Abstract
AIM: To develop a nomogram for predicting calf muscular venous thrombosis (CMVT) in patients with acute ischemic stroke (AIS). DESIGN: A cross-sectional study. METHODS: From July 2021 to December 2023, a total of 1084 cases of AIS patients were selected from the hospital. The R software was used to randomly divide the data into training and validation sets in a 7:3 ratio. The training set was utilized for constructing the prediction model, while the validation set was employed to verify its performance. Logistic regression analysis, including single-factor and multiple-factor logistic backward stepwise regression methods based on the principle of minimizing Akaike information criterion (AIC), was conducted to identify predictors for CMVT in AIS patients. A nomogram model visualization was created, and ROC curves were generated for both the training and validation sets with corresponding calculation of area under the curve (AUC) to evaluate model differentiation. Calibration curves were plotted, and Hosmer-Lemeshow test was performed to assess model calibration. RESULTS: Age, gender, antiplatelet therapy, D-dimer and lower limb paralysis are identified as predictors of CMVT in AIS patients. The internal model validation results demonstrate an area under the ROC curve of 0.868 (95% CI:0.810–0.926), a Model Brier value of 0.080, and a Hosmer-Lemeshow test p-value of 0.179. CONCLUSION: We developed a forecasting model, the model has good prediction efficiency, help to predict the probability of CMVT AIS patients in fast. CLINICAL TRIAL NUMBER: Not applicable.