Spinal/Epidural Hematoma after Neuraxial Anesthesia

Bleeding is a recognized complication associated with the placement of a regional anesthetic block in the anticoagulated patient. The most significant complication, however, is the development of a spinal axis hematoma. The true incidence of neurologic complications caused from bleeding after spinal axis anesthesia is unknown; however, Tryba, in his classic article, reported the estimated incidence to be less than 1 per 220,000 for spinal anesthesia and less than 1 per 150,000 for epidural anesthesia. A newer and more detailed study by Moen and colleagues has shown that the predictions made by Tryba on the incidence of spinal and epidural hematomas were probably correct when all patients receiving spinal or epidural blocks are considered; however, if special subsets, such as elderly women undergoing TKR, are analyzed, the original calculations by Tryba grossly underestimate the incidence by as much as almost 100-fold.

In a review published in 1994 by Vandermeulen and colleagues, the following possible risk factors were discussed for 61 previously reported spinal hematomas in patients receiving neuraxial anesthesia between 1906 and 1994:

• •
  

  At the time of anesthetic administration, 42 of the 61 patients (68%) developing spinal hematoma had impaired coagulation. In 25 of 42 of the cases, some form of heparin therapy was present. An additional five of 42 patients had undergone a major vascular procedure in which heparin was likely used but not reported. The remaining 12 of 42 patients had a variety of medical conditions that could have produced an impairment in their ability to form a quality clot. These conditions included thrombocytopenia, hepatic dysfunction, and renal insufficiency, or they had been treated with another anticoagulant/antiplatelet agent at the time the bleeding occurred.

  
  
• •
  

  The needle placement was reported as difficult in 15 of 61 patients (25%) and/or it was bloody in another 15 (25%) of the cases.

  
  
• •
  

  Multiple punctures were reported in 12 of 61 (20%) of the cases.

  
  
• •
  

  Pregnancy was noted in five of 61 (8%) of the cases.

  
  
• •
  

  Anatomic abnormalities, such as spina bifida occulta and vascularized tumor, were noted in four of 61 (6.5%) of the cases.

  
  
• •
  

  An epidural technique was used in 46 of 61 (75%) of the cases, and an epidural catheter was placed in 32 of 46 (70%). In 15 of 32 (47%) of the epidural catheter cases, the bleeding occurred immediately on removal of the catheter.

  
  
• •
  

  A spinal technique was involved in 15 (25%) of the cases.

The Vandermeulen study has two major shortcomings: it is a retrospective review of the literature and does not evaluate any primary data; of more importance, probably less than one in 10 adverse events that occur are ever reported in the literature.

