Developing and validating risk predicting models to assess venous thromboembolism risk after radical cystectomy

Area of Science:

• Urology
• Oncology
• Cardiovascular Medicine

Background:

• Radical cystectomy (RC) patients face a high risk of venous thromboembolism (VTE).
• Existing predictive models, like the Caprini risk assessment (CRA) model, have limitations in accuracy.
• There is a need for improved VTE risk prediction in the post-RC population.

Purpose of the Study:

• To develop and validate novel predictive models for VTE risk following radical cystectomy.
• To compare the performance of the new models against the established CRA model.
• To identify a more accurate and clinically applicable tool for VTE risk stratification in RC patients.

Main Methods:

• A single-center retrospective study of 272 RC patients (2010-2019).
• Patients were randomly assigned to training and testing cohorts.
• Multivariate and stepwise logistic regression were used to develop two novel predictive models (Model A and Model B), compared against the CRA model using NRI, IDI, and ROC analyses.

Main Results:

• 36 out of 272 patients developed VTE post-RC.
• Both Model A and Model B demonstrated superior predictive accuracy over the CRA model (ROC AUCs 0.806 and 0.833, respectively).
• The novel models showed significant improvements in classification and prediction (NRI >0, IDI >0, P<0.01) and favorable clinical utility in decision curve analysis.

Conclusions:

• The developed models significantly outperform the current CRA model for VTE risk prediction after RC.
• Model A, with fewer variables, is particularly advantageous for ease of use and rapid risk assessment.
• These models can facilitate timely interventions, potentially improving outcomes for high-risk RC patients.