Introduction: Kawasaki disease (KD) is the leading cause of acquired heart disease in children, and early identification of patients at risk for severe inflammation and coronary artery complications remains a major clinical challenge. This study aimed to define predictors of severe KD and to identify early determinants of persistent or delayed coronary artery normalization.
Methods: This retrospective, cross-sectional, single-center cohort study included 104 children with KD followed between 2009 and 2025. Clinical, laboratory, echocardiographic, and therapeutic characteristics were compared between severe and non-severe KD groups, defined according to coronary involvement, treatment refractoriness, and need for intensive therapy. Predictors of severe KD were assessed using univariate and multivariate logistic regression. ROC analyses evaluated the discriminatory performance of clinical and laboratory parameters for predicting persistent coronary abnormalities, while Kaplan–Meier models assessed factors associated with time to coronary normalization.
Results: Severe KD accounted for 22.1% (n=23) of the cohort and was associated with younger age, diagnostic delay, higher inflammatory burden, and more intensive treatment requirements. Medium/giant aneurysms and intracoronary thrombosis were observed exclusively in severe cases. In multivariate logistic regression, low serum albumin emerged as the only independent predictor of severe KD (OR=0.041; p=0.040). Baseline coronary Z-scores were the strongest predictors of persistent coronary abnormalities (LMCA AUC=0.963; RCA AUC=0.850). Prolonged hospitalization and delayed laboratory normalization also showed high discriminatory accuracy. Kaplan–Meier analyses demonstrated significantly prolonged coronary normalization times in severe disease, low albumin, diagnostic delay, and higher Z-score strata.
Conclusion: In this cohort, hypoalbuminemia was the sole independent predictor of severe KD, while baseline coronary Z-scores strongly predicted persistent coronary abnormalities. These findings support the use of early laboratory and echocardiographic markers for risk stratification and individualized management in KD.
Keywords: Coronary Artery Abnormalities; Echocardiography; Hypoalbuminemia; Intravenous Immunoglobulins; Kawasaki Disease