More recently, Moen and colleagues conducted a retrospective review of all central neuraxial blocks placed in Sweden between 1990 and 1999. The study encompassed two phases. First, a postal survey letter was sent to the chairperson of all anesthesia departments in which they were asked to provide the number of spinal and epidural blocks placed in their department during 1998. In addition, they were asked to provide the number of block-related complications that occurred in their department during the decade 1990-1999. The specific complications that the study addressed were epidural hematoma, epidural abscess, meningitis, and cauda equina syndrome. The researchers then went to the National Board of Health and Welfare (NBHW) and reviewed the quality assurance files associated with each complication. By Swedish law all serious complications must be reported to the NBHW. During the study period, Moen and colleagues ascertained that 1,260,000 spinal blocks were performed and 450,000 epidural blocks were administered, including 200,000 labor epidural blocks. As a result of these blocks, 127 serious complications occurred, and 85 of 127 of these patients sustained permanent neurologic damage. Of the 127 complications, 33 were spinal axis hematomas, 32 were cauda equina syndrome, 29 were meningitis, 13 were epidural abscesses, and 20 were miscellaneous. The results of the Moen study mirror those of Tryba in that the incidence of complications after epidural blockade was much more frequent than after spinal blockade. In addition, more complications than expected were found by Moen and colleagues. The incidence of epidural hematoma in association with a labor epidural block was quite low (1 : 200,000), whereas those placed in women undergoing knee arthroplasty were very high (1 : 3600) and mirror the predictions made by Schroeder during the First ASRA Consensus Conference in 1998. Perhaps the most alarming information reported by Moen and colleagues is the fact that “one third of all spinal hematomas were seen in patients receiving thromboprophylaxis in association with a central neuraxial block (CNB) in accordance with the current guidelines and in the absence of any previously known risk factors.” Consequently, adherence to the guidelines regarding LMWH and CNB may reduce but not completely abolish the risk of spinal hematoma after CNB. This latter fact is further reinforced by the case report by Sandhu and colleagues in which they essentially followed the ASRA guidelines and still had an epidural hematoma occur in their patient 1 day after the removal of her epidural catheter. However, it must be noted that the patient in the Sandhu case report had two risk factors. The patient was elderly, age 79 years, and female. In brief, the Sandhu article reports the placement of an epidural catheter on the third attempt, hours after the last 5000-unit dose of subcutaneous UH. The catheter was removed on the third postoperative day, 6 hours after the last 5000-unit dose of UH. The patient developed a symptomatic epidural hematoma the next day that required surgical evacuation. Her platelet count and aPTT were within normal limits at all times. Sandhu and colleagues highlighted the need for all clinicians to be vigilant about the timing of epidural placement and removal, even in patients receiving standard-dose UH therapy, and they encouraged the routine and repeated monitoring of coagulation status.

The aforementioned two studies (the Moen and Sandhu studies) in which all guidelines were followed suggest that the incidence of these and similar catastrophic events are still very likely underreported. However, this impression may be tempered by a recent meta-analysis by Kreppel and colleagues. Kreppel and colleagues analyzed 613 case studies of all neuraxial hematomas published between 1826 and 1996 and ascertained that, in about one third of the cases (29.7%), no etiologic factor could be identified as the cause of the bleeding. This idiopathic spontaneous hemorrhage group formed the largest group of patients who developed spinal/epidural hematomas. Spinal and epidural anesthetics placed in conjunction with anticoagulation therapy were actually the fifth most common cause of spinal/epidural hematomas, and spinal and epidural anesthetic procedures alone were the tenth most common etiologic factor. The second largest group comprised cases in which the patients were undergoing anticoagulation therapy (17%) in the absence of neuraxial blocks. Unlike Moen and colleagues, Kreppel and associates found elderly men between 55 and 70 years to be at the greatest risk of the development of spinal hemorrhage. In the Kreppel series, 64% of the patients were men; however, all causes of spinal hemorrhages in patients in the Kreppel study were included, not just causes of hemorrhages in those undergoing total joint replacement under spinal/epidural anesthesia in conjunction with perioperative anticoagulation. In this latter scenario, elderly women are clearly at greatest risk.

Spinal hematoma is a rare and catastrophic complication associated with both epidural and spinal anesthesia. It may occur with bleeding into either the epidural space or the subarachnoid space. The prominent epidural venous plexus accounts for the majority of hematomas being formed in the epidural space. In addition, the radicular vessels along nerve roots can bleed either into the intrathecal or epidural space.

