Development and validation of a Nomogram for predicting intravenous immunoglobulin-responsive coronary progression in Kawasaki Disease.

Abstract

Background: Coronary artery lesions in children with Kawasaki disease patients have long been a critical clinical concern. While extensive research has focused on predicting intravenous immunoglobulin resistance and optimizing therapeutic strategies, this retrospective study identified a distinct clinical subgroup characterized by intravenous immunoglobulin responsiveness (defined as sustained defervescence without recurrence within 36 hours post-intravenous immunoglobulin therapy) yet exhibiting persistent progression of coronary artery lesions. We have termed this phenomenon intravenous immunoglobulin-Responsive Coronary Progression. Early intervention in this subgroup is crucial for prognosis improvement; however, validated predictive tools are currently lacking.

Methods: Risk factors for intravenous immunoglobulin-responsive coronary progression were analyzed in Kawasaki disease patients meeting American Heart Association criteria. Predictors were selected via Least Absolute Shrinkage and Selection Operator and multivariable logistic regression. Internal validation used bootstrapping (1000 resamples) and five-fold cross-validation. Model performance was assessed via receiver operating characteristic curve, calibration curves, decision curve analysis, and clinical impact curves.

Results: Six predictors were identified: age (years), time interval from fever onset to intravenous immunoglobulin (days), c-reactive protein, erythrocyte sedimentation rate, albumin, and sodium. The Nomogram showed strong discrimination 0.831 (95% CI: 0.788-0.893), and the cross-validation yielded a mean C-index of 0.82. Decision curve analysis and clinical impact curves confirmed clinical utility at threshold probabilities>16%. A web-based calculator was developed for real-time predictions.

Conclusion: This study developed a validated Nomogram incorporating six pretreatment variables to predict intravenous immunoglobulin-responsive coronary progression risk in Kawasaki disease patients, providing a clinically actionable tool for early targeted intervention.