Research on risk factors for deep vein thrombosis in elderly patients with upper limb fractures and establishment of Nomogram prediction model

Abstract

Purpose In this study, we aimed to investigate the risk factors for the development of deep vein thrombosis (DVT) in elderly patients with upper extremity fractures, construct and validate a Nomogram prediction model. Methods Based on the inclusion and exclusion criteria, we retrospectively analyzed 359 patients with upper limb fractures over the age of 70 who were admitted to the Affiliated Hospital of Shandong University of Traditional Chinese Medicine from September 2020 to September 2023. All the patients underwent color Doppler vascular ultrasound of all four limbs to determine the presence or absence of upper and lower limb venous thrombosis before surgery. The patients were divided into DVT and non-thrombosis groups based on thrombosis. Along with the prevalent complications in clinical elderly patients, the sex, smoking, the use of indwelling needle, hypertension, fracture site, hyperlipidemia, atrial fibrillation, diabetes, premature beat, body mass index (BMI), preoperative plasma D-dimer levels, and hemoglobin levels of the patients in the two groups were analyzed by performing univariate and multivariate analysis. We also investigated the relationship between different factors and DVT in patients. All patients were divided into a modeling group (n = 251) and validation group (n = 108) in a ratio of 7:3. Logistic regression was used to construct a Nomogram prediction model and internal validation was performed. The receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) were drawn to evaluate the predictive efficiency, accuracy and clinical benefit of the nomogram model, and the validation group was used to evaluate the feasibility of the nomogram. Results In total, 38 cases of DVT were found in 359 patients, and the incidence rate of thrombosis was 10.58%. High blood pressure, hyperlipidemia, diabetes, anemia, BMI > 25kg/m² and shoulder periarticular fracture are independent risk factors for deep vein thrombosis in upper limb fractures. The results of univariate and multivariate analysis revealed significant differences in terms of smoking, hyperlipidemia, diabetes, atrial fibrillation, anemia, fracture location, indwelling needle, and BMI between the two groups (p < 0.05). The periarticular fracture of the shoulder joint is a common site for the occurrence of DVT. No significant difference was found in terms of plasma D-dimer levels and premature beats with thrombosis between the two groups (p < 0.05). Furthermore, we also found that DVT is not associated with a history of hypertension but is associated with blood pressure levels, which showed significant differences between the two groups (p < 0.05). The ROC curve analysis showed that the area under the curve (AUC) of the prediction nomograph model was 0.972(95%CI: 0.952–0.992)for the modeling group and 0.860༈95%CI༚0.699-1.000༉for the validation group, with a sensitivity of 96.3% and 81.8%, and a specificity of 88.8% and 87.6%, respectively, with a good discrimination. The calibration curve of two groups showed that the predictive probability of the model was in high consistency with the actual incidence. The decision curve analysis (DCA) results in both groups showed that the nomograph model had good net returns. Conclusions The high-risk factors for the development of DVT in elderly patients with upper extremity fracture include smoking, high blood pressure, diabetes mellitus, hyperlipidemia, atrial fibrillation, anemia, BMI > 25kg/m², indwelling needle, and periprosthetic shoulder fracture. These factors should be paid attention to in clinical settings, and the relevant evaluation methods should be improved, to prevent the occurrence of DVT and pulmonary embolism (PE).