anti-factor Xa activity is ≤ 0.1 IU/mL [22].
The first postoperative dose of therapeutic dose LMWH should not be administered until at least four hours after neuraxial catheter removal and a minimum of 24 hours after the initial needle/catheter placement. Neuraxial catheters should be removed prior to initiation of therapeutic dosing.
LMWH thromboprophylaxis — Many surgical patients receive LMWH for thromboembolism prophylaxis. Neuraxial anesthesia (including continuous catheters) may be used in these patients if they are not receiving any other medications that affect hemostasis. Recommendations are summarized in the table (table 1).
Prophylactic doses include:
●Enoxaparin 30 mg every 12 hours
●Enoxaparin 40 mg daily
●Dalteparin 2500 to 5000 units daily
●Nadroparin 2850 units daily
●Nadroparin 38 units/kg daily
●Tinzaparin 50 to 75 units/kg daily
●Tinzaparin 3500 units daily
A neuraxial anesthetic may be inserted at least 12 hours after the last dose of LMWH (prophylactic). Postoperative twice daily dosing of enoxaparin is associated with an increased risk of SEH [17].
●Twice daily prophylactic LMWH – ASRA guidelines recommend removing the epidural catheter prior to initiation of twice daily prophylactic LMWH heparin therapy (enoxaparin 30 mg twice daily). The first postoperative dose should be administered no earlier than 12 hours after neuraxial procedure and at least four hours after catheter removal [17].
The European Society of Anesthesiology (ESA) guidelines allow for maintenance of an epidural catheter during twice daily prophylactic dosing, but, prior to catheter removal, one scheduled dose should be omitted, and the catheter removed 24 hours after the last dose [39].
●Once per day prophylactic LMWH
•For a single injection neuraxial procedure, the first postoperative dose of LMWH should be administered no earlier than 12 hours after the neuraxial procedure.
•Indwelling neuraxial catheters may be maintained during once daily prophylaxis as long as no other drugs that affect hemostasis are administered. For neuraxial catheters, the first postoperative dose of LMWH should be administered at least 12 hours after the neuraxial procedure, and the second dose should be administered no sooner than 24 hours after the first. The catheter should be removed at least 12 hours after the last dose of LMWH, and the subsequent dose should be administered at least four hours after catheter removal.
WARFARIN
General issues — The mechanism of warfarin action differs from other anticoagulants, both at the time of anticoagulant discontinuation and initiation, when effects on clotting factors may not be accurately reflected in the prothrombin time/international normalized ratio (PT/INR). The response to an initial dose of warfarin depends on the dose and patient factors that may increase sensitivity (eg, female sex, advanced age, frailty, liver or kidney disease, dietary vitamin K deficiency, and certain medications). The biology of warfarin and its effect on clotting factors during discontinuation and initiation of therapy are discussed in detail separately. (See "Warfarin and other VKAs: Dosing and adverse effects", section on 'Initial dosing'.)
The INR is not specifically validated for documenting normal coagulation status, and the "normal range" for the INR is not well-defined. Some individuals have a baseline INR slightly above 1 (range 1.2 to 1.3). In an individual who has been taking warfarin, an INR of 1.3 may represent that individual's baseline INR, or it may signify that their coagulation status has not returned to normal. This variation is reflected in the various INR cutoffs for neuraxial anesthesia used by different societies (eg, "normal," INR is used in the American Society of Regional Anesthesia and Pain Medicine (ASRA) and European Society of Anesthesiology and Intensive Care/European Society of Regional Anesthesia (ESAIC/ESRA) guidelines, and INR <1.4 is used in the British guidelines) [17,21]. In practical terms, the cutoff used to allow neuraxial anesthesia should be based on institutional guidelines, laboratory normal ranges, and the judgment of the individual performing the procedure. This subject is discussed in more detail separately. (See "Clinical use of plasma components", section on 'Settings in which plasma is not appropriate'.)
Preoperative warfarin discontinuation — For individuals receiving warfarin preoperatively, the drug needs to be discontinued with sufficient time to allow coagulation status to return to normal before performing neuraxial anesthesia. In rare cases (eg, for urgent or emergency surgery), a reversal agent such as vitamin K or a prothrombin complex concentrate (PCC) may be used (see "Perioperative management of patients receiving anticoagulants", section on 'Warfarin interruption'). Recommendations from anesthesia societies regarding neuraxial anesthesia in patients who have received or will receive warfarin vary in significant ways, as discussed below.
