Introduction: A prospective study was designed to elucidate incidence and predictors of deep venous thrombosis (DVT) in patients undergoing craniotomies.
Materials and methods: Ninety-two patients who underwent craniotomies received pre- and postoperative venous ultrasonography and/or contrast-enhanced spiral computed tomography for diagnosis of DVT. The primary endpoint was DVT occurrence. Serial levels of serum D-dimer, soluble fibrin, and thrombin-antithrombin complex (TAT) were analyzed.
Results: Twenty-four of 92 patients (26.1%) had DVT, of whom 10 (41.7%) were diagnosed preoperatively. In patients with preoperative DVT, age, incidence of decreased performance status and leg paresis, levels of D-dimer, soluble fibrin, and TAT were significantly greater. In patients with postoperative DVT, length of surgery, incidence of decreased postoperative performance status, levels of D-dimer on postoperative days (POD) 3, 7, and 14, and TAT on POD7 were significantly greater. Patients with postoperative DVT had elevated D-dimer levels on POD 7 compared with POD 3. The D-dimer cutoff of 2.65 μg/mL at POD 7 could be used to identify DVT with 85.7% sensitivity and 72.3% specificity. A cutoff of 5.25 μg/mL at POD 7 yielded a specificity of 96.9%. Decreased performance status and elevated D-dimer were independent predictors for preoperative DVT, prolonged operation time, and elevated D-dimer on POD 7 for postoperative DVT.
Conclusions: DVT frequently was observed in patients before and after undergoing craniotomies. Patients with decreased performance status should be preoperatively screened for DVT by checking D-dimer levels. Elevated D-dimer levels on POD 7 compared with POD 3 and D-dimer levels greater than 2.65 μg/mL at POD7 suggest the presence of DVT.
Keywords: Deep vein thrombosis; Neurosurgery; Perioperative period; Venous thromboembolism.
Copyright © 2017. Published by Elsevier Inc.