Incremental value of ultrasound-based carotid plaque-RADS beyond stenosis degree in discriminating symptomatic carotid atherosclerosis
摘要
To determine whether ultrasound-based Carotid Plaque‑RADS adds incremental value beyond stenosis degree in distinguishing symptomatic from asymptomatic carotid atherosclerosis.
MethodsPatients with carotid artery stenosis diagnosed by ultrasound were retrospectively enrolled. Plaque-RADS scores and stenosis degree were independently assessed by two blinded radiologists. Binary logistic regression identified factors associated with symptoms. Model discrimination was evaluated by AUC with DeLong comparison. Incremental value was quantified by continuous NRI and IDI with bootstrap validation. Calibration was assessed using bootstrap-corrected slopes, and decision curve analysis evaluated clinical net benefit.
ResultsA total of 161 patients were included (mean age 70.99 ± 9.86 years; 73.3% male). Plaque-RADS (OR = 3.51, P = 0.0058), stenosis (OR = 4.12, P = 0.0002), and age (OR = 1.07, P = 0.0053) were independent symptom predictors. Adding Plaque-RADS to stenosis improved AUC from 0.762 to 0.807 (P = 0.039). The combined model (Plaque-RADS, stenosis, age) achieved the highest AUC (0.838), though the gain over the model (stenosis plus age) was not significant (P = 0.080). All models showed good calibration (slopes 0.951–0.999). Adding Plaque-RADS to stenosis plus age significantly improved reclassification (continuous NRI = 0.432, P = 0.004; IDI = 0.058, P = 0.002). Decision curve analysis showed net benefit at higher thresholds (0.20–0.30), but no added benefit at lower screening thresholds.
ConclusionUltrasound-based Plaque-RADS provides independent incremental discriminative value beyond stenosis, primarily improving identification of symptomatic patients, supporting its utility as an adjunctive risk stratification tool for clinical decision-making.
Highlights