Recurrently Back to the Drawing Board on Thromboprophylaxis After Spine Surgery

Consensus has been chronically elusive with regard to the timing of chemical thromboprophylaxis (TPx) after spine surgery. In general, “too early” might risk bleeding and a return to the operating room, while “too late” or “not at all” could possibly risk an increased rate of venous thromboembolism (VTE) complications, such as deep vein thrombosis and pulmonary embolism, particularly in patients with reduced mobility, spinal cord injuries, trauma, and other risk factors. Postoperative TPx risk-benefit profiles in elective spine surgery can differ from those in the surgical management of traumatic spinal cord injury. Hence there is a very heterogeneous body of literature, and often “mixed messages” in this field. This might suggest that future clarity will require ever more nuanced subgroup analyses, which could eventually guide individualized care more optimally—while ironically requiring larger and larger studies in order to achieve that goal. A 2024 analysis1 of 8,704 patients who underwent elective spine surgery, one of many large insightful studies, retrospectively compared cohorts who had received heparin under usual care immediately postoperatively versus at 1 to 2 days postoperatively and found similar rates of VTE, but fewer hematomas, in the groups not receiving immediate postoperative heparin. As with many other studies, those timelines are typical for postoperative TPx. The present large database study examined a retrospective observational cohort of 5,325 patients treated across 463 trauma centers between 2013 and 2020, focusing on patients who underwent surgery following a complete traumatic cervical spinal cord injury (American Spinal Injury Association [ASIA] type A). The data revealed a significant practice trend of earlier administration of postoperative chemical TPx (5.2 hours earlier per year, on average, coming down under usual care from 114 hours to 82 hours postoperatively). This coincided with an annualized 6.2% reduction in the odds of postoperative VTE (the estimated incidence of 13% came down to 9%). Trends like these cannot proceed indefinitely because a point will be reached where the improvements in VTE rates would taper off and the complications (bleeding, etc.) would likely rise, perhaps when postoperative TPx administration time becomes less than 1 to 2 days, one might guess. This study’s data were enlightening in that usual care was to administer TPx within a postoperative time range when there was still room for improvement in VTE rates (as has been empirically shown here). Of course, the postoperative bleeding question matters, too, but it cannot be answered because of the lack of granular data on such in this dataset and the analysis only of patients with paralysis (who would not necessarily exhibit neurologic decline if experiencing a postoperative epidural hematoma, for example). Future studies will certainly need to look at other neurologic subgroups, and perform detailed analysis of postoperative bleeding. In this era of heightening interest in pseudo-statistical modeling approaches (machine learning, artificial intelligence, etc.), the interpretation and wise analysis of large retrospective databases will be increasingly important in medical research. The present study stays on track to answer specific questions within the scope of its dataset, while leaving open many related topics, which will be crucial to study in future endeavors. Subgroups definitely appear to be an important factor in this field, since the trends shown here are one indication that a “one-size-fits-all” guideline approach will not be the main way forward. One thing appears certain, though: the subject of postoperative TPx is surprisingly far from settled in the realm of spine surgery, particularly for traumatic spinal cord injury. This paper provided a lot of food for thought and is appreciated.