Spinal or epidural anesthesia or analgesia may be used in patients who have discontinued warfarin, ideally at least five days earlier (ie, last dose is given on day minus 6), or who have had warfarin reversed, if coagulation status is normal as indicated by a PT within (or below) the normal range for the testing laboratory, or INR ≤ 1.2, at the time of neuraxial needle or catheter insertion. (See "Perioperative management of patients receiving anticoagulants", section on 'Warfarin interruption'.)
Postoperative initiation of warfarin — The use of neuraxial techniques in patients who receive warfarin for postoperative thromboprophylaxis is controversial. The anesthesia societies disagree on the performance of neuraxial anesthesia techniques, and on the use of epidural catheters postoperatively, in patients who receive warfarin for postoperative thromboprophylaxis. ASRA guidelines allow for performance of neuraxial anesthesia techniques within 24 hours after a single dose of warfarin and for maintenance of neuraxial catheters during initiation of postoperative warfarin thromboprophylaxis [17], whereas the ESA [39] and British guidelines [21] recommend removing an epidural catheter prior to administration of postoperative warfarin.
Warfarin was widely used for postoperative thromboprophylaxis before the advent of LMWH and direct oral anticoagulants for this purpose. There are two case reports of spinal epidural hematoma (SEH) in patients who had epidural catheters removed during full postoperative anticoagulation with warfarin, one with an INR over 6 [40], and another who had the catheter removed after four days of warfarin with a PT of 17.3 (normal range 11.2 to 14.4) [41]. There are several large observational studies on the concomitant use of postoperative epidural analgesia with initiation of warfarin. While they provide some measure of reassurance that epidural analgesia may be used safely during early initiation of warfarin therapy, much larger studies would be required to draw definitive conclusions, given the rarity of SEH.
●In a multi-institutional study including 4365 patients who underwent total hip or knee arthroplasty with postoperative epidural analgesia and thromboprophylaxis with warfarin, and in all cases with removal of the epidural catheter with an INR >1.4, there were no cases of SEH [42]. The only antithrombotic medication in effect at the time of surgery was a nonsteroidal antiinflammatory drug (NSAID) in some patients. In all patients, the first warfarin dose was administered on the night of surgery, at a dose that varied with the individual institutional protocol. Most catheters were removed on postoperative day 2 (4090 patients). Mean INR at the time of catheter removal was 1.9, with a range of 1.5 to 5.9.
●In a single institution retrospective database review of records from approximately 13,000 patients who underwent total hip or knee arthroplasty with epidural analgesia and postoperative warfarin, there were no cases of SEH [43]. The first dose of warfarin was administered the night of surgery. The epidural catheter was removed within 48 hours of surgery unless the INR was >2, in which case removal was delayed. Among 1038 case records that were reviewed in detail, the mean INR at the time of catheter removal was 1.54 (range 0.93 to 4.25).
We agree with the ASRA recommendations regarding the management of neuraxial catheters during administration of warfarin for postoperative thromboprophylaxis, as follows:
●The epidural catheter should be removed prior to administration of warfarin in patients who have received any other medication that affects hemostasis, including aspirin or other NSAIDs.
●Spinal or epidural anesthesia and analgesia may be initiated or an epidural catheter may be removed without documentation of a normal PT or INR when a single warfarin dose was given <24 hours earlier, before an anticoagulant effect has occurred. If a second dose has been given, or >24 hours have elapsed since the dose, a normal PT or INR should be documented prior to placing a neuraxial needle or catheter. (See "Warfarin and other VKAs: Dosing and adverse effects", section on 'Initial dosing'.)
●The PT/INR should be checked daily in individuals receiving warfarin who have a neuraxial catheter in place.
●An epidural catheter kept in place beyond 24 hours after a dose of warfarin should be removed only when the INR is <1.5.
●ASRA makes no recommendation regarding catheter removal in patients with INR between 1.5 and 3. We remove epidural catheters in these patients as long as other drugs affecting hemostasis were not used in combination with warfarin, and perform neurologic monitoring for 24 to 48 hours after catheter removal, or until the INR has stabilized at the desired level.
●For patients with an INR >3 (likely supratherapeutic), we recommend holding or reducing the warfarin dose in patients with indwelling epidural catheters, and delaying catheter removal until the INR is <3. Management of patients with a supratherapeutic INR is discussed separately. (See "Management of warfarin-associated bleeding or supratherapeutic INR".)
●Sensory and motor function should be monitored closely during epidural analgesia for patients receiving warfarin, and for 24 hours after catheter removal. We suggest using the lowest effective concentration of local anesthetic to facilitate evaluation of sensory and motor neurologic function. (See 'Neurologic monitoring' below.)
DIRECT ORAL ANTICOAGULANTS (DOACS) — DOACs include oral direct factor Xa inhibitors and the direct thrombin inhibitor dabigatran. (figure 1).