Spinal hematoma is often occult, delaying both diagnosis and treatment. The presenting symptom of spinal hematoma is not always radicular back pain. Vandermeulen and colleagues found the presenting symptoms to be lower extremity weakness (46%), radicular back pain (38%), and paresthesia (14%). The diagnosis is frequently complicated and delayed because of residual paresthesia or anesthesia produced by the neuraxial block. The use of a short-acting local anesthetic agent for intraoperative anesthesia and then unilateral or motor block–sparing techniques for postoperative analgesia can avert such delays in diagnosis. There is likely a temporal relationship between the onset of paraplegia, surgical evacuation of hematoma, and recovery ( Table 50-1 ). Full recovery of neurologic function appears less likely if surgery is postponed or delayed for more than 8 to 12 hours. Similar to the patients in the Vandermeulen series, patients in the Kreppel series who underwent rapid diagnosis and surgical evacuation obtained the most ideal recovery of neurologic function. In the Kreppel study, 31 of 47 patients who received surgical treatment within 12 hours of the onset of their symptoms recovered completely (66%); more than half of the patients who did not obtain surgical decompression until 13 to 24 hours had elapsed did not recover any neurologic function. The exact treatment for a neuraxial hematoma, however, remains controversial. In the classic study by Moen and colleagues, five of the six patients who recovered total neurologic function were conservatively managed; that is, they did not have surgical evacuation of their hematomas once they were diagnosed. However, among the 27 other patients with hematomas who did not recover neurologic function, 11 received a laminectomy and six were under consideration for a decompressive laminectomy that was not ultimately undertaken because of the delay from the time of symptom recognition until the time of diagnosis. In light of the mixed messages contained in these three studies, if an epidural hematoma is suspected, promptly seek the advice of a consultant. Moreover, an emergent MRI and a neurosurgical consultation are imperative; the advice of consultants should be followed. Of note, it is possible for a select group of patients to recover totally from an epidural hematoma without surgery, and an epidural blood patch is, in fact, an epidural hematoma. On the other hand, the literature would suggest that a decompressive laminectomy is the treatment of choice if one is faced with this catastrophic complication.

Safety of Neuraxial Anesthesia in Patients Receiving Low-Dose Subcutaneous Unfractionated Heparin

Multiple studies have demonstrated the relative safety of neuraxial anesthetic techniques in the presence of DVT prophylaxis with low-dose subcutaneous UH; in addition, there is little increased risk of spinal hematoma associated with this therapy. Five series have been published involving more than 9000 patients receiving this therapy without any complications. Allemann and colleagues and Lowson and Goodchild similarly reported no cases of spinal hematoma in 204 epidural blocks and 119 spinal blocks in patients who had received 5000 units of UH subcutaneously 2 hours before needle placement. The large amount of data suggests that subcutaneous heparin for DVT prophylaxis is both safe and efficacious in patients undergoing lower extremity orthopedic procedures and general, urologic, and gynecologic operations with a neuraxial block.

Currently, only three cases of spinal hematoma after neuraxial block in the presence of low-dose subcutaneous UH have been reported in the literature, two of which involved a continuous epidural anesthetic technique. In one of these case reports, an epidural catheter was placed despite elevation of the patient’s aPTT. In another, blood was aspirated from the catheter during placement. In the last case, spinal anesthesia was attempted multiple times.

ASRA 2010 Guidelines for the Use of Neuraxial Anesthesia and Low-Dose Subcutaneous Unfractionated Heparin

The material presented at the third annual meeting of the ASRA in Vancouver, British Columbia, Canada, in April 2007 appears to have served as the basis for most of the changes made to the earlier (1998 and 2002 ) guidelines on the use of UH and LMWH in conjunction with neuraxial anesthesia and have acted as the platform for the drafting of the newest ASRA guidelines (2010).

During subcutaneous (mini-dose) prophylaxis (5000 units, twice daily), no contraindication exists to the use of neuraxial techniques. The risk of neuraxial bleeding may be reduced by delaying the heparin injection until 1 to 2 hours after the block, and it may be increased in debilitated patients or after prolonged therapy. For the provision of the best possible patient care, it is imperative that every patient’s chart is reviewed on a daily basis to determine that patients are not receiving concurrent medications such as LMWH, oral anticoagulants, or antiplatelet agents that could affect other components of the clotting cascade. Because heparin-induced thrombocytopenia may occur, patients receiving heparin for more than 4 days should have a platelet count assessed before neuraxial block.

• •
  

  Avoid neuraxial techniques in patients with other coagulopathies.

  
  
• •
  

  Heparin administration should be delayed for 1 hour after needle placement.

