This study included a total of 169 patients who underwent lumbar fusion surgery, consisting of 80 male patients and 89 female patients. Among them, 144 patients did not experience lower limb DVT postoperatively, while 25 patients developed lower limb DVT, resulting in an incidence rate of 14.79%. Notably, all cases of DVT were found to be in the calf muscle veins. To conduct external validation, an additional 100 patients were collected for use as an external validation dataset.
General characteristics
The comparison of general characteristics between the DVT and non-DVT groups is presented in Table 1. There were significant differences observed in terms of gender. Age of the DVT group was 61(IQR 68) years old, the non-DVT group 58(IQR 63.2) years old. DVT group was significantly older than non-DVT group(P < 0.1). Preoperative hospitalization time stay in DVT group was 2(3) days and in non-DVT group 3(3) days. The difference between them was statistically significant(P < 0.1). In terms of physical signs, there were 39 cases of lower limb muscle weakness in the non-DVT group and 11 cases in the DVT group, which was significantly different between the two groups (P < 0.1).
Table 1
Comparison between general data
| Non-DVT group (n = 144) | DVT group (n = 25) | P-value |
---|
Gender (n, %) | | | 0.036 |
Female | 71 (49.3) | 18 (72) | |
Male | 73 (50.7) | 7 (28) | |
Age (yrs) (median, IQR) | 58 (53.8,63.2) | 61 (58,68) | 0.012 |
BMI(kg/m2) (average ± SD) | 25.5 ± 2.8 | 25.9 ± 3.1 | 0.533 |
Days before surgery (median, IQR) | 3 (2, 3) | 2 (2, 3) | 0.06 |
Duration of disease (months) (median, IQR). | 2 (1, 4) | 1 (1, 4) | 0.724 |
Defecation disorders (yes) (n, %) | 17 (11.8) | 2 (8) | 0.742 |
Urination disorders (yes) (n, %) | 7 (4.9) | 1 (4) | 1.000 |
Sleep disorders (yes) (n, %) | 23 (16) | 2 (8) | 0.378 |
Weakened lower extremity muscle strength (yes) (n, %) | 39 (27.1) | 11 (44) | 0.087 |
Walking impairment (yes) (n, %) | 75 (52.1) | 16 (64) | 0.270 |
Smoking history (yes) (n, %) | 25 (17.4) | 2 (8) | 0.375 |
History of alcohol consumption (yes) (n, %) | 20 (13.9) | 2 (8) | 0.536 |
History of allergies (yes) (n, %) | 10 (6.9) | 0 (0) | 0.361 |
History of trauma (yes) (n, %) | 13 (9) | 4 (16) | 0.285 |
History of surgery (yes) (n, %) | 51 (35.4) | 10 (40) | 0.66 |
Hypertension (yes) (n, %) | 51 (35.4) | 7 (28) | 0.471 |
Diabetes mellitus (yes) (n, %) | 7 (4.9) | 3 (12) | 0.169 |
Coronary heart disease (yes) (n, %) | 8 (5.6) | 1 (4) | 1.000 |
History of cerebral infarction (yes) (n, %) | 7 (4.9) | 1 (4) | 1.000 |
Cerebral ischemia (yes) (n, %) | 1 (0.7) | 1 (4) | 0.275 |
Note: SD: standard deviation; median: median; IQR: quartile; yrs: years old |
Ultrasound and Laboratory test information
A comparative analysis of the ultrasound results and laboratory indicators for the two groups is presented in Table 2. Prior to the surgery, it was observed that the proportion of patients with lower limb venous reflux/varicose veins in the non-DVT group was significantly lower than that in the DVT group (P < 0.1).