Direct oral factor Xa inhibitors — There is limited experience with direct oral factor Xa inhibitors and neuraxial anesthesia. Spinal epidural hematoma (SEH) has been reported with the use of these agents [17], but the risk factors and incidence have not been determined, nor has the exact timing after a dose to reach a level of anticoagulant effect safe for neuraxial anesthesia.
American Society of Regional Anesthesia and Pain Medicine (ASRA) recommends avoiding spinal or epidural techniques until nearly all of the medication has been eliminated (eg, five elimination half-lives for 97 percent elimination). For all of the oral direct factor Xa inhibitors, ASRA recommends an interval of 72 hours between the last dose and neuraxial anesthesia, despite widely differing half-lives for different factor Xa inhibitors. (See "Management of bleeding in patients receiving direct oral anticoagulants", section on 'Interval since last dose'.)
The European Society of Anesthesiology and Intensive Care/European Society of Regional Anesthesia (ESAIC/ESRA) guidelines make different recommendations for patients who receive low versus high dose rivaroxaban or apixaban, as described in the sections below (table 2) [22].
Routine laboratory testing is not used for determining anticoagulation status in patients receiving direct factor Xa inhibitors. Prolonged coagulation tests can be helpful in determining residual anticoagulant effect, but a normal prothrombin time/international normalized ratio (PT/INR) cannot be used as evidence that the anticoagulant effect has resolved. Specific assays for anti-factor Xa activity that are calibrated to the individual anticoagulant may be used if available in a timely fashion; if this testing is normal, then the result can be presumed to show that there is no residual anticoagulant effect. (See "Management of bleeding in patients receiving direct oral anticoagulants", section on 'Assessment of anticoagulation status'.)
The US Food and Drug Administration (FDA) has issued black box warnings on the risk of SEH with neuraxial anesthesia in patients treated with the direct oral factor Xa inhibitors [44].
●Rivaroxaban (Xarelto) – Rivaroxaban has a half-life of five to nine hours, which may be prolonged to 11 to 13 hours in older patients. Dose adjustments at required for patients with renal or hepatic insufficiency. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Rivaroxaban'.)
The ASRA guidelines suggest that neuraxial anesthesia should be performed no sooner than 72 hours after the last dose of rivaroxaban. A rivaroxaban specific anti-factor Xa level may be measured, but a safe level of activity (other than zero) has not been determined. The first postoperative dose should be administered at least six hours after a neuraxial procedure or catheter removal; neuraxial catheters should be removed prior to the first postoperative dose.
In the event of that rivaroxaban is mistakenly given to a patient with a neuraxial catheter in place, rivaroxaban should be withheld for 22 to 26 hours, or a rivaroxaban specific anti-factor Xa level should be measured to confirm lack of effect, prior to catheter removal.
The ESAIC/ESRA guidelines define low dose rivaroxaban as 2.5 mg twice daily or 10 mg once daily; they define high dose as 15 mg twice daily or 20 mg once daily when used for venous thromboembolism (VTE) treatment or atrial fibrillation [22]. They consider 20 mg once daily when used for extended thromboprophylaxis to be low dose [22]. The guideline recommends waiting at least 24 hours after low dose, and at least 72 hours after high dose rivaroxaban before a neuraxial procedure. However, for individuals receiving the 20 mg once daily dose, it may be prudent to observe an interval of 72 hours, regardless of indication, based on the pharmacokinetics of the drug. The first postoperative dose should be administered at least six hours after the procedure or catheter removal, longer after a traumatic procedure.
The FDA warning stipulates that if a traumatic neuraxial procedure occurs, administration of rivaroxaban should be delayed for at least 24 hours [45].
●Apixaban (Eliquis) – The half-life of apixaban is approximately 12 hours (range, 8 to 15 hours), and may be prolonged in patients with renal dysfunction and in older patients. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Apixaban'.)
ASRA guidelines suggest that neuraxial anesthesia should occur no sooner than 72 hours after the last dose of apixaban. The first postoperative dose should be administered at least six hours after a neuraxial procedure or catheter removal; neuraxial catheters should be removed prior to the first postoperative dose. If an unanticipated dose of apixaban is administered while a neuraxial catheter is in place, apixaban should be withheld for 26 to 30 hours, or an apixaban-specific anti Xa level should be measured. The FDA warning on apixaban stipulates that if a traumatic neuraxial procedure occurs, apixaban should be withheld for at least 48 hours [46].
The ESAIC/ESRA guidelines define low dose apixaban as 2.5 mg twice daily (as long as certain criteria are absent, such as age ≥80 years, body weight <60 kg, or chronic kidney impairment), and high dose as 5 or 10 mg twice daily [22]. The guideline recommends waiting at least 36 hours after low dose apixaban, and at least 72 hours after high dose apixaban, before performing a neuraxial procedure. The first postoperative dose should be administered at least 6 hours after the procedure or catheter removal, longer after a traumatic procedure.
