25

The most significant spinal complication resulting from systemic anticoagulant or antiplatelet therapy is spontaneous spinal hemorrhage (SSH). This complication, though rare, must be considered in patients with elevated coagulation cascade times (prothrombin time [PT]/partial thromboplastin time [PTT]) who present with local or referred spinal pain associated with neurologic deficits. In these patients, an urgent magnetic resonance imaging (MRI) is vital to confirm the diagnosis. Spinal hemorrhage may vary in location by anatomic compartment. Spinal epidural hematoma (SEH) is the most common SSH type and accounts for 75% of all such hemorrhages, whereas subdural and subarachnoid bleeding occur with much less frequency; hematomyelia, or bleeding within the spinal cord, is very uncommon.1

The estimated incidence of SSH in the general population is very difficult to delineate but is known to be very low. As of 1996, there had only been 613 total reported cases. Kreppel et al,² in the most comprehensive SSH review to date, performed a meta-analysis of these 613 patients. In their analysis they noted that 29.7% of spontaneous spinal hemorrhage was idiopathic in origin, whereas medicinal anticoagulation was recognized as the second most common etiology and accounted for 17% of SSH occurrences. Interestingly, given that patient anticoagulation could be implicated in only a minority of SSH instances, Kreppel et al speculated that anticoagulation alone is likely to be insufficient to promote SSH without an additional factor such as elevated vertebral venous plexus pressure. To further support the notion that anticoagulation is an exceedingly rare cause of SSH, Angstwurm and Frick3 found a frequency of spinal hematoma of only 1% in 10,441 patients who experienced hemorrhage of any type while on therapeutic anticoagulation. Another study found no spinal hemorrhage among a total of 3,126 patients who were followed for any complication of anticoagulation therapy.⁴

Thus, due to the extremely low incidence of SSH combined with a lack of high-quality data, it is not possible to define the specific relative risk of developing an SSH while on anticoagulation therapy. However, based on the available retrospective population studies, the risk would appear to be very low (< 1%) in patients with well-controlled anticoagulation therapy, and is likely not statistically more probable to occur than that in the general nonanticoagulated population.

Risk of Anticoagulant or Antiplatelet Agents as a Source of Spontaneous Spinal Hemorrhage in Patients with Spinal Column or Spinal Cord Disorders

Broadly, disorders of the spinal column or spinal cord can be categorized as traumatic injury of the spinal column or spinal cord injury (SCI), degenerative or spondylotic conditions, deformity, neoplasm, infection, or vascular lesions of the spine or spinal cord. In our review we attempted to determine the risks of anticoagulant and antiplatelet agents in the context of these various categories of spinal disorders. Unfortunately, there is a paucity of studies delineating the relative safety of anticoagulant or antiplatelet agents in the setting of distinct spinal pathologies.

Spondylotic or degenerative spinal disorders have a very high population prevalence and accordingly warrant specific discussion. In our review, we found no evidence to suggest that degenerative spinal disease inherently increases the risk of spinal hemorrhage in patients requiring therapeutic anticoagulation beyond the general population receiving anticoagulation. Similarly, there is no evidence that patients with spinal deformity are at an increased risk of spinal hemorrhage in the face of anticoagulant therapy. Thus, patients with spinal degenerative disease or deformity can be placed on anticoagulants or antiplatelet agents as indicated for other pertinent medical conditions.

Although there are anecdotal reports of spontaneous spinal hemorrhage in patients with spinal neoplasms on anticoagulants, given the sporadic case report nature of these data, there are no guidelines for management in this situation. However, as will be discussed in detail in a later section, it is clear that patients with malignant disease are at a high risk of VTE, and thus the risk-to-benefit ratio, although difficult to quantify specifically, would appear to be highly in favor of initiation of anticoagulant therapy as indicated for other medical conditions.

