Comparative Analysis of Padua and Caprini Scores in Predicting Venous Thromboembolism Risk Among Nonagenarians: A Cross-Sectional Study From Rugao

Abstract

Objective

To evaluate and compare the predictive performance of Padua and Caprini scores for venous thromboembolism (VTE) risk assessment in individuals aged ≥ 90 years.

Methods

A cross-sectional study was conducted among 511 nonagenarians in Rugao, China. Participants underwent comprehensive clinical assessments including both Padua and Caprini risk scoring. VTE events were monitored through the follow-up period. The predictive efficacy of both scoring systems was analyzed using receiver operating characteristic (ROC) curves, and risk factors were evaluated through multivariate logistic regression.

Results

During follow-up, 31 participants (6.07%) developed VTE. The VTE group demonstrated significantly higher mean Padua scores (4.97 ± 2.21 vs. 4.11 ± 2.45, p = 0.0463), Caprini scores (6.39 ± 2.42 vs. 5.02 ± 2.39, p = 0.0044), and D-dimer levels (median 2.79 vs. 1.31 mg/L, p = 0.0133) compared to the non-VTE group. Both scoring systems showed moderate predictive capability, with the Padua score achieving an area under the curve (AUC) of 0.625 (95% CI: 0.533–0.717) and the Caprini score showing an AUC of 0.679 (95% CI: 0.590–0.768). Optimal cutoff values were 3.5 points for the Padua score (sensitivity 80.65%, specificity 50.42%) and 4.5 points for the Caprini score (sensitivity 77.42%, specificity 51.46%). Multivariate analysis identified atrial fibrillation (OR 4.130, 95% CI: 1.667–9.673, p = 0.001) and elevated Caprini score (OR 1.310, 95% CI: 1.073–1.582, p = 0.006) as significant independent risk factors for VTE. Interestingly, hypertension showed an unexpected protective association with VTE risk (OR 0.400, 95% CI: 0.162–0.907, p = 0.035).

Conclusions

While both Padua and Caprini scores demonstrate moderate predictive value for VTE risk in nonagenarians, their accuracy suggests the need for age-specific refinement.