●Edoxaban (Savaysa, Lixiana) – The half-life of edoxaban is approximately 6 to 11 hours. It is primarily excreted by the kidneys, and dose adjustment is recommended for patients with renal insufficiency. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Edoxaban'.)
ASRA and ESAIC/ESRA guidelines suggest that neuraxial anesthesia should occur no sooner than 72 hours after the last dose of edoxaban [17,22]. The first postoperative dose should be administered at least six hours after a neuraxial procedure or catheter removal; neuraxial catheters should be removed prior to the first postoperative dose. If an unanticipated dose of edoxaban is administered while a neuraxial catheter is in place, edoxaban should be withheld for 20 to 28 hours, or an edoxaban specific anti Xa level should be measured, before the catheter is removed.
Reversal of direct factor Xa inhibitors is discussed separately. (See "Management of bleeding in patients receiving direct oral anticoagulants", section on 'Factor Xa inhibitors'.)
Dabigatran (Pradaxa) — Dabigatran is an oral direct thrombin inhibitor. Dabigatran is administered for postoperative thromboprophylaxis after orthopedic surgery, prevention and treatment of deep vein thrombosis, and stroke prevention in nonvalvular atrial fibrillation. There are no reported cases of SEH related to neuraxial anesthesia, but there are reports of a spontaneous SEH [47] and one related to a fall [48] in patients taking dabigatran. The FDA has issued a black box warning for SEH with neuraxial anesthesia or lumbar puncture, without recommendation for optimal timing.
The half-life of dabigatran is approximately 12 to 17 hours in patients with normal kidney function. It is highly dependent on renal excretion, and the ASRA's recommendations for anesthetic management are based partly on estimates of the patient's renal function. For intervals less than those specified in the recommendations, a dilute thrombin time (dTT) or ecarin clotting time may be used to determine residual dabigatran activity, though safe levels have not been determined. (See "Clinical use of coagulation tests", section on 'Monitoring argatroban and hirudin' and "Management of bleeding in patients receiving direct oral anticoagulants", section on 'Coagulation testing'.)
If necessary, dabigatran can be reversed with idarucizumab, which is discussed separately. (See "Management of bleeding in patients receiving direct oral anticoagulants", section on 'Dabigatran reversal'.)
ASRA recommends the following:
•For patients in whom creatinine clearance (CrCl) has not be determined, perform neuraxial anesthesia no sooner than five days after the last dose.
•For patients in whom CrCl has been determined, and without other risk factors for bleeding (eg, age >65 years, hypertension, concomitant antiplatelet therapy), recommendations are as follows:
-For patients with CrCl <30 mL/min who have received dabigatran, do not perform neuraxial anesthesia.
-For patients with CrCl 30 to 49 mL/min perform neuraxial anesthesia no sooner than five days after the last dose.
-For patients with CrCl 50 to 79, perform neuraxial anesthesia no sooner than four days after the last dose.
-For patients with CrCl ≥80 mL/min, perform neuraxial anesthesia no sooner than three days (72 hours) after the last dose.
•The first postoperative dose of dabigatran should be administered at least six hours after a neuraxial procedure or catheter removal; neuraxial catheters should be removed prior to the first postoperative dose.
•If an unanticipated dose of dabigatran is administered while a neuraxial catheter is in place, dabigatran should be withheld for 34 to 36 hours or a dTT or ECT assessed before the catheter removal.
The ESAIC/ESRA guidelines make a distinction between low and high dose dabigatran [22], as follows:
•For patients who receive low dose dabigatran (220 mg once daily, or 150 mg once daily if CrCl is 30 to 50 mL/minute or age ≥75), perform neuraxial anesthesia at least 48 hours after the last dose.
•For patients who receive high dose dabigatran (150 mg twice daily, or 110 or 150 mg twice daily if CrCl is < 50 mL/min or age 75 to 80, or 110 mg twice daily if age ≥ 80), perform neuraxial procedure at least 72 hours after the last dose. For patients with CrCl <50 mL/min, perform neuraxial anesthesia after direct thrombin inhibitor level <30 ng/mL or thrombin time is normal. In contrast with the ASRA guidelines, ESAIC/ESRA recommends administering the first postoperative dose at least 24 hours after a neuraxial procedure.