Table 2
Comparison between groups of laboratory test results
| Non-DVT group (n = 144) | DVT group (n = 25) | P-value |
---|
Varicose veins/venous reflux of lower extremities (n, %) | 38 (26.4) | 11 (44) | 0.073 |
White blood cells (10^9/L) (median, IQR) | 5.8 (4.8,7.1) | 5.8 (5,6.6) | 0.757 |
Lymphocytes (10^9/L) (median, IQR) | 1.8 (1.4,2.2) | 1.7 (1.5,2.4) | 0.75 |
Red blood cells (10^12/L) (mean ± SD). | 4.5 ± 0.5 | 4.4 ± 0.4 | 0.234 |
Neutrophil (median, IQR) | 3.3 (2.6,4.2) | 3.1 (2.7,3.9) | 0.472 |
NLR (median, IQR) | 1.9 (1.5,2.4) | 1.7 (1.3,2.1) | 0.205 |
Monocytes (10^9/L) (median, IQR). | 0.4 (0.3,0.6) | 0.4 (0.3,0.5) | 0.506 |
Plate (10^9/L) (median, IQR) | 235 (192,271) | 234 (194,270) | 0.712 |
PDW (fL) (median, IQR) | 11.3 (10.2,12.3) | 11.7 (10.4,13.5) | 0.227 |
MPV (fL) (median, IQR) | 10.1 (9.5,10.6) | 10.2 (9.6,11.1) | 0.392 |
ESR (mm/h) (median, IQR) | 8.5 (4, 17) | 15 (8,28) | 0.008 |
CRP > 5mg/L (n, %) | 10 (6.9) | 4 (16) | 0.132 |
PT(s) (median, IQR) | 10.9 (10.5,11.3) | 11 (10.2,11.3) | 0.623 |
APTT(s) (median, IQR) | 30.1 (28.8,32.2) | 31.2 (29,33.4) | 0.291 |
FIB(g/L) (median, IQR) | 2.8 (2.6,3.2) | 3.2 (2.6,3.4) | 0.041 |
TT(s) (mean ± SD) | 15.1 ± 1 | 15.3 ± 1 | 0.5 |
D-dimer (ug/ml) (median, IQR) | 0.3 (0.2,0.6) | 0.4 (0.3,0.5) | 0.142 |
ALT(U/L) (median, IQR) | 17 (13,25) | 13 (11,22) | 0.145 |
Blood glucose (mmol/L) (median, IQR)) | 5 (4.7,5.6) | 5 (4.6,5.2) | 0.334 |
AST(U/L) (median, IQR) | 19 (15,22) | 18 (15,20) | 0.419 |
Albumin (g/L) (mean ± SD). | 44 ± 3.2 | 43.5 ± 2.7 | 0.524 |
Urea nitrogen (mmol/L) (median, IQR) | 5.6 (4.7,6.6) | 5.6 (5,6.4) | 0.922 |
Total cholesterol (mmol/L) (mean ± SD). | 4.9 ± 0.9 | 5 ± 0.9 | 0.448 |
LDL-C(mmol/L) (median, IQR) | 2.8 (2.3,3.3) | 2.9 (2.4,3.7) | 0.228 |
HDL-C(mmol/L) (median, IQR) | 1.2 (1.1,1.4) | 1.2 (1.1,1.4) | 0.738 |
Hcy(umol/L) (median, IQR) | 11.2 (9.7,14.9) | 11.1 (9.9,12.4) | 0.559 |
Blood potassium (mmol/L) (mean ± SD) | 4.1 ± 0.3 | 4.1 ± 0.3 | 0.827 |
Blood sodium (mmol/L) (median, IQR) | 142 (140.8,143) | 142 (141,144) | 0.483 |
Blood magnesium (mmol/L) (mean ± SD) | 0.9 ± 0.1 | 0.9 ± 0.1 | 0.85 |
Blood calcium (mmol/L) (median, IQR) | 2.3 (2.2,2.4) | 2.3 (2.2,2.4) | 0.713 |
Blood chlorine (mmol/L) (median, IQR) | 106 (104,107) | 106 (104,107) | 0.952 |
NLR: Neutrophil-lymphocyte ratio, PDW: Platelet distribution width, MPV: Mean platelet volume, CRP: C reactive protein, PT: Prothrombin time, APTT: Activated partial thromboplastin time, TT: Thrombin time, FIB: Fibrinogen, LDL-C: Low-density lipoprotein cholesterol, HDL-C: High-density lipoprotein cholesterol, ALT: Alanine aminotransferase, AST: Aspartate aminotransferase, Hcy: Homocysteine, BMI: Body mass index |
Regarding preoperative laboratory results, patients in the non-DVT group had significantly lower ESR and FIB than patients in the DVT group (P < 0.05). Additionally, CRP in the non-DVT group was significantly higher than those in the DVT group before surgery (P < 0.05).
