Background context

There is significant variability in postoperative chemoprophylaxis protocols amongst spine providers due to perceived risks and benefits, but limited data on the topic. At our institution, both orthopaedic spine and neurosurgery departments utilize unfractionated subcutaneous heparin in identical dosages and frequency, with the only difference being time to initiation postoperatively.

Purpose

To evaluate the rate of symptomatic venous thromboembolism (VTEs) and unplanned reoperation for hematoma based on timing of chemoprophylaxis initiation.

Study design/setting

Single institution retrospective cohort study.

Patient sample

Patients undergoing elective spine surgery, excluding patients undergoing surgery in the setting of trauma, malignancy, or infection.

Outcome measures

Outcome measures included the diagnosis of a venous thromboembolism within 90 days of surgery and unplanned reoperation for a hematoma.

Methods

Patients undergoing elective spine surgery from 2017 to 2021 were grouped based on chemoprophylaxis protocol. In the "immediate" group, patients received subcutaneous heparin 5000 units every 8 hours starting immediately after surgery, and in the "delayed" group, patients received chemoprophylaxis starting postoperative day (POD)-2 for any decompressions and/or fusions involving a spinal cord level (ie, L2 and above) and POD-1 for those involving only levels below the spinal cord (ie, L3 to pelvis). A cox proportional hazards model was created to assess independent predictors of venous thromboembolic events, while a logistic regression was utilized for unplanned reoperations for hematoma.

Results

Of 8,704 patients, a total of 98 (1.13%) VTE events occurred, of which 43 (0.49%) were pulmonary embolism. Fifty-four patients (0.62%) had unplanned reoperations for postoperative hematomas. On cox proportional hazards model analysis, immediate chemoprophylaxis was not protective of a venous thromboembolism (Hazard Ratio: 1.18, p=.436), but, it was a significant independent predictor for unplanned reoperation for hematoma on multivariable logistic regression modeling (Odds Ratio: 3.29, p<.001).

Conclusions

Both chemoprophylaxis protocols in our study resulted in low rates of VTE and postoperative hematoma. However, our findings suggest that the delayed chemoprophylaxis protocol may mitigate postoperative hematoma formation without increasing the risk for a thrombotic event.