Development of a predictive model for deep vein thrombosis in burn patients based on the Caprini Risk Assessment Scale.

Abstract

Objective: To explore the applicability of the Caprini Risk Assessment Scale in burn patients for evaluating the risk of deep vein thrombosis (DVT).

Methods: A retrospective case-control study was conducted involving 278 burn patients from March 2021 to March 2023, with an additional independent test set of 119 patients for external validation. Patients were stratified into DVT and non-DVT groups based on the DVT incidence within one month after burn. The Caprini Risk Assessment Scale was employed to calculate scores and determine risk factors. Multivariate logistic regression analyses identified significant risk factors, and receiver operating characteristic (ROC) curves evaluated the model's predictive power.

Results: The mean Caprini score was significantly higher in the DVT group (6.61 ± 2.64) compared to the non-DVT group (4.89 ± 2.36; P < 0.001). Key risk factors included older age, increased body mass index (BMI), and a personal or family history of thrombosis. DVT patients were more prone to higher Caprini scores and classified as 'very high risk'. Logistic regression demonstrated a positive correlation between Caprini scores, risk stratification, and DVT incidence (β = 0.284, OR = 1.329; β = 0.466, OR = 1.594, respectively). The predictive model displayed strong discriminatory power, with an area under the curve (AUC) of 0.853 in the training set and 0.937 in the test set.

Conclusion: The Caprini Risk Assessment Scale is an effective tool for predicting DVT risk in burn patients, aiding in risk stratification and targeted prophylaxis.