Development of an Efficient Nomogram for Prediction of Early Deep Vein Thrombosis Following Intracerebral Hemorrhage

Objective: Our objective was to identify the risk factors associated with deep vein thrombosis (DVT) subsequent to intracerebral hemorrhage (ICH), as well as devise a highly effective and personalized predictive model for anticipating DVT occurrence in patients diagnosed with ICH. Methods: Our research recruited a total of 875 consecutive patients diagnosed with ICH, who were then randomly split in a ratio of 3:1. The derivation group consisted of 584 individuals, while the validation cohort included 291 participants. Variables were screened from demographics, comorbidities, clinical symptoms, neuroimaging features, initial treatment in hospital, and laboratory tests. In the derivation cohort, we developed a prediction model with multivariable logistic regression analysis. In the validation cohort, we assessed the model performance and compared it to previously reported models. Results: Multivariable logistic regression analysis showed that older age (OR 1.067, 95%CI 1.040-1.095), impaired consciousness (OR 1.485, 95%CI 1.008-2.189), and infection (OR 1.905, 95%CI 1.075-3.379), were independent risk factors for early DVT. In contrast, rehabilitation (OR 0.227, 95%CI 0.118-0.438) and intact muscle stress (OR 0.828, 95%CI 0.757-0.905) were independent protective factors. We integrated these variables to construct a nomogram model. Compared to previous models, our model is well calibrated, showing higher C statistics in decision curve analysis and better reclassification ability. Conclusions: We developed a new and efficient clinical prediction model for the rapid prediction of the probability of early DVT in patients with ICH.