Factors associated with venous thromboembolic events following ICU admission in patients undergoing spinal surgery: An analysis of 1269 consecutive patients

OBJECTIVE Venous thromboembolic events (VTEs) are a common cause of morbidity and mortality after spine surgery. Patients admitted to the intensive care unit (ICU) following spine surgery exhibit high-risk clinical characteristics. METHODS The authors retrospectively analyzed 1269 ICU patients who had undergone spine surgery between January 1, 2009, and May 31, 2015. Relevant demographic, procedural, and outcome variables were collected. RESULTS Patients admitted to the ICU postoperatively had a postoperative VTE rate of 10.2%, compared to 2.5% among all spine surgery patients during the study period. ICU patients had a higher comorbid disease burden (odds ratio [OR] 1.45, p < 0.001), and were more likely to have a history of a bleeding disorder (2.60% vs 0.46%, OR 2.85, p = 0.028), receive a transfusion (OR 4.81, p < 0.001), have a fracture repaired (OR 4.30, p < 0.001), have an estimated blood loss > 500 ml (OR 1.95, p = 0.009), have an osteotomy (OR 20.47, p = 0.006), or have a corpectomy (OR 3.48, p = 0.007) than patients not admitted to the ICU. There was a significant difference in time to VTE between patients undergoing osteotomy and patients undergoing scoliosis corrections without osteotomy (p = 0.0431), patients with fractures (p = 0.0113), and patients undergoing fusions for indications other than scoliosis or fracture (p = 0.0056). Patients who developed a deep vein thrombosis (DVT) during their ICU stay were more likely to have received a prophylactic inferior vena cava filter placement (OR 8.98, p < 0.001), have undergone an interbody fusion procedure (OR 2.38, p = 0.037), have a history of DVT (OR 3.25, p < 0.001), and have shorter surgery times (OR 0.30, p = 0.002). Patients who developed a pulmonary embolism (PE) during the ICU stay were more likely to have a history of PE (OR 12.68 p = 0.015), history of DVT (OR 5.11, p = 0.042), fracture diagnosis (OR 7.02, p = 0.040), and diagnosis of scoliosis (OR 7.78, p = 0.024). Patients with higher BMIs (OR 0.85, p = 0.036) and those who received anticoagulation treatment (OR 0.16, p = 0.031) were less likely to develop a PE during their ICU stay. CONCLUSIONS Patients admitted to the ICU following spine surgery have a higher rate of VTE than non-ICU patients. Time to VTE varied by pathology. Factors independently associated with VTE in the ICU are distinct from factors otherwise associated with VTE. Some factors are independently associated with VTE throughout the 30-day postoperative period, while others are associated with VTE specifically during the initial ICU stay or after leaving the ICU.