Spinal vascular malformations are very uncommon entities and intuitively may be perceived to have an increased relative risk of spinal hemorrhage in the setting of anticoagulation therapy. Interestingly, the medical literature contains no description of the relative risk of hemorrhage from a spinal vascular malformation while being anticoagulated. Accordingly, one must balance the natural history of the disease requiring anticoagulation versus the apparent very low increase in risk of spinal vascular malformation hemorrhage.

Before one can make an informed decision about the relative risk-to-benefit ratio underlying prophylactic postoperative anticoagulation for patients undergoing spinal surgery, it is imperative to understand the incidence of deep venous thrombosis (DVT) and pulmonary embolus (PE) without anticoagulation in this patient population. In an attempt to address this question, Cheng and colleagues5 performed a meta-analysis examining the risk of DVT and PE in the perioperative period for patients who did not receive prophylactic anticoagulation. Specifically, they attempted to define particular high-risk VTE subpopulations stratified according to the type of underlying etiologic spinal condition for which spinal surgical occurred.⁵ These authors noted that for patients undergoing elective operations for deformity in three studies, the rate of DVT was 5.3% and the rate of PE was 2.7%; in non-SCI spinal column trauma patients undergoing spinal surgery in two studies, the risk of DVT was 6.0% (it was 18% in one study); in patients undergoing surgery for degenerative conditions in seven studies, the rate of DVT was 2.0%. Fatal PE was reported only twice over the 14 total studies that were analyzed—once in a trauma patient and once in a degenerative disease patient after an anterior lumbar procedure. In summary, Cheng and colleagues concluded that the risk of DVT in elective spine surgery without chemical prophylaxis was 1 to 2% but up to 18% in the trauma population, whereas the risk of fatal PE was extremely low in elective spinal surgery (0.05%) but occurred in 2% of reported spinal surgery for spinal column trauma. It must be emphasized that there were a limited number of studies available for analysis, and thus low numbers in the different surgical groups significantly weakened the power of the subgroup analysis. Despite the low level of evidence associated with these numbers the apparent increased risk of VTE complications in those patients undergoing surgery for spinal column fractures is compelling. Accordingly, we feel that it is prudent to consider spinal column trauma patients, irrespective of the presence of SCI, as at a distinct increased risk of VTE.

Two additional reviews have attempted to define the prevalence of VTE in spinal surgery.6,7 Sansone et al⁶ noted a DVT incidence of 1.09% and a PE incidence of 0.06% in a heterogeneous group of patients undergoing elective spine surgery (a subset of whom received DVT prophylaxis). An older literature review by Catre⁷ reported a raw incidence of VTE complications in a heterogeneous group of patients undergoing elective spinal surgery as 7.1%. Catre noted that the reviewed studies were of poor quality and thus felt that the derived incidence was suspect.

The incidence of PE after major thoracolumbar spine surgery (posterior fusion, anterior fusion, combined procedures (anterior and posterior) in patients treated with elastic antiembolism stockings and sequential compression devices until ambulatory (but no chemoprophylaxis) was found to be 2.2% in another study.⁸ A combined anterior-posterior approach was noted to have a statistically significant higher incidence of PE (6.0%) than in patients undergoing posterior procedure alone (p < 0.01).⁸

Boakye et al⁹ used the National Inpatient Sample to study complications of 58,115 patients (with variable prophylaxis) undergoing spinal fusion for cervical spondylotic myelopathy. Findings included higher VTE in patients undergoing fusion for cervical spondylotic myelopathy (0.73%) versus fusion for cervical spondylosis alone (0.25%). Moreover, in patients undergoing fusion for cervical spondylotic myelopathy, posterior fusion had a higher rate of DVT/PE (1.38%) when compared with anterior fusion (0.60%). Other traditional VTE risk factors such as older age, prior VTE, immobilization, thrombophilia, and malignancy also increase VTE risk in spinal surgery. One retrospective study of more than 35,000 patients undergoing spinal surgery found the VTE risk to be 0.5% (95% confidence interval [CI], 0.4–0.5%) for nonmalignant disease, and 2.0% (95% CI, 1.4–2.6%) for malignant disease.¹⁰

To summarize, although the available studies are largely of poor quality, it is readily apparent that while there is a low (but nonnegligible) incidence of VTE complications in spinal surgery patients as a whole, there are distinct subgroups of spinal surgery patients (e.g., spinal column trauma, deformity correction, combined anterior-posterior surgical approach, and individuals with malignant disease) subjected to significant increased VTE risk.

Risk of Perioperative Spinal Hemorrhage in Patients Undergoing Spinal Surgery

Following the identification of particular spinal surgery groups at particular risk of VTE complications, one must also consider the risk of bleeding complications in the postoperative spinal surgery patient population before one can safely administer anticoagulant prophylaxis to ameliorate VTE risks. In particular, it is important to recognize subgroups of patients who are at higher risk of bleeding complications either to adjust therapy accordingly, or to provide closer surveillance and earlier investigation of potential bleeding complications. To this end, there have been two retrospective studies examining both the incidence and predisposing risk factors for spinal hemorrhage following spinal surgery.^15,16 Kou et al¹⁶ found 12 incidences of SEH in 12,000 spinal procedures (giving an incidence of 0.1%). Similarly, Awad et al15 found 32 incidences of SEH in nearly 15,000 operations (incidence of 0.2%). This incidence is consistent with previously reported estimates as reviewed by Aono and colleagues.¹⁷ Awad et al further identified significant risk factors for postoperative SEH as age > 60, more than five 5 operative levels, Intraoperative blood loss > 1 L, international normalized ratio (INR) > 2.0 in the first 48 hours postoperatively, and recent preoperative nonsteroidal anti-inflammatory drug (NSAID) use (although the exact timing was unspecified).¹⁵ Kou et al identified multilevel procedures (likelihood ratio [LR = 12.42) and coagulopathy (LR = 38.78) as having statistically significant higher risk for developing postoperative SEH.¹⁶ Factors analyzed in these studies^15,16 but not found to significantly affect the incidence of SEH were the use of drains, intraoperative durotomy, body mass index (BMI), smoking, and well-controlled anticoagulation for DVT or cardiovascular prophylaxis (though the exact regimens were not specifically delineated).

Intuitively, intraoperative drain insertion should lower the risk of SEH as excess blood is evacuated; however, the findings of these retrospective studies^15,16 agree with the only prospective study examining this topic,¹⁸ which did not show any benefit of drain insertion in terms prevention of SEH, mortality, or infection rate in patients undergoing single level laminectomy.

Most recently, a 2011 study examined the incidence of symptomatic SEH in over 6,000 patients undergoing spine surgery and found an overall incidence of 0.41%.¹⁷ Aono and colleagues¹⁷ further examined the incidence of SEH among different types of spinal surgical procedures and noted the following incidences: 0% (none of 1,568) in standard lumbar diskectomy, 0.50% (eight of 1,614) in lumbar laminectomy, 0.67% (eight of 1,191) in posterior lumbar interbody fusion (PLIF), 4.46% (five of 112) in thoracic laminectomy, 0.44% (four of 910) in cervical laminoplasty, and 0.21% (1 of 466) in anterior cervical decompression and fusion. Thus, although thoracic laminectomy appeared to be associated with an increased incidence of symptomatic SEH, likely related to the small thoracic spinal canal diameter relative to the cervical and lumbar spines (even following laminectomy), there was no observed increased SEH risk between “simple” spinal procedures such as lumbar laminectomy and “complex” procedures such as PLIF.¹⁷

Postoperative Venous Thromboembolic Prophylaxis for Spinal Surgery

There are very few randomized trials comparing the efficacy of prophylactic VTE strategies in spinal surgery patients. Those trials that do exist are limited by small sample size and the difficulties in detection of asymptomatic DVT.¹⁹ A meta-analysis including these trials as well as other neurosurgical trials (many of which also included a subset of spinal surgery patients) has also examined this topic.²⁵ Clinical guidelines reported by Gould et al¹⁹ in 2012 suggest inferring treatment decisions based on this pooled analysis.²⁵ This large mixed meta-analysis (7779 patients in 18 randomized trials and 12 cohort studies were included) offers an accurate picture of prophylactic intervention efficacy but is confounded by the presence of cranial neurosurgical patients. The authors found a 59% reduction in DVT (relative risk [RR], 0.41; 95% CI, 0.21–0.78) when using intermittent pneumatic compression (IPC) devices. The reduction in PE was not significant with IPCs. Also, there was a 40% relative risk reduction in DVT when comparing low molecular weight heparin (LMWH) versus compression stocking (RR, 0.60; 95% CI, 0.44–0.81). Other pooled analyses based on comparative prophylactic regimes did not show any significant difference between IPC versus compression stockings, LMWH versus IPC, LMWH versus unfractionated heparin (UFH), and UFH versus placebo.25

Further, the pooled neurosurgical data showed no significant increased risk in major hemorrhage when using heparin compared with mechanical methods. However, there was a trend toward increased risk of minor hemorrhage with LMWH compared with mechanical prophylaxis alone (RR, 2.06; 95% CI, 0.18–5.09).²⁵ Interestingly, a small prospective trial using only IPC devices found only one DVT/PE in 100 consecutive patients undergoing single-level anterior corpectomy/fusion.²⁶

Timing of Venous Thromboembolic Chemoprophylaxis Initiation for Patients Undergoing Spinal Surgery

There is a single retrospective cohort study that analyzed preoperative VTE chemoprophylaxis with respect to the incidence of postoperative DVT/PE and SEH.²⁷ The authors concluded that although preoperative prophylaxis with UFH or LMWH did not significantly increase the risk of postoperative SEH, it also did not reduce the incidence of postoperative VTE complications.

Is early postoperative DVT prophylaxis with LMWH associated with an increased risk of hemorrhage? Gerlach et al²⁸ retrospectively evaluated 1,954 patients undergoing a wide range of spine surgical procedures and in whom early (< 24 hours postoperative) subcutaneous LMWH prophylaxis was administered, and noted a SEH rate of 0.4%. Thus, it appears that early LMWH prophylaxis is not associated with an increased risk of hemorrhage postoperatively and, accordingly, is safe to start within 24 hours of surgery. This is in agreement with the mixed neurosurgical meta-analysis that did not show a difference in major hemorrhage rate when the first dose of prophylactic heparin was administered preoperatively, intraoperatively, or postoperatively.²⁵ Accordingly, as is reflected in Gould et al’s 2012 guidelines,¹⁹ it does appear that preoperative or early postoperative chemoprophylaxis is safe in patients who are at high risk of perioperative VTE complications.

Recommendations for Venous Thromboembolism Prophylaxis in Spinal Surgery Patients Without Spinal Column or Spinal Cord Injury

Based on the pooled analysis of Collen et al,²⁵ in combination with identification of particular subgroups of nontrauma spinal surgery patients at higher risk of VTE complications, Gould et al’s¹⁹ 2012 clinical guidelines make the following specific recommendations for VTE prophylaxis for patients without traumatic spinal column or cord injuries undergoing spinal surgery:

• For patients undergoing spinal surgery, mechanical prophylaxis, preferably with IPC, is recommended. There is no compelling evidence to recommend UFH or LMWH.

• For patients undergoing spinal surgery at high risk for VTE (including those with malignant disease and those undergoing surgery with an anterior or combined anterior-posterior approach), we suggest adding pharmacological prophylaxis (UFH or LMWH) to mechanical prophylaxis once adequate hemostasis is established and the risk of bleeding decreases.

