This study reviewed the appropriateness of VTE chemoprophylaxis following spine surgery. Conclusions were drawn from a methodology designed to address clinical scenarios regarding VTE prevention, based on existing evidence and clinical expertise from a multidisciplinary panel. Scenarios that were appropriate with agreement are likely reasonable to consider. Those that are rarely appropriate are likely unreasonable to consider. Uncertainty or disagreement in ratings likely arise when the strength of the current literature is weaker, and clinical experience with the intervention has been variable. These are often scenarios in which success has been realized but results are more inconsistent and/or controversial, or the potential for harm is higher. In these cases, more scientific study is needed to clarify the direction of recommendation. This is not a declaration that these interventions are appropriate or inappropriate but suggests that they should be considered with greater scrutiny on a case-by-case basis.

The definition of ‘appropriate’ should be reiterated: chemoprophylaxis is reasonable to consider given the scenario. There are scenarios where the risk of VTE is extremely low, such that a rater would not use prophylaxis in their own practice, yet still considers it reasonable. In the current study, the median ratings never found VTE prophylaxis to be rarely appropriate for any scenario, although uncertainty surrounded some scenarios, usually when the risk of VTE was lower and/or the risk of bleeding complications was perceived to be higher. It is also noteworthy that high bleeding risk was not specifically used as a modifier; thus, concerns about bleeding are only indirectly reflected in uncertain ratings.

There are previous systematic reviews and studies that have not recommended routine chemoprophylaxis in low risk patients due to the infrequency of VTE.63, 64 Shapiro et al suggested using an algorithm to selectively utilize chemoprophylaxis in higher risk patients only.44 Using this algorithm, they started the medication 24 to 36 hours after surgery and found a very low incidence of both VTE and wound complications.

Spinal cord injury was the strongest predictor of appropriateness for recommending VTE prophylaxis within 24 hours of surgery. This is consistent with several prior studies.21,35, 54, 65-68 Additionally, scenarios listed as ‘high-risk’ were significantly more likely to receive a final rating as ‘appropriate’ for initiating VTE prophylaxis within 24 hours, also consistent with several prior studies.22, 32, 52, 54-56, 69-83 Various risk factors have been identified in the literature, including prior VTE, cancer, baseline coagulopathy, prolonged hospital stay, delayed postoperative mobilization, cervical myelopathy, obesity, and advanced age. Additionally, there is data that suggest that pharmacologic prophylaxis more selectively decreases rate of DVT than pulmonary embolus (PE)84, while other studies suggest improvement in PE complications as well with early prophylaxis.45 Lastly, there are studies that have concluded that anterior thoracolumbar surgery is a risk factor for VTE39 although anterior surgery was not a substantial contributor to decision making in this AUC (P=0.2), possibly due to the lack of differentiation between anterior cervical and thoracolumbar surgery.

In the current study, the timing for initiating chemical prophylaxis was most controversial. There was nearly unanimous support for appropriateness of anticoagulation on a delayed basis, defined as greater than 24 hours after surgery. However, recommendation for initiating anticoagulation immediately, defined in this study as less than 24 hours following surgery, was considerably more uncertain and generated more disagreement (Table 3, Figure 1). Early intervention varies in the literature but is often considered to be within 48 or 72 hours, which was slightly different than our definition.85 It is known that VTE tends to occur early following the insult (injury or surgery) such that rates of VTE are directly related to timing of prophylaxis.10, 86, 87 Thus, there appears to be an advantage to immediate prophylaxis if it can be administered with a low complication profile.

A review of 112,000 patients from the NSQIP database plus a home institution review found no difference in VTE rate using chemoprophylaxis after spine surgery, but increased bleeding complications in elective patients.36 In another study, use of anticoagulation decreased incidence of VTE but also increased in the risk of bleeding complications (Zeng X-J). Both of these studies highlight concerns for bleeding complications when using chemoprophylaxis after spine surgery.

Conversely, most studies suggest that anticoagulation did not increase the rate of bleeding complications, although timing was variable. In the CLOTT study, there was a subgroup of patients with SCI that benefited from prophylaxis65, with 21% receiving VTE prophylaxis within 24 hours and no patients were reported to have clinically relevant bleeding complications in the SCI cohort. A small study compared two anticoagulants in 179 patients with the first dose given preoperatively.51 These were small elective procedures in low-risk patients with average duration of approximately one hour, and no bleeding complications were observed. Similarly, in a review of patients undergoing elective spine surgery over a 5-year period, there were no bleeding complications after starting anticoagulation 6 hours after surgery.17 A small pilot study administered chemoprophylaxis less than 24 hours prior to surgery in patient with SCI and found no bleeding complications in 47 patients.89 Two anticoagulants were compared in 556 total patients undergoing either elective or trauma surgery.24 Both received their first dose 6 hours following surgery, and there were 2 total bleeding complications, one in each group. In a review of their trauma database, Sharpe et al compared VTE and hemorrhagic complications in patients that did or did not have at least 50% of their chemoprophylaxis before surgery.45 Average time from injury to surgery was 3.8 days. There was variability in chemoprophylaxis from time of last dose to surgery as well as starting time after surgery. They found a small bleeding risk after surgery, but it was not different between the two groups. Another study looked at bleeding complications following consecutive spine surgeries over a 3 year period.20 This was a mix of primarily degenerative cases but also included tumor and infection. Patients receiving chemoprophylaxis after 24 hours of surgery had similar rates of hematoma to those that had their first dose prior to 24 hours. In summary, while many studies suggest that early administration of VTE chemical prophylaxis is safe, there are conflicting conclusions across studies concerning the risk of hemorrhage with early administration.

Several studies have investigated the use of anticoagulation on a more delayed basis. In 367 patients following elective cervical or lumbar laminectomy who received anticoagulation 24 to 36 hours following the procedure, no bleeding complications were noted.90 Chang et al retrospectively compared their trauma patients receiving prophylaxis before vs after 48 hours following surgery.5 They found reduction in VTE particularly for patients with SCI without difference in hematoma at 30 days. There are several other studies that found a small bleeding risk after surgery, not necessarily greater than controls, although timing was not well defined.6,7, 16, 91 Other studies found no significant bleeding complications starting medication after 48 hours following surgery.19, 29 Kim et al also found no bleeding complications in a 5-year review of their trauma registry.31 There was substantial variability in time to initiate prophylaxis, although some patients received their first dose as early as 12 hours postoperatively. Two systematic reviews found a low rate of hematoma and VTE with use of chemoprophylaxis following spine surgery92, 93, while another review of patients receiving low molecular weight heparin (LMWH) after elective spine surgery for degenerative conditions did not find any postoperative hematoma complications.63

The results of this study should be considered with its limitations. This study did not consider type of pharmacologic prophylaxis. Across the literature reviewed, the type of chemical prophylaxis varied and included platelet inhibitors, direct oral anticoagulants, warfarin, subcutaneous heparin, and low molecular weight heparin. The purpose of this study was to define appropriateness for any chemical VTE prophylaxis and for that reason, no attempt was made to differentiate efficacy of one form of prophylaxis over another. As previously noted, we also did not differentiate between anterior cervical and anterior lumbar spine surgery, which could bias the final ratings. Lastly, we assume that the spine specialist can identify patients that have high risk factors, and we chose to limit ‘risk factors’ to either ‘high’ or ‘low’ risk. This creates the risk for misclassification bias, as risk in true clinical scenarios lies on a spectrum.