  
  
• •
  

  Remove the catheter 1 hour before any subsequent heparin administration or 2 to 4 hours after the last heparin dose.

  
  
• •
  

  Monitor the patient postoperatively to provide early detection of motor blockade, and consider the use of a minimal concentration of local anesthetic to enhance the early detection of a spinal hematoma.

  
  
• •
  

  Although a bloody or difficult neuraxial needle placement may increase the risk of neuraxial bleeding, data do not support mandatory cancellation of a case. Clinical judgment is needed. If a decision is made to proceed, full discussion with the surgeon and careful postoperative monitoring are warranted.

Safety of Neuraxial Anesthesia in Patients Receiving Therapeutic or Full-Dose Unfractionated Heparin

Therapeutic or full-dose management modalities usually involve the injection of moderate amounts (5000 to 10,000 units) of intravenous (IV) UH intraoperatively. Injection during vascular cases may prevent thrombus formation during arterial cross-clamping. Alternatively, 20,000 to 30,000 units of UH may be injected during a cardiac procedure to facilitate cardiac bypass. In both these situations, high levels of UH are transient.

Several studies have demonstrated that spinal or epidural anesthesia followed by systemic UH administration is relatively safe. Rao and El-Etr reported on the outcomes of 3146 patients receiving continuous epidural anesthesia and 847 patients receiving continuous spinal anesthesia for lower extremity vascular procedures. UH was administered 50 to 60 minutes after catheter placement to achieve an activated clotting time (ACT) of twice the normal value. The UH was given every 6 hours throughout the period of anticoagulation therapy, and the catheters were removed the next day, 1 hour before the administration of the next maintenance dose of UH. None of the patients developed spinal hematoma. This UH therapy was closely monitored, and catheters were removed when UH levels were relatively low.

In 1998 Liu and Mulroy reported on more than 1000 patients undergoing full intraoperative anticoagulation who had also received either a single-bolus spinal injection of opioids or an epidural opioid infusion without any incidence of spinal hematoma. The authors noted that communication with the surgeon regarding traumatic attempts and subsequent management of anticoagulation was critical. Similarly, in 1998 Sanchez and Nygard reported on 558 patients undergoing cardiac surgery who had epidural catheters placed following strict guidelines. These guidelines mandated placement of the epidural catheters the day before surgery, use of a paramedian approach, obtaining an initial normal coagulation profile, carefully screening for preoperative drug use, and limiting catheter placement to two attempts. No incidence of spinal hematoma occurred in this study.

Baron and colleagues published a retrospective review in 1987 that evaluated 912 patients who had received continuous epidural analgesia while undergoing major vascular reconstruction of a lower extremity. The patients all received transient, full anticoagulation with UH at a dose of 75 IU/kg, in addition to a maintenance dose of 1000 IU/hr. None of these patients developed neurologic evidence of spinal hematoma. In this review, 71% of the patients were male, the average age was 68.7 years, and the following hematologic studies were obtained preoperatively: hemoglobin level, platelet count, prothrombin time (PT), and aPTT. No reference was made to the timing of either catheter placement or removal.

The potential usefulness of thoracic epidural analgesia in patients undergoing cardiothoracic surgery has been shown in multiple studies. In 2000 Ho and colleagues published a statistical analysis suggesting that, at most, one spinal hematoma secondary to epidural catheter placement would occur for every 1520 patients receiving epidural analgesia for coronary bypass surgery. This analysis was based on a zero incidence of spinal hematoma in more than 1500 reported uses of epidural analgesia in patients undergoing cardiac surgery. Thus studies purporting the safety of epidural anesthesia in the fully anticoagulated patient may be tainted by small sample sizes and type II statistical error.

