Kawasaki disease (KD) is an acute pediatric vasculitis in which 10%–20% of patients show resistance to initial intravenous immunoglobulin (IVIG) therapy, increasing the risk of coronary artery complications. Several Japanese risk scores—Kobayashi, Egami, and Sano—are widely used to predict IVIG resistance, but their applicability to other populations remains uncertain. This study aimed to validate these Japanese risk models in a large nationwide Korean cohort and examine their age-specific performance. This nationwide retrospective cohort study analyzed data from the Korean nationwide KD survey (2015–2017), including 10,352 children diagnosed with KD who received standard IVIG treatment. IVIG resistance was defined as persistent or recurrent fever ≥ 36 hours after IVIG completion. The Kobayashi, Egami, and Sano scores were calculated using pre-treatment laboratory parameters, and predictive performance was assessed using sensitivity, specificity, positive and negative predictive values (PPV/NPV), accuracy, and area under the receiver operating characteristic curve (AUC). Subgroup analyses were performed by age (< 6 months, 6–11 months, 12–59 months, ≥ 60 months). Overall IVIG resistance occurred in 16.3% of patients. The proportion classified as high risk was 26.3% (Kobayashi), 30.0% (Egami), and 16.7% (Sano). Across all ages, sensitivities ranged from 0.33–0.47, specificities from 0.73–0.87, and AUCs from 0.64–0.66. Infants aged 6–11 months showed higher discrimination, with AUCs increasing from 0.63 to 0.68., whereas no subgroup exceeded an AUC ≥ 0.70. All three scores demonstrated high NPV (0.87–0.88) but low PPV (0.26–0.33), indicating stronger ability to exclude rather than predict IVIG resistance. In this nationwide Korean cohort, Japanese IVIG-resistance risk scores exhibited consistently low sensitivity, high specificity, and only modest discrimination. These findings confirm limited cross-population transportability of Japanese models and underscore the need for recalibrated or newly developed Korean-specific prediction tools to improve early identification of IVIG-resistant KD patients.
No takes yet. Share an insight, caveat, or question.
Choi et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: