Early risk stratification for coronary artery aneurysms in Kawasaki disease: a predictive modeling approach

Area of Science:

• Cardiology
• Pediatrics
• Rheumatology

Background:

• Medium-to-giant coronary artery aneurysms (MGCAA) represent a severe complication of Kawasaki disease (KD).
• Current risk stratification tools are inadequate for predicting MGCAA, lacking validation and specific design for this outcome.
• Early identification of KD patients at high risk for MGCAA is crucial for timely intervention.

Purpose of the Study:

• To develop and validate a transparent clinical prediction model for medium-to-giant coronary artery aneurysms (MGCAA) in Kawasaki disease.
• To provide a tool for personalized risk assessment to guide clinical management and follow-up strategies.

Main Methods:

• Development of a transparent prediction model incorporating six clinical variables: hemoglobin, diagnosis time, mucosal changes, rash, triglycerides, and neutrophil percentage.
• Validation of the model in two independent Chinese cohorts.
• Recalibration of the model for potential broader applicability.

Main Results:

• The transparent model demonstrated predictive capability for MGCAA in KD patients.
• Validation in Chinese cohorts confirmed the model's performance.
• Simple recalibration methods were effective for adapting the model.

Conclusions:

• A transparent, six-variable model offers a validated approach for predicting MGCAA in Kawasaki disease.
• This model, accessible via a web tool, supports personalized risk assessment and clinical decision-making for monitoring and follow-up.
• The tool assists clinicians in managing KD patients at risk for severe coronary artery complications without relying on fixed treatment thresholds.