It is important to recognize that other members of the care team may institute an inappropriate therapeutic intervention with catastrophic results. In a 2004 case report, a junior intensive care house officer administered an antithrombotic medication to a pediatric patient who had a functioning epidural catheter in place. The patient had been ambulating before the administration of the alteplase. Almost immediately after the administration of the drug, the child developed severe back pain, and blood was noted in the epidural catheter. The house officer immediately removed the epidural catheter and within minutes the patient developed lower extremity sensory and motor losses. The anesthesia care team was promptly notified, and a timely laminectomy and clot evacuation resulted in total recovery of neurologic function in the child 6 weeks later. This case report reinforces the need for all members of the care team involved in complex cases to be familiar with the guidelines for the management of epidural or other indwelling catheters. Moreover, this event occurred after the Rosen team had placed and managed slightly more than 1500 epidural catheters in infants and children undergoing total heparinization and cardiopulmonary bypass.

Davignon and colleagues and Chaney questioned the benefits of neuraxial blocks in patients undergoing cardiopulmonary bypass. The risks and benefits must always be carefully balanced.

In an article on the risks of neuraxial techniques and UH, Ruff and Dougherty reported the occurrence of spinal hematomas in seven of 347 patients who had initially had signs of cerebral ischemia. After subarachnoid bleeding had been ruled out, each patient immediately underwent a diagnostic lumbar puncture with a 20-gauge needle, followed by the institution of IV UH therapy. Unfortunately, the amount of UH administered was not reported in the article. The authors concluded that traumatic needle placement, initiation of IV UH within 1 hour of lumbar puncture, and concomitant aspirin therapy were all risk factors that led to the development of the spinal hematomas.

The therapeutic benefits of UH are limited by an increased risk of bleeding, which is at least a partially dose-dependent phenomenon. To optimize the balance between efficacy and bleeding complications, physicians have adopted two dosing practices. The first is estimation of UH plasma concentrations using frequent serial evaluations of the aPTT, a relatively inexpensive laboratory test. However, with repeated serial testing, cost may become an issue. The second is continuous IV administration of UH, in an attempt to allow multiple rapid dosage adjustments guided by aPTTs.

ASRA 2010 Guidelines for Neuraxial Anesthesia and Full-Dose Unfractionated Heparin

Currently, insufficient data and experience are available to determine whether the risk of neuraxial hematoma is increased when neuraxial techniques are combined with the full anticoagulation of cardiac surgery. Postoperative monitoring of neurologic function and selection of neuraxial solutions that minimize sensory and motor block are recommended to facilitate detection of new or progressive neurodeficits.

Prolonged therapeutic anticoagulation appears to increase the risk of spinal hematomas, especially if combined with other anticoagulants or thrombolytics. Therefore neuraxial blocks should be avoided in this clinical setting.

• •
  

  If systemic anticoagulation therapy is begun with an epidural catheter in place, it is recommended that catheter removal be delayed for 2 to 4 hours after therapy discontinuation and evaluation of coagulation status. The concurrent use of medications that affect other components of the clotting mechanisms may increase the risk of bleeding complications for patients receiving standard heparin. These medications include antiplatelet medications, LMWH, and oral anticoagulants.

  
  
• •
  

  It is important to note that approximately half of the spinal hematomas that have involved epidural catheters have occurred on the removal of the catheter. Epidural catheter removal carries the same risk as catheter placement, and the same guidelines should be followed for both procedures.

European Guidelines for Neuraxial Blockade and Unfractionated Heparin

Tryba found a low incidence of spinal hematomas in the large numbers of European patients receiving a spinal anesthetic and concurrent anticoagulation therapy. On the basis of their unique experiences and available experimental data, two European countries have recently promulgated new guidelines regarding the dosing of the heparins and heparin-like drugs in patients receiving neuraxial anesthetics. The European Society of Anaesthesiology (ESA) and Belgian Association for Regional Anesthesia (BARA) have both recently updated their guidelines, and they are now the standard by which other European nations manage and monitor the use of UH, LMWH, fondaparinux, and rivaroxaban in conjunction with neuraxial and deep plexus blocks. Only a few minor differences exist between the ESA and BARA guidelines.