D-dimer screening for deep venous thrombosis in traumatic cervical spinal injuries.

BACKGROUND CONTEXT: Deep venous thrombosis (DVT) in spinal cord injury (SCI) patients is a life-threatening comorbidity. Despite its seriousness, prophylaxis and screening for DVT in SCI patients are still not sophisticated.

PURPOSE: This study aimed to define the epidemiology and incidence of DVT in acute traumatic cervical SCI in a Japanese population, determine the best timing for DVT screening, and determine the optimal D-dimer threshold level for use as an easy and minimally invasive screening tool.

STUDY DESIGN: This is a prospective clinical study.

PATIENT SAMPLE: The patient sample included acute traumatic cervical SCI patients who were admitted to our facility within 2 weeks after injury.

OUTCOME MEASURE: Multivariate logistic regression was performed for outcome measure.

METHODS: We enrolled 268 patients (223 men and 45 women), from April 2007 to December 2012. After excluding early drop-out patients, 211 patients remained. Assessment for neurological status and blood chemistry, especially blood coagulation levels (prothrombin time, prothrombin time-international normalized ratio, activated partial thromboplastin time, and serum D-dimer), was performed every week until 1 month after injury. Ultrasonography was performed for DVT detection every 2 weeks.

RESULTS: Deep venous thromboses were detected in 22 patients (10.4% of patients studied). All DVT-positive patients demonstrated severe paralysis classified as C or greater on the American Spinal Injury Association (ASIA) Impairment Scale. Multivariate logistic regression of clinical and laboratory parameters revealed that only the D-dimer level at 2 weeks after injury was an accurate predictor of DVT formation. The optimal threshold of D-dimer for prediction was determined to be 16 μg/dL. The sensitivity and specificity for detecting DVT were 77.3% and 69.2%, respectively.

CONCLUSIONS: D-dimer levels may be used to predict the likelihood of DVT development in patients with acute cervical SCI. Furthermore, the optimal timing for screening test by D-dimer is 2 weeks after injury, and optimal threshold level for D-dimer for diagnosing DVT is 16 μg/dL. Such a screening test would be cost-efficient and simple to administer and could then be followed with additional investigations, such as ultrasonography or venography.