Risk of Venous Thromboembolism in Spinal Surgery Patients with Traumatic Injury to the Spinal Column or Spinal Cord

Although traumatic injuries are heterogeneous in nature, they are universally associated with a high risk of VTE. Indeed, a review of the incidence of VTE complications in trauma patients noted a range from 5 to 63%, dependent on the specific type of traumatic injury, modality of prophylaxis, and the particular method of detection employed.11 Despite this tangible risk of VTE complications, decision making regarding prophylaxis in trauma patients remains challenging due to the potentially calamitous consequences of bleeding complications in the trauma population, particularly in cases of visceral, spinal, and head injury.

Numerous studies have found that spinal column fractures and SCI are potent risk factors for the development of VTE complications. This is reflected in their inclusion as individual risk factors in the VTE risk assessment profile (RAP) developed by Greenfield et al.³⁸ As has been discussed previously, Cheng et al⁵ noted a 6% risk of DVT in non-SCI spinal trauma patients not receiving prophylaxis, which was higher than any other subset of spinal surgery patients. Even in the face of various prophylactic regiments VTE risks remain high for non-SCI spinal trauma patients, with a pooled incidence of 2.2%.¹⁹ A systematic review by Velmahos et al³⁹ further noted that patients with spinal fractures had an odds ratio of 2.3 for VTE in comparison to other patients with trauma, a cohort that already has a recognized elevated risk.

Despite the elevated risk of VTE in spinal trauma patients, individuals with SCI represent the cohort of either spinal surgery patients or trauma patients at the highest risk of VTE complications. It is estimated that 67 to 100% of patients with acute SCI will develop VTE (symptomatic or asymptomatic) in the first 3 months following their injury.¹² VTE is a significant cause of morbidity and mortality in SCI patients and is responsible for 9.7% of all deaths in the first year following acute SCI.⁴⁰ Given the marked elevation of VTE risk, it is widely recognized the prophylactic regimens must be considered in the spinal trauma and SCI patient populations.

Risk of Perioperative Spinal Hemorrhage in Spinal Surgery Patients with Traumatic Injury to the Spinal Column or Spinal Cord

Very few studies have examined the bleeding complications associated with spinal column or spinal cord injury. To date, only three trials in patients with trauma reported bleeding complications in patients who did not receive chemical VTE prophylaxis. In those trials, the pooled risk of bleeding complication was 0.7%, a number that likely represents the lower baseline bleeding risk in the trauma population.¹⁹

Similarly, few studies have examined the bleeding complications of spinal column or SCI patients associated with VTE chemical prophylaxis. Christie et al,²⁴ in their review, noted a low rate of bleeding complications among SCI patients receiving VTE prophylaxis, ranging from 0 to 2.6% in a limited number of available studies. Boakye et al,¹³ in a 2008 National Inpatient Sample (NIS) study, retrospectively examined over 31,000 SCI patients with the goal of evaluating outcomes and complications, stratified according to the level of SCI and management (nonoperative versus laminectomy alone versus laminectomy and fusion). Though not explicitly stated, it is assumed that these patients were treated with some variety of VTE prophylactic strategy for an appropriate duration. DVT/PE occurred in 1.4% of nonoperative cases, 2.64% of laminectomy alone cases, and 2.46% in laminectomy and fusion cases. Postoperative hematoma occurred in 3.37% of laminectomy alone and 3.82% of laminectomy and fusion. Interestingly, postinjury hematoma was also reported in 0.84% of patients undergoing nonoperative management. These estimates are similar to the 1% incidence of symptomatic SEH in patients suffering spinal trauma that was quoted in a previous study.14 In this particular study, T7-T12 injuries had the highest incidence of postoperative hematoma at 4.78%, whereas non–fracture-associated SCI at that same levels had the lowest incidence at 1.5%. For operatively managed patients, the incidence of DVT/PE was lowest with nonfracture SCI at the level of T1-T6 (0%) and highest among fracture-related injuries at the same levels (4.71%).¹³ For nonoperatively managed patients, the incidence of DVT/PE was lowest with nonfracture cord injuries at the level of T7–12 (0.98%) and highest among fracture-related injuries at T1-T6 (3.3%).⁸ Overall, there was a trend toward lower incidences of DVT/PE and postoperative hemorrhage with cord injuries secondary to non–fracture-related mechanisms than those with fracture-related mechanisms.