FONDAPARINUX (ARIXTRA) — Fondaparinux is an injectable drug (subcutaneous) that may be used as an alternative to heparin for postoperative venous thromboembolism prophylaxis. It inhibits factor Xa indirectly by binding to antithrombin. Fondaparinux is a long-acting drug with plasma half-life of 17 to 21 hours [49], which is prolonged in patients with renal dysfunction [26]. (See "Prevention of venous thromboembolism in adults undergoing hip fracture repair or hip or knee replacement", section on 'Approach to pharmacologic thromboprophylaxis (low-risk bleeding)' and "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients", section on 'Pharmacologic dosing'.)
Experience with fondaparinux in patients who undergo neuraxial anesthesia is limited. During initial dose finding studies, one case of spinal epidural hematoma (SEH) was reported in a patient who received high dose fondaparinux after difficult epidural catheter placement (five attempts) [50].
For patients who have received fondaparinux, the European Society of Anesthesiology and Intensive Care/European Society of Regional Anesthesia (ESAIC/ESRA) guidelines recommend that neuraxial anesthesia should not be performed until 36 hours after the last dose (or until 72 hours after the last dose if the CrCL is <50 mL/minute) [22]. American Society of Regional Anesthesia and Pain Medicine (ASRA) guidelines make no recommendation on preoperative administration; they recommend that fondaparinux should not be administered for postoperative thromboprophylaxis after difficult neuraxial procedures (ie, more than one needle pass, difficulty threading a neuraxial catheter), and should not be administered for patients with an indwelling neuraxial catheter. A neuraxial catheter should be removed at least six hours prior to the first postoperative dose of fondaparinux.
Fondaparinux-specific anti-factor Xa assays are available, but optimal therapeutic levels and safe levels for performance of neuraxial anesthesia have not been determined.
PARENTERAL DIRECT THROMBIN INHIBITORS — This class of anticoagulants includes argatroban and bivalirudin, which are used for anticoagulation in patients with heparin-induced thrombocytopenia (HIT). Neuraxial anesthesia is not recommended in patients who are receiving these medications. The anticoagulant effects of direct thrombin inhibitors may be monitored with an activated partial thromboplastin time (aPTT). The parenteral direct thrombin inhibitors do not have specific reversal agents, but their half-lives are very short and thus their effect on coagulation status reverses relatively soon after discontinuation.
Neither the American Society of Regional Anesthesia and Pain Medicine (ASRA) nor the European Society of Anesthesiology and Intensive Care/European Society of Regional Anesthesia (ESAIC/ESRA) guidelines make recommendations on the timing of neuraxial anesthesia after discontinuation of argatroban or bivalirudin [17,39]. The ESIAC/ESRA guidelines note that these drugs are primarily used in critically ill patients who are not likely to receive regional anesthesia. Nevertheless, the 2010 ESRA guidelines suggested delaying neuraxial anesthesia for at least four hours after discontinuation of argatroban, and waiting two hours after a neuraxial procedure or catheter removal to administer a subsequent dose [39]. Regardless of the interval, a normal aPTT should be documented prior to neuraxial anesthesia.
ANTIPLATELET DRUGS — The antiplatelet drugs include aspirin, nonsteroidal antiinflammatory drugs (NSAIDS), the direct and indirect platelet P2Y12 receptor blockers, the platelet GP IIb/IIIa receptor antagonists, and other individual antiplatelet agents. Antiplatelet therapy is increasingly used for management and prevention of cardiac and cerebrovascular ischemia. (See "Long-term antithrombotic therapy for the secondary prevention of ischemic stroke" and "Acute non-ST-elevation acute coronary syndromes: Early antiplatelet therapy" and "Acute ST-elevation myocardial infarction: Antiplatelet therapy" and "Long-term antiplatelet therapy after coronary artery stenting in stable patients".)
Aspirin and other nonsteroidal antiinflammatory drugs — As single drugs, aspirin and other NSAIDs can be used in patients receiving spinal or epidural anesthesia or analgesia; there is no evidence of an increased risk of spinal epidural hematoma (SEH) [51].
In patients receiving other antithrombotic medications, aspirin should be stopped 7 to 10 days prior to neuraxial anesthesia if clinically appropriate (because aspirin causes dysfunction for the life of the platelet), and other NSAIDs should be stopped three days before a neuraxial anesthetic to allow for full return of platelet function [52]. Cyclooxygenase-2 (COX-2) inhibitors have minimal effects on platelet function and may be used together with antithrombotic medications, with the exception of warfarin, in patients having neuraxial anesthesia. The concomitant use of COX-2 inhibitors and warfarin may increase the risk of hemorrhagic complications [53]. (See 'Multiple antithrombotic drugs' above.)
