Incidence and Risk Factors for Venous Thromboembolism in Patients with Spinal Cord Injury: A Retrospective Study

Introduction: The true incidence of venous thromboembolism (VTE) in patients with acute spinal cord injury (SCI) is unclear. The management of thromboprophylaxis varies among clinicians and is challenging due to the balance against bleeding. There is limited data on the risk factors associated with VTE in patients with an acute SCI. Methods: We performed a retrospective chart review of consecutive adult patients with acute traumatic or non-traumatic SCI presenting within 1 week of injury from 2009 to 2015. We excluded patients who were already on therapeutic oral anticoagulation, those who had a short hospital admission (<7 days), and patients who had early transfer to a different hospital location (<30 days from the initial admission). The primary outcome was incidence of symptomatic, objectively verified deep vein thrombosis (DVT) and/or pulmonary embolism (PE) within 90 days. Secondary outcomes were major bleeding, all-cause mortality, and fatal PE. Step-wise cox modeling analyses were used to identify risk factors for VTE. Results: A total of 211 eligible patients were initially screened and 60 patients were excluded: 34 patients had short hospital admissions, 17 were transferred to a different hospital location (median time of 11 days, range 5 to 29 days), 3 had another reason for admission, and 6 patients were already on therapeutic anticoagulation. The remaining 151 patients with acute SCI were analyzed. Median age was 51 (range 17 to 91 years) and 106 (70%) were males. Patients were followed for a median of 90 days (range 1 to 90 days) and the median length of hospital stay was 79 days (range 1 to 90 days). Hundred and eleven patients (73.5%) had paraplegia or tetraplegia, 10 patients (6.6%) had no neurological impairment, and the degree of impairment was unspecified in 30 patients (19.9%). A total of 112 patients (74.2%) had a traumatic SCI with a fracture, 34 (22.5%) had a traumatic SCI without a fracture, and 5 (3.3%) had a non-traumatic SCI. Ninety four patients (62.2%) had a SCI alone versus 57 (37.8%) who had other sites of injury in addition to a SCI. The median duration of thromboprophylaxis was 65 days (range 2 to 90 days). Majority of the patients (59.6%) received low-dose low-molecular-weight heparin (LMWH) either alone (85 of 151, 56.3%) or sequentially with warfarin (5 of 151, 3.6%) with international normalized ratio range of 2 to 3 compared to 48 patients (31.8%) who received increased intensity LMWH either alone (9 of 151, 6%) or sequentially with low-dose LMWH (39 of 151, 25.8%). Thirteen patients (8.6%) had no thromboprophylaxis. Of the 151 patients included in the analysis, 17 patients (11%) had symptomatic VTE (9 PEs, 6 lower extremity DVT, 1 upper extremity DVT, and 1 with both a DVT and PE). The median time of VTE occurrence was 18 days (range 2 to 65 days) (Figure 1). There was no statistical difference in rate of VTE between the standard prophylactic LMWH group compared to those who received increased intensity LMWH (13.3% versus 8.3%, respectively; odds ratio [OR] 0.59, 95% confidence interval [CI], 0.13-2.1, p = 0.58). Six patients (4%) had major bleeding and none had a fatal PE. The median time of major bleeding occurrence was 17 days (range 8 to 40 days). The all-cause mortality rate was 13.9%. The median time of all-cause death was 6 days (range 1 to 61 days). In the univariate analyses, male sex (OR 14.95; 95% CI, 1.05-49.1, p = 0.003) and having other sites of injuries along with SCI (OR 2.6; 95% CI. 0.8- 8.7, p = 0.049) were significantly associated with the risk of VTE. In stepwise Cox modeling, independent contributors to risk for VTE were other sites of injuries (hazard ratio [HR] 6.07), age (HR 1.05 per year) and to some extent the presence of leg paresis/para- or tetraplegia (HR 2.7), whereas hypertension appeared to reduce the risk (HR 0.18). The gender variable could not be included due to zero events for females. Conclusions: Symptomatic VTE is still today a frequent complication in patients with acute SCI and can occur early after injury. Male sex, age and presence of other sites of injuries along with SCI were independent risk factors for symptomatic VTE. Future randomized trials assessing the role of direct oral anticoagulants versus the current standard of care using low-dose LMWH among these higher risk patients are needed. Figure Survival without symptomatic VTE after SCI Figure. Survival without symptomatic VTE after SCI Disclosures No relevant conflicts of interest to declare.