Venous Thromboembolic Prophylaxis for Patients with Traumatic Injury to the Spinal Column or Spinal Cord

Given the high rate of VTE in these patient populations, combined with the apparent low rate of major bleeding complications, the literature supports the routine use of VTE prophylaxis for patients with spinal column or spinal cord traumatic injury. Of course, it must be recognized that there are commonly accepted relative contraindications to the use of chemical thromboprophylaxis that include severe head injury, nonoperatively managed liver or splenic injuries, renal failure, SCI associated with epidural hematoma, severe thrombocytopenia, and coagulopathy.¹⁹

Teasell et al²⁹ conducted a large systematic review and meta-analysis of VTE prophylaxis in SCI patients, specifically examining chemical prophylaxis, mechanical VTE prophylaxis, and combination methods. They analyzed 23 separate studies that met their predefined inclusion criteria, including 13 studies that examined VTE chemical prophylaxis in SCI patients. Ultimately, they were able to reach several conclusions regarding thromboprophylaxis in SCI patients. First, they concluded that prophylaxis with UFH of 5000 IU twice daily is not superior to placebo alone in preventing DVT post-SCI, based on the combined findings of one small randomized controlled trial (RCT)³⁰ and one nonrandomized trial.³¹ They further noted, on the basis of a single RCT,³² that dose-adjusted UFH appeared superior to standard (5000 IU) twice daily dosing in the prevention of DVT (31% to 7%) but had a significantly higher risk of bleeding complications (0 vs 7%). Teasell et al also concluded that LMWH is superior to standard dose UFH in preventing VTE and has a lower rate of bleeding complications, based on two RCTs^22,33 and two nonrandomized trials.^21,34 The LMWH used in the majority of patients in these trails was enoxaparin. This conclusion is slightly different from that reached in a meta-analysis by Ploumis et al,¹² who noted a similar odds ratio following thromboprophylaxis with UFH versus LMWH (2.8 and 2.7, respectively) for the development of PE. However, they state that UFH is significantly associated with increased bleeding risk compared with LMWH, though other heparin-related complications were not different between the two.

Other studies have attempted to determine whether there was superiority of a particular LMWH for VTE prophylaxis in the setting of SCI. One randomized trial of 129 patients determined that 30 mg of enoxaparin administered twice daily via a subcutaneous route showed no differences in VTE rates or bleeding complications in comparison with 40 mg of subcutaneous enoxaparin administered once daily.35 Similarly, Chiou-Tan et al,³⁵ in a randomized trial of 95 patients, noted no difference in either VTE or bleeding complications between dalteparin 5000 IU subcutaneously once daily versus enoxaparin 30 mg subcutaneously twice daily.

In addition to chemical VTE prophylaxis, mechanical prophylaxis has also been examined in SCI patients by Winemiller et al.³⁷ Based on this case series with multivariate analysis, there is limited evidence that IPC or a gradient elastic stocking (GES) independently decrease the risk of DVT/PE.³⁷ Interestingly, there is no high-quality evidence that chemical VTE prophylaxis combined with mechanical compression devices is superior to mechanical compression alone for acute SCI patients, although this scenario has been poorly studied.¹² However, in deference to the conclusion of Winemiller et al, this does at least imply that mechanical compression devices have some efficacy in VTE prevention and, accordingly, in situations where anticoagulation is contraindicated (such as intracranial hemorrhage, hemothorax, and other active bleeding), mechanical VTE prophylaxis should be implemented as soon as feasible. Similarly, as noted by Teasell et al,²⁹ there is little evidence that combination (i.e., mechanical and chemical) methods are superior to LMWH alone, simply because this particular paradigm has not been studied specifically in the SCI population. However, given the well-established significantly elevated risk of VTE in this patient population, combination VTE prophylaxis would intuitively be at least as effective, if not marginally more effective, in the prevention of VTE with no added bleeding risk. Accordingly, a combination prophylaxis paradigm should be considered in the acute spinal injured population, at least until a high-quality analysis is available to support or refute this practice.