Recommendations regarding aspirin differ in the American Society of Regional Anesthesia and Pain Medicine (ASRA) and European Society of Anesthesiology and Intensive Care/European Society of Regional Anesthesia (ESAIC/ESRA) guidelines. The ASRA guidelines do not recommend changes to the timing of neuraxial procedures or catheter removal in patients receiving aspirin alone [17]. In contrast, the ESAIC/ESRA guidelines recommend discontinuing high dose aspirin (defined as ≥200 mg/day) for three days in patients with normal platelet counts and seven days in others, prior to performing neuraxial anesthesia, and waiting at least six hours after neuraxial procedures or catheter removal before administering the first postoperative dose [22]. The variation in opinions between the guideline panels may relate to different interpretations of clinical or laboratory data regarding the mechanism of action of aspirin at different doses.
Platelet P2Y12 receptor blockers — The thienopyridines (clopidogrel, prasugrel, and ticlopidine) irreversibly antagonize ADP-induced platelet aggregation by inhibition of the P2Y12 receptor after metabolism via hepatic CYP 450 system [54]. Their effects last 4 to 10 days after the drugs have been discontinued. (See "Platelet biology", section on 'Clopidogrel and ticlopidine'.)
Ticagrelor reversibly inhibits platelet aggregation by a direct action on the platelet P2Y12 receptor. Spontaneous SEH has been reported during clopidogrel therapy [55], and SEH has been reported after neuraxial anesthesia during clopidogrel or ticlopidine therapy [56-58].
The ASRA guidelines for neuraxial anesthesia with these medications are consistent with guidelines from the ESAIC/ESRA and the American College of Chest Physicians (ACCP) [17,22,59], as follows:
●The time intervals between the last dose of the medication and neuraxial anesthesia should be as follows:
•Clopidogrel (Plavix) – 5 to 7 days
•Prasugrel (Effient) – 7 to 10 days
•Ticlopidine (not available in the United States) – 10 days
•Ticagrelor (Brilinta) – 5 to 7 days
●Neuraxial catheters should not be maintained with prasugrel or ticagrelor because of rapid onset of action. Epidural catheters may be maintained for one or two days after reinstitution of therapy with clopidogrel or ticlopidine, as long as a loading dose is not administered; if a loading dose is to be administered, it should be administered at least six hours after a neuraxial procedure or catheter removal.
Cangrelor (Kengreal) is a short acting P2Y12 inhibitor that is administered by infusion. It may be used for bridging therapy after oral agents are discontinued. ASRA suggests that cangrelor should be discontinued at least three hours prior to neuraxial anesthesia. Neuraxial catheters should be removed prior to starting cangrelor therapy, and the first postoperative dose should be administered at least eight hours after a neuraxial procedure or catheter removal.
Platelet GPIIb/IIIa receptor antagonists — These are intravenous drugs that block platelet aggregation (see "Platelet biology", section on 'GPIIb/IIIa inhibitors'). They are used primarily for therapy of acute coronary syndrome, usually in combination with aspirin and heparin. Neuraxial anesthesia should be avoided for four to eight hours after administration of eptifibatide (Integrilin) or tirofiban (Aggrastat) and for 24 to 48 hours after abciximab.
Use of these drugs is contraindicated for four weeks after major surgery [60-62]; patients should be carefully monitored neurologically if a dose is given in the month following a neuraxial needle or catheter placement, and any neuraxial catheter infusions adjusted to eliminate motor block and allow neurologic monitoring.
Cilostazol (Pletal) — Cilostazol reversibly inhibits platelet aggregation by inhibiting PDE IIIA. It is used primarily for peripheral arterial disease. The elimination half-life is approximately 11 to 13 hours and is prolonged in patients with renal failure [63]. SEH has been reported following epidural catheter removal during cilostazol therapy [64]. The risk of bleeding associated with neuraxial anesthesia during cilostazol therapy is unknown. Based on elimination half-life, ASRA suggests that neuraxial anesthesia should be performed no sooner than two days after the last dose. The first postoperative dose of cilostazol should be administered at least six hours after a neuraxial procedure or catheter removal; neuraxial catheters should be removed prior to the first postoperative dose.
Dipyridamole (Persantine) — Dipyridamole impairs platelet function by inhibiting the activity of adenosine deaminase and phosphodiesterase, and may be used for stroke prevention, particularly in combination with aspirin. The half-life of elimination is 10 to 12 hours. (See "Long-term antithrombotic therapy for the secondary prevention of ischemic stroke", section on 'Aspirin plus dipyridamole'.)
Spinal hematoma has not been reported after neuraxial anesthesia, but hematomas have been reported after peripheral nerve block during dipyridamole therapy [65,66].
