Utility of postoperative D-dimer thresholds and Caprini score in predicting deep venous thrombosis following cranial and spinal surgery

Background: Deep venous thrombosis (DVT) is a major cause of morbidity and mortality in neurosurgical patients. Doppler ultrasound is the diagnostic gold standard but is often limited in resource-constrained settings. The conventional D-dimer cutoff of 500 ng/mL, derived largely from Western populations, may be inappropriate for Sub-Saharan African patients with higher baseline levels, particularly postoperatively. This study aimed to determine postoperative D-dimer thresholds for predicting DVT in cranial and spinal surgery patients and to assess the utility of combining D-dimer with the Caprini score. Methods: This prospective cohort study enrolled patients aged ≥40 years undergoing elective cranial or spinal surgery over 12 months. Plasma D-dimer levels and Caprini scores were measured preoperatively and on postoperative days 1, 3, and 7. Duplex Doppler ultrasonography served as the reference standard. Receiver operating characteristic (ROC) analysis and Youden index were used to derive optimal thresholds. Results: Of 67 patients, 12 (17.9%) developed DVT. In cranial surgeries, optimal D-dimer thresholds were 5020 ng/mL (day 1, area under the ROC curve [AUROC] 0.750), 7450 ng/mL (day 3, AUROC 0.958), and 5770 ng/mL (day 7, AUROC 0.900). In spinal surgeries, thresholds were 7540 ng/mL (day 1, AUROC 0.960), 5830 ng/mL (day 3, AUROC 0.865), and 4520 ng/mL (day 7, AUROC 0.798). Combining Caprini scores with D-dimer did not improve diagnostic accuracy. Conclusion: Distinct postoperative time-based plasma D-dimer trajectories were observed between cranial and spinal neurosurgical patients who developed lower-limb DVT, with a delayed peak discriminatory value on postoperative day 3 after cranial surgery and an early peak on postoperative day 1 after spinal surgery. Within this exploratory cohort, the corresponding optimal D-dimer thresholds were 7450 ng/mL and 7540 ng/mL, respectively. The addition of Caprini scoring did not demonstrate meaningful incremental discriminatory benefit over D-dimer alone.