A predictive model for long-term coronary artery lesion risk in Kawasaki disease

Area of Science:

• Cardiology
• Pediatrics
• Medical Informatics

Background:

• Kawasaki disease (KD) is a significant pediatric illness.
• Coronary artery lesions (CALs) are a major complication of KD, potentially leading to severe cardiovascular events.
• Identifying risk factors for CALs is crucial for timely management and improved patient outcomes.

Purpose of the Study:

• To develop and validate a nomogram model for predicting the long-term risk of CALs in pediatric patients with KD.
• To identify key clinical and laboratory predictors associated with the development of CALs in KD.

Main Methods:

• Retrospective analysis of 2,481 pediatric KD patients' clinical data, lab results, and echocardiographic findings.
• Multivariate logistic regression to identify risk factors and construct a predictive nomogram.
• Model performance evaluation using ROC curves, AUC, calibration curves, and DCA.

Main Results:

• Male sex, prolonged hospitalization, prolonged fever, decreased hemoglobin (Hb), decreased hematocrit (HCT), and hyponatremia were identified as significant predictors of long-term CAL risk.
• The nomogram demonstrated good predictive performance with AUCs of 0.801 (training) and 0.796 (validation).
• The model showed high accuracy, good calibration, and significant clinical utility as indicated by DCA.

Conclusions:

• The developed nomogram is an accurate and effective tool for physicians to identify KD patients at high risk for long-term CALs.
• This predictive model can facilitate early detection and intervention strategies for KD patients.
• The findings underscore the importance of considering specific clinical and laboratory parameters in assessing CAL risk in KD.