ASRA suggests that neuraxial anesthesia should be performed no sooner than 24 hours after the last dose of extended release dipyridamole. The first postoperative dose should be administered at least six hours after a neuraxial procedure or catheter removal; neuraxial catheters should be removed prior to the first postoperative dose.
THROMBOLYTIC OR FIBRINOLYTIC THERAPY — Plasminogen activators (streptokinase, urokinase, alteplase, tenecteplase) activate plasminogen to plasmin to dissolve fibrin clot, resulting in reduced levels of circulating plasminogen and fibrinogen. Despite relatively short plasma half-lives, the thrombolytic effect of these drugs may persist for days. The incidence of spinal epidural hematoma (SEH) in patients receiving thrombolytic/fibrinolytic therapy is unknown. However, both spontaneous spinal or epidural hematoma and neuraxial anesthesia-related SEH have been reported during or prior to thrombolytic therapy [67-72]. There are no data addressing the length of time that neuraxial anesthesia should be avoided after discontinuation of these drugs. A cautious approach is to avoid spinal or epidural techniques for five elimination half-lives (ie, 97 percent elimination). The American Society of Regional Anesthesia and Pain Medicine (ASRA) suggests avoiding neuraxial anesthesia for 48 hours after discontinuation of thrombolytic/fibrinolytic medications, and documentation of normal clotting parameters and fibrinogen level. In patients who require thrombolytic therapy near the time of neuraxial block, frequent neurologic monitoring should performed to assess for signs of SEH.
No recommendations are available to guide removal of a neuraxial catheter in patients who received thrombolytic therapy after initial catheter placement. In this extremely rare situation, the fibrinogen level may be measured to assist in timing of catheter removal, though a safe level of fibrinogen has not been determined.
HERBAL PRODUCTS — Several widely used herbal products affect platelet function, including garlic, ginkgo, and ginseng, although the effects are not well quantified and may vary with different preparations. There are reports of spontaneous neuraxial bleeding in patients who have ingested garlic [73] and ginkgo biloba [74], and reports of other significant bleeding in patients who were taking ginkgo biloba. The risk of bleeding theoretically may be further increased if herbal substances (eg, garlic, ginseng, ginger, ginkgo biloba) are taken together or in combination with medications that affect hemostasis. (See "Clinical use of ginkgo biloba", section on 'Safety' and "Clinical use of ginkgo biloba", section on 'Potential interactions'.)
There are no reported cases of spinal epidural hematoma (SEH) after neuraxial anesthesia thought to be related to the use of herbal products. Therefore, we agree with the guidelines from American Society of Regional Anesthesia and Pain Medicine (ASRA) that herbal products do not need to be withheld prior to neuraxial anesthesia. The joint guidelines on interventional pain procedures in patients taking antiplatelet and anticoagulant medications recommends discontinuation of herbal products that affect hemostasis prior to high-risk interventional procedures (eg, spinal cord simulator placement, intrathecal catheter placement, vertebral augmentation), but not before intermediate-risk procedures (eg, epidural steroid injection, sympathetic blocks) or low-risk procedures (eg, peripheral nerve blocks, trigger point injections, implantable pulse generator replacement) [1].
There are no data regarding the risk of SEH in patients who take a combination of herbal products or the combination of herbal product with antithrombotic medication.
PREVENTION OF NEUROLOGIC DAMAGE FROM SPINAL HEMATOMA — The best chance of neurologic recovery after spinal epidural hematoma (SEH) depends on rapid diagnosis and management. Strategies for reducing the risk of SEH are shown in a table (table 3). (See 'Typical presentation' above.)
Neurologic monitoring — Sensory and motor function should be monitored in all patients receiving neuraxial anesthesia or analgesia, whether or not they receive antithrombotic therapy. (See "Management of acute perioperative pain in adults", section on 'Monitoring the patient who has received neuraxial analgesia'.)
Frequent checks (ie, every one to two hours) are necessary during the 6 to 12 hours after initiation of a new anticoagulant regimen for patients with an epidural catheter in place. Neurologic checks can occur less frequently thereafter. Any change in neurologic status that is not clearly explained by epidural bolus or change in local anesthetic concentration should immediately trigger clinical assessment and further evaluation.
During epidural analgesia, use of the smallest effective concentration of local anesthetic will allow for earlier recognition of sensory or motor changes.
Evaluation, management, and prognosis — Emergency magnetic resonance imaging (MRI; or emergency computed tomography scan [CT] if MRI is contraindicated or unavailable) is recommended as soon as SEH is suspected. If SEH is detected, any residual anticoagulant should be reversed rapidly if appropriate (ie, if anticoagulant effect is present). Reversal strategies are discussed separately. (See "Management of warfarin-associated bleeding or supratherapeutic INR", section on 'Serious/life-threatening bleeding' and "Management of bleeding in patients receiving direct oral anticoagulants", section on 'Major bleeding'.)