Timing of Venous Thromboembolic Chemoprophylaxis Initiation for Spinal Surgery Patients with Traumatic Injury to the Spinal Column or Spinal Cord

Christie et al²⁴ conducted a systematic review with the goal of determining the ideal timing (i.e., after SCI or after subsequent spinal surgery) for initiation of VTE prophylaxis in the acute SCI patient. They subsequently only identified a single study that was specifically designed to determine the effects of timing of prophylaxis after acute SCI.²⁰ In this study, Aito and colleagues²⁰ compared 275 SCI patients divided into either an early VTE prophylaxis group (within 72 hours) or a late group (initiation after 72 hours). All patients received mechanical prophylaxis with IPC or GES, and chemical prophylaxis with LMWH (nadroparin). They noted a substantially lower incidence of DVT, as detected by Doppler ultrasonography, in the early group (2%) as compared with the late group (26%). In this same study, a subgroup analysis noted that American Spinal Injury Association (ASIA) grade A SCI patients were more substantially likely to develop DVT as compared with ASIA grade D SCI patients (incidence of 36% vs 7%). On the basis of this compelling data, Christie et al recommend early (< 72 hours postinjury) initiation of VTE prophylaxis in the setting of acute SCI.24

In addition, Christie et al also reviewed the use of chemical VTE prophylaxis in the perioperative period for acute SCI patients. Two prospective trials^21,22 and one retrospective review²³ were identified that addressed this issue. The combined findings of these studies noted a low incidence of major bleeding complications (0–2.6%) and an incidence of PE of 0 to 5.2% and symptomatic DVT of 0 to 1.7% that was not substantially elevated above baseline. Accordingly, based on a balance of risk and with the caveat that any recommendation is supported by weak data, Christie et al ultimately recommended withholding LMWH on the morning of surgery and resuming within 24 hours postsurgery.

Recommendations for Venous Thromboembolism Prophylaxis in Spinal Surgery Patients with Spinal Column or Spinal Cord Injury

The following recommendations for patients with SCI are made by adapting the clinical guidelines for all major trauma patients, including spinal column and cord injuries, of Gould et al,¹⁹ in combination with the recommendations of other systematic reviews of VTE prophylaxis in SCI patients^24,29:

• The use of UFH, LMWH, or IPC mechanical prophylaxis is suggested over no prophylaxis.

• Chemoprophylaxis should be instituted within 72 hours of injury.

• Low molecular weight heparin should be withheld on the morning of surgery and resumed within 24 hours following surgery.

• For major trauma patients at high risk for VTE, including all patients with acute SCI and requiring spinal surgery for trauma (without or without SCI), combination pharmacological prophylaxis and mechanical prophylaxis is recommended when not contraindicated by lower-extremity injury.

• For major trauma patients, including all patients with acute SCI and requiring spinal surgery for trauma (without or without SCI), in whom LMWH and low-dose unfractionated heparin (LDUH) are contraindicated, mechanical prophylaxis, preferably with IPC, is suggested over no prophylaxis when not contraindicated by lower-extremity injury. Addition of pharmacological prophylaxis should commence when the risk of bleeding diminishes or the contraindication resolves.

• In patients with acute SCI requiring spinal surgery for trauma (without or without SCI) inferior vena cava (IVC) filters should not be used for primary VTE prevention.