Objective:To evaluate the Caprini risk assessment model (Caprini RAM) for predicting lower extremity deep vein thrombosis (DVT) after a spinal cord injury (SCI).Methods:Five hundred and one SCI patients were divided into low- (0-1 points, n=130), medium- (2 points, n=43), and high-risk (≥3 points, n=328) groups according to their Caprini scores. The data covering all 501 included gender, age, cause of injury, injury site, American Spinal Injury Association classification, smoking history, surgical history, concurrent pulmonary or urinary tract infections, indwelling catheterization, comorbid type 2 diabetes or hypertension, D-dimer level, fibrinogen level, prothrombin time (PT), activated partial thromboplastin time, international normalized ratio), red blood cell (RBC) count, white blood cell count, platelet (PLT) count and neutrophil percentage. DVTs were detected using lower extremity ultrasound. The Caprini scores were treated as both categorical and continuous variables alongside the other risk factors in multivariate logistic regressions predicting DVT incidence. Generalized additive models were used for curve fitting and threshold saturation analysis, and log-likelihood ratio tests were applied to evaluate the linear relationships observed between Caprini scores and DVT incidence. Results:When the Caprini score was used as a categorical variable, a high risk score ( OR=7.264), age ( OR=1.050), D-dimer ( OR=1.11) , PT ( OR=1.494), PLT count ( OR=1.004) and lung infection ( OR=1.83) were found to be significant independent predictors of DVT. RBC was a protective factor ( OR=0.509). When the Caprini score was used as a continuous variable, lung infection lost its predictive utility. After adjusting for the risk and protective factors, the Caprini scores and the incidence of DVT showed a significant curvilinear increasing association, with a breakpoint of 3. When the Caprini score was less than 3, the incidence of DVT increased 1.83 times with a 1 point increase in the Caprini score. Beyond 3 the multiplier was 1.06. Conclusions:With the independent risk factors accounted for, Caprini scores demonstrate a curvilinear increasing relationship with DVT risk. Clinically, special attention should be given to SCI patients with Caprini scores ≥3.

Objective:To evaluate the Caprini risk assessment model (Caprini RAM) for predicting lower extremity deep vein thrombosis (DVT) after a spinal cord injury (SCI).Methods:Five hundred and one SCI patients were divided into low- (0-1 points, n=130), medium- (2 points, n=43), and high-risk (≥3 points, n=328) groups according to their Caprini scores. The data covering all 501 included gender, age, cause of injury, injury site, American Spinal Injury Association classification, smoking history, surgical history, concurrent pulmonary or urinary tract infections, indwelling catheterization, comorbid type 2 diabetes or hypertension, D-dimer level, fibrinogen level, prothrombin time (PT), activated partial thromboplastin time, international normalized ratio), red blood cell (RBC) count, white blood cell count, platelet (PLT) count and neutrophil percentage. DVTs were detected using lower extremity ultrasound. The Caprini scores were treated as both categorical and continuous variables alongside the other risk factors in multivariate logistic regressions predicting DVT incidence. Generalized additive models were used for curve fitting and threshold saturation analysis, and log-likelihood ratio tests were applied to evaluate the linear relationships observed between Caprini scores and DVT incidence. Results:When the Caprini score was used as a categorical variable, a high risk score ( OR=7.264), age ( OR=1.050), D-dimer ( OR=1.11) , PT ( OR=1.494), PLT count ( OR=1.004) and lung infection ( OR=1.83) were found to be significant independent predictors of DVT. RBC was a protective factor ( OR=0.509). When the Caprini score was used as a continuous variable, lung infection lost its predictive utility. After adjusting for the risk and protective factors, the Caprini scores and the incidence of DVT showed a significant curvilinear increasing association, with a breakpoint of 3. When the Caprini score was less than 3, the incidence of DVT increased 1.83 times with a 1 point increase in the Caprini score. Beyond 3 the multiplier was 1.06. Conclusions:With the independent risk factors accounted for, Caprini scores demonstrate a curvilinear increasing relationship with DVT risk. Clinically, special attention should be given to SCI patients with Caprini scores ≥3.