Emergency neurosurgical consultation should be arranged to evaluate for decompressive surgery. Neurologic recovery is more likely if decompressive laminectomy is performed within eight hours of symptom onset. In a retrospective case series of 61 SEHs, patients with decompressive surgery less than eight hours after onset of paraplegia had better neurologic outcomes (6/13 good recovery, 4/13 partial recovery, 3/13 poor recovery) than if surgery was over 24 hours after symptom onset (2/12 good recovery, 10/12 poor recovery) [28]. Even with prompt diagnosis and decompression, many patients have permanent neurologic deficits. The final neurologic outcome depends on [28]:
●The time span between hematoma formation and surgical decompression
●The speed with which the hematoma develops
●The severity of the preoperative neurologic deficit
●The size of the hematoma
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Local and regional anesthesia".)
SUMMARY AND RECOMMENDATIONS
●Risk of SEH with neuraxial procedures – Patients using anticoagulants and certain antiplatelet agents are at increased risk for spinal epidural hematoma (SEH) after neuraxial anesthesia. The risk is estimated to be 1 in 18,000 for epidural anesthetics, and 1 in 158,000 for spinal anesthetics. Additional risk factors include bleeding diatheses, timing of antithrombotic drugs in relation to neuraxial needle placement or catheter removal, difficult or traumatic (bloody) placement, spinal abnormalities, female sex, and possibly older age. (See 'Incidence and risk factors for SEH after neuraxial anesthesia' above.)
●Strategies to reduce risk – Patients with multiple risk factors for SEH may have the risk decreased by using a smaller needle (spinal rather than epidural) and by avoiding a continuous catheter technique. (See 'Decision to use neuraxial anesthesia in patients on antithrombotic medication' above and 'Spinal versus epidural technique' above.)
●Recommendations for specific medications
•More than one antithrombotic – Use of more than one antithrombotic medication increases the risk of SEH. In patients with more than one antithrombotic medication (including aspirin or another nonsteroidal antiinflammatory drug [NSAID] in effect at the time of the procedure), we recommend avoiding neuraxial anesthetic techniques (Grade 1C). (See 'Multiple antithrombotic drugs' above.)
•Anticoagulants – Patients receiving an anticoagulant during the periprocedural period require attention to the timing of drug administration and drug dosing when a spinal or epidural catheter is used; this includes unfractionated and low molecular weight heparin (LMWH), warfarin, direct oral anticoagulants (dabigatran, apixaban, edoxaban, rivaroxaban), and fondaparinux. Recommendations based on the guidelines of the American Society of Regional Anesthesia (ASRA; 2018) are summarized in the table (table 1). Recommendations from the European Society of Anaesthesiology and Intensive Care and the European Society of Regional Anaesthesia are summarized in a separate table (table 2). (See 'Guidelines for timing of neuraxial anesthesia procedures' above.)
•Antiplatelet agents – There should be a delay of 5 to 7 days between clopidogrel or ticagrelor and neuraxial anesthesia; longer intervals are required for prasugrel and ticlopidine and shorter intervals for dipyridamole and GPIIb/IIIa inhibitors. Single agent aspirin or another NSAID does not appear to increase the risk of SEH after a neuraxial technique and does not require discontinuation. (See 'Antiplatelet drugs' above.)
•Herbal products – Patients using herbal products that affect platelet function (eg, garlic, ginkgo, and ginseng) may be considered for neuraxial anesthesia since there is no evidence of increased risk of SEH. Administration of combinations of herbal medicines that are thought to affect hemostasis or administration of these medicines along with other drugs that affect hemostasis may increase the risk of bleeding. (See 'Herbal products' above.)
●Monitoring for complications – Patients having neuraxial anesthesia who have received any antithrombotic drug in the periprocedural period need frequent neurologic examinations, particularly noting motor and sensory function at and below the level of the neuraxial technique. Especially in patients with multiple risk factors for SEH, the lowest effective dose of local anesthetic for perioperative analgesia will allow for earlier recognition of motor or sensory loss caused by SEH. (See 'Neurologic monitoring' above and 'Typical presentation' above.)
●Evaluation for symptoms – Patients with significant symptoms leading to suspicion of SEH should have emergency imaging with magnetic resonance imaging (MRI) or computed tomography (CT), and emergency neurosurgical evaluation for possible decompressive surgery. Long-term neurologic outcome of SEH is better if decompressive surgery is performed less than eight hours after symptom onset. (See 'Evaluation, management, and prognosis' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff thank Dr. John Stanec for his contributions as an author to previous versions of this topic review.
