Anticoagulation in Elective Spine Cases

Rates of Hematomas Versus Thromboembolic Disease

Dharani Rohit Thota, BA; Carlos A. Bagley, MD; Mazin Al Tamimi, MD; Paul A. Nakonezny, PhD; Michael Van Hal, MD


Spine. 2021;46(13):901-906. 

In This Article


There is much debate on the importance of postoperative anticoagulation. While PE's can be deadly, postoperative hematomas can also be debilitating and devastating complications. It does raise an interesting question on when and how should patients be protected from both postoperative complications. It is a fine line to walk between increasing a bleeding risk and increasing risk for potentially deadly VTE events.

In this study, pharmacologic anticoagulation did not significantly decrease the symptomatic VTE incidence rate, but it did significantly increase hematoma rates and the return to OR rates. This finding is in line with recent literature findings in regarding to spinal surgery and pharmacologic anticoagulation administration.[10,11] However, in the review article by Kepler et al[1] there was a significant decrease in VTE incidence following pharmacologic anticoagulation in spine surgery.

Some of the literature studies have different methods of VTE detection. Some of them use active screening which increases the number of non-symptomatic VTE events reported.

While our study does have noted limitations, it does raise the question if anticoagulation should be used in routine cases as it does not appear to lead to lower VTE rates, but it does appear to have an adverse effect on the rate of return to OR for hematomas.

Notably, the timing of the VTE may be an area of further investigation. While most of the hematoma risk in this study appeared to be in the relatively early postoperative period, there appears to be a trend toward symptomatic VTE occurring later in the postoperative course.

Our study has a few key limitations, namely the surgeons were different. While still at the same institution, each surgeon practices independently. There may be inherent differences such as the use of tranexamic acid, the approach to blood loss. Durotomies are another large potential confounder. These are hard to control for in any study due to the way they are reported. The matching process would help control for this by controlling for complexity and case duration, but it is a weakness of its retrospective nature.

Future studies will need to have a more rigorous methodology for determining the need for anticoagulation. Until, a randomized control trial is available to answer this question, this study provides some insight into the potential risks of anticoagulation while not clearly reducing the major target of this therapeutic. While the current study is not a randomized controlled trial, we attempted to overcome this by using propensity matching for the medical risk factors of the patient as well as the case severity. The propensity score plays an important role in balancing the study groups to make them comparable. Rosenbaum and Rubin[8] showed that treated and untreated subjects with the same propensity scores have identical distributions for the included covariates. This "balancing property" means that, if we control for the propensity score when we compare the groups, we have effectively turned the observational study into a randomized block experiment, where "blocks" are groups of subjects with the same propensities. This bolsters the internal validity of the study. Propensity scores also offer a promising way to assess the similarity between a sample and a target population of interest, which can bolster external validity.