The purpose of this study was to retrospectively analyze the risk factors associated with the development of DVT in patients with degenerative spinal lesions undergoing vertebral fusion. In different reports in the literature, many of the conclusions of the risk factor analysis are different, which may be the result of applying different statistical methods and screening tests. Therefore, there are also overlaps and conflicts between our findings and the available reports
In our univariate analysis, BMI ≥24 [OR = 2.800, 95% CI (1.196-6.555), P = 0.018] was statistically significant. In a study by Tan et al[10], the causal relationship between BMI level and increased risk of DVT was investigated by analyzing published studies, and they found a significant causal association between high BMI and DVT. This may involve several biological mechanisms: high BMI may affect venous return and increase chronic venous insufficiency, thus allowing venous blood stasis to occur[11]. In addition, high BMI may increase fibrinogen activator inhibitor-1 (PAI-1) levels, thereby inhibiting clot lysis and ultimately impairing fibrinolytic function[12]. Therefore, reducing the incidence of obesity through interventions may be effective in preventing the development of DVT. Trauma [OR = 7.887, 95% CI (3.579-17.381), P = 0.001] was also a risk factor in the results of the univariate analysis. Yumoto et al[13] found that the risk of DVT in Japanese trauma patients was about 30%. This may be related to the patient's prolonged rest with blood in a hypercoagulable state. This also suggests that antithrombotic therapy should be given to trauma patients as early as possible after surgery. Multifactorial logistic regression analysis revealed that spinal cord injury, diabetes mellitus, preoperative muscle strength (≥ grade 4) and preoperative D-dimer abnormalities were associated as risk factors for the development of DVT after surgery. The cause of spinal cord injury [OR = 3.165, 95% CI (1.187-8.439), P = 0.021] was due to the Virchow triad of blood stasis, vascular endothelial cell injury, and a hypercoagulable state of blood.[14] This is in agreement with the retrospective study by Agarwal et al[15]. They also found no correlation between the occurrence of DVT and injury level, ASIA classification and limb spasticity status. Interestingly, in a study by Lv et al[16], they found lower extremity muscle strength < grade 3 as one of the risk factors for DVT, and they suggested that both prolonged bed rest and motor impairment lead to reduced lower extremity activity, leading to circulatory stagnation in association. This conflicts with our results of grade 4 in preoperative muscle strength [OR = 0.072, 95% CI (0.009-0.601), P = 0.015], and grade 5 in preoperative muscle strength [OR = 0.036, 95% CI (0.004-0.331), P = 0.003]. Perhaps this is related to our application of low molecular calcium, which we will investigate in depth in a prospective trial at a later stage. Diabetes [OR = 5.191, 95% CI (1.723-15.640), P = 0.003] proved to be significant for the development of postoperative DVT, which is consistent with the study of Deng et al[17]. This may be related to abnormalities in some of the coagulation and fibrinolytic indicators that contribute to the propensity for thrombosis[18]. In a study by Albert[19] on the relationship between risk factors for coronary heart disease and DVT reported that both diabetes and obesity significantly increased the risk of developing DVT independently of age, race and gender. D-dimer is a degradation product of cross-linked fibrin and is therefore a biomarker of coagulation activation and fibrinolysis. Abnormal levels of D-dimer indicate that the patient's blood is in a hypercoagulable state, which plays an important role in thrombus formation. In our study results, preoperative D-dimer abnormalities [OR = 2.770, 95% CI (1.059-7.245), P = 0.038] also confirmed this view. In the perioperative period, we need to closely monitor patients' D-dimer and other related coagulation indicators, and timely antithrombotic treatment for abnormal patients can effectively reduce the incidence of DVT.
Doppler ultrasound technology is commonly used in the diagnosis of DVT. Clinicians can use this technique to identify DVT early and take appropriate treatment, which helps a lot in detecting asymptomatic and lethal DVT. Therefore, it is necessary to use Doppler ultrasound technology preoperatively and regularly postoperatively. This technique is simple, inexpensive, noninvasive and safe, but does not accurately assess the proximal extent of pelvic vein thrombosis in patients with DVT proximal to the inguinal ligament[20]. MRI and venography are more accurate, but have significant limitations of their own. MRI is more expensive and is not indicated for most patients with cardiac stent placement or other metal implants. Venography is invasive and its potential for nephrotoxicity may affect its practical use in the clinic[21]. Among the various similar studies that have been reported, there is currently a wide variety of diagnostic methods, and no definitive best diagnostic method has been developed yet.
In the studies of Yoshioka et al[22] and Yang et al[23], they found that advanced age and hypertension were also risk factors, which is contrary to the results of our univariate analysis. This may be caused by the limited number of samples we selected and the fact that patients did not stop using antihypertensive drugs after surgery. Chronic hypertension causes vascular endothelial damage, and recent studies have shown that vascular endothelial damage is also a risk factor for DVT.[24, 25] We speculate that hypertension may also be a contributing factor to DVT, and we can subsequently verify this prediction by counting the duration of hypertension in patients and animal experiments. In a study by Sebastian et al[22], they concluded that operative time >193 minutes and hospital stay ≥6 days were also independent risk factors. Thus, it seems that the importance of improved surgical efficiency in minimizing postoperative complications has been demonstrated. Our results are contrary to the above findings and may be influenced by the complexity of the case leading to the presence of other confounding factors. Taken together, this because requires that we should monitor more carefully the occurrence of DVT postoperatively in cases with longer operative times. Notably, their study also found that patients with ≥6 days of hospitalization were more than 4 times more likely to have a DVT event than those with ≥6 days of hospitalization. This may be related to the prolonged lack of exercise of the lower extremities due to prolonged postoperative bed rest in patients. Also, the mean time to diagnosis of DVT was 10 days postoperatively, and a large proportion of patients may have been diagnosed with DVT after discharge but were not included in the statistics. Referring to their findings, this also gives us hints about the importance of educating patients about the signs and symptoms of postoperative DVT and emphasizing monitoring in the outpatient setting.
Our study also had some limitations. Retrospective analysis resulted in our lack of randomization (i.e., randomized grouping), and failure to conduct prospective studies may have biased our results. Second, in order to collect complete data, chart review relied on manual extraction and accurate reporting and coding by the treatment team. Some patients may have received other treatments outside of our system that could have an impact on the DVT diagnosis. Third, because DVT testing is only used for diagnosis in the presence of symptoms and never for routine asymptomatic surveillance, the occurrence of DVT may be underreported. In addition, the study results may be biased due to the limited number of cases we collected.