VTE has become the third most prevalent vascular disease following acute coronary syndrome and stroke; it is also a common complication following surgery. However, the risk assessment model for VTE in patients with spinal fractures is scarce. Our analysis revealed that six factors, including age, spinal cord injury, Postoperative bed time, D-DI, FIB and FDP, were correlated to the occurrence of VTE, respectively. Importantly, the AUG of the joint model for VTE risk assessment was 0.88 with higher sensitivity and specificity compared to the D-D, FIB and FDP, respectively.

It was reported that various risk factors contributed to the occurrence of VTE in patients with spinal fractures, which could be divided into non-genetic and genetic factors. In elderly patients, hardening of blood vessels, a decrease in elasticity of blood vessel wall and venous valve function led to slow of blood flow. In addition, hypertension, diabetes and other basic diseases, accompanied by functional or organic changes of multiple organs, are likely to lead to hypercoagulability of the body's blood, consequently increasing the risk of VTE [15, 16]. Previous studies showed that the incidence of lower limb DVT increases by 0.5–0.6 percent when age increases from 45 to 89 years old [17]. Masuda et al. [18] found that thrombosis incidence with over 70 years old significantly increased compared with patients with lower than 50 years old in patients after spinal surgery. Our findings also unveiled that the VTE incidence was higher in patients over 60 years old, which was in line with the previous findings. Although some studies suggested that the occurrence of VTE is not related to age [19], the Chinese thrombosis prevention recommendations state unequivocally that the older one is the greater the risk of lower limb DVT.

Patients with spinal fractures, often complicated with spinal cord injury, are more prone to VTE, compared to patients without spinal cord injury. In previous investigations, spinal cord damage is a significant risk factor for VTE among patients obtaining spinal fracture surgery. Moreover, spinal cord damage is frequently associated with traumatic events, neurosurgical operations, and protracted immobility, all of which are high-risk factors for VTE [20]. Our study revealed that the percentage of VTE was 24% in patients with spinal fractures complicated by spinal cord injury, which was consistent with previous research [21]. Besides, it was reported that longer operative duration was associated with an increased risk of VTE [22]. In our study, the operation duration was significantly correlated to the incidence of VTE, however, the OR of operation duration for VTE occurrence in the multivariate logistic analysis showed that the patients with longer operation duration over 3 h did not have higher risk for VTE. Additionally, slow blood flow caused by postoperative bed time increased the incidence of lower limb DVT. Our study showed that patients with postoperative bed time of more than 5 days had a higher incidence of VTE.

The non-genetic factors, including age, spinal cord injury, operative duration and postoperative bed time, were proved to be associated with VTE occurrence. Meanwhile, APTT, PT and TT also attributed to non-genetic factors for VTE occurrence. Interestingly, other studies presented that APTT and PT had a significant difference between NVTE and VTE groups, nevertheless, TT did not have a significant difference between NVTE and VTE groups [23]. Our findings revealed that all of these three parameters, including APTT, PT and TT, did not have differences among NVTE and VTE groups, which was not consistent with other studies. The possible reason might be due to the differences in the number of patients included in the analysis. Another possible reason was that the exclusion criteria were different from other studies. Therefore, whether these three parameters were closely correlated to the VTE occurrence needs more patients for analysis.

Except for the non-genetic factors, the genetic factors associated with the coagulation were also involved in the VTE occurrence. When fibrin is degraded by plasmin, D-dimer is produced, which is a product of early fiber degradation in thrombosis. The elevation of DD has been widely confirmed as a risk factor for VTE occurrence [24, 25]. Our finding also unveiled that DD was significantly elevated in the VTE group compared to the NVTE group, which indicated that DD could be an independent risk factor for VTE incidence. Meanwhile, it was also demonstrated that high FIB was associated with susceptibility to thrombosis in patients with posttraumatic deep vein thrombosis [26]. FIB level was significantly increased in the patient with VTE compared to the patient without VTE in our study. Additionally, FDP is also a product of fibrinogen degradation. Wu et al. [27] analyzed 569 patients with femoral and pelvic fractures, indicating that a high level of FDP is a high-risk factor for DVT in the perioperative period. It was also proved that FDP was dramatically increased in patients with VTE in our study. Taken together, DD, FIB and FDP could be the risk factor for VTE incidence in patients with spinal fractures separately. Importantly, the combination of these three factors was more effective to predict the incidence of VTE in patients with spinal fractures.

Although our study revealed the non-genetic factors and genetic factors were closely correlated to VTE incidence through chi-square test and OR analysis, the number of included patients was limited so that the bias could not be excluded from the analysis. More patients would be included in the risk assessment of VTE incidence.