Logistic regression analysis
To further evaluate the impact of various indicators on the occurrence of DVT, a single-factor logistic regression analysis was conducted on all indicators. The analysis indicated that the following variables had statistical significance: preoperative hospitalization days, gender, age, ESR, FIB, preoperative ultrasound indicating lower limb venous varicose veins/reflux and lower limb muscle weakness (Table 3).
Table 3
Univariate logistic regression analysis of some indicators
| coefficient | P-value | OR (95% confidence interval). |
---|
age | 0.074 | 0.012 | 1.077 (1.019,1.144) |
man | -0.972 | 0.041 | 0.378 (0.140,0.926) |
Days before surgery | -0.448 | 0.050 | 0.639 (0.397,0.974) |
ESR | 0.024 | 0.059 | 1.025 (0.988,1.052) |
Venous reflux /Varicose veins | 0.785 | 0.078 | 2.192 (0.901,5.241) |
Weakened lower extremity muscle strength | 0.749 | 0.092 | 2.115 (0.871,5.051) |
FIB | 0.422 | 0.093 | 1.525 (0.914,2.549) |
Model building
Variable screening: LASSO regression
The candidate variables were included in the LASSO regression using the training set data. Figure 2a illustrates the change in variable coefficients for different values of λ (lambda). The optimal λ value was determined through 10-fold cross-validation and selected as the largest λ value within one standard error of the minimum mean squared error (Fig. 2b), resulting in an optimal λ value of 0.0496994.
Ultimately, the selected variables included age, preoperative hospitalization days, FIB, gender, and lower limb venous varicose veins/reflux.
Model construction: Logistic regression
The variables screened by LASSO regression were incorporated into the Logistic regression model. Table 4 summarizes the coefficient values of each item in the model.
Table 4
Logistic regression model coefficients: training set
| Coefficient | The standard error of the coefficient estimate | z-value | P-value |
---|
Constant | -4.59694 | 1.73888 | -2.64361 | 0.00820 |
Days before surgery | -0.70344 | 0.20059 | -3.50680 | 0.00045 |
Age | 0.06820 | 0.02559 | 2.66452 | 0.00771 |
FIB | 0.90016 | 0.32671 | 2.75525 | 0.00586 |
Venous reflux/ Varicose veins | 0.51120 | 0.38916 | 1.31360 | 0.18898 |
male | -2.00199 | 0.40273 | -4.97104 | 0.00000 |
Nomogram & Dynamic nomogram
Age, preoperative hospitalization days, FIB, gender, and varicose veins/venous reflux of low extremities were selected as predictive indicators. A nomogram (Fig. 3) was created to facilitate risk prediction for LEDVT.
Based on the established model, a web-based dynamic nomogram was also developed (Fig. 4). The results are presented in two formats. The first format is graphical, which includes the prediction probability values and the corresponding 95% confidence intervals. The second format is numerical, providing more detailed prediction values. Additionally, each calculation result performed before closing the page is displayed simultaneously, allowing for dynamic observation.
This web tool is deployed on the official server of shinyapps.io. The web address is: https://dvt-risk-prediction.shinyapps.io/DynNomapp/
Model evaluation
The model was evaluated using different datasets. ROCs were generated using the training set (Fig. 5a), validation set (Fig. 5b) and external validation set (Fig. 5c), with AUC of 0.803, 0.800 and 0.791 respectively. These AUC values represent the model's performance in distinguishing between cases with and without DVT.
To further evaluate the model, calibration curves were drawn for the training set, validation set, and external validation data (Fig. 6). The calibration curves for all three sets closely aligned with the diagonal line, indicating that the predicted probabilities generated by the model were close to the actual probabilities. This suggests that the model has good accuracy.
For assessing the clinical benefit of the model, decision curves analysis was employed (Fig. 7). The red curve remained above both the "Treat All" and "Treat None" curves when the threshold probability was less than 60%, indicating that the model can obtain some clinical benefits.