Hibiscus score: Developing and validating a predictive tool for intravenous immunoglobulin treatment resistance in Malaysian children with Kawasaki disease

BACKGROUND: Children with intravenous immunoglobulin (IVIG) resistant Kawasaki disease (KD) are at higher risk of developing coronary artery (CA) aneurysm. Early identification of high-risk patients using a predictive tool would allow for earlier interventions to prevent cardiac complications.

METHODS: Children with KD who were admitted to five selected hospitals in Malaysia between 2008 and 2018 and received 2 g/kg of IVIG within 10 days from the onset of illness were included. Predictors of IVIG resistance in KD were determined using multiple logistic regression analysis. An optimal cut-off point was set using receiver operative characteristic curve and a final multiple logistic regression analysis was performed entering these cut-off points. A new scoring system was constructed.

RESULTS: A total of 276 patients were included. IVIG resistance occurred in 9.1 % of them. Total bilirubin [OR 7.37; 95 % CI (2.18, 24.83)], male sex [OR 0.34; 95 % CI (0.10, 1.19)], C-reactive protein (CRP) [OR 0.17; 95 % CI (0.02, 1.38)] and neutrophils [OR 0.25; 95 % CI (0.05, 1.21)] were found to be significant predictors for IVIG resistance. The findings led to the development of a new predictive tool called the Hibiscus score, which scored 1 point each for neutrophils ≥60 %, CRP ≥80 mg/L, and male sex, while total bilirubin ≥9.4 μmol/L scored 2 points. A cut-off point of ≥4 with this prediction score yielded a sensitivity of 78.9 % and specificity of 80.5 %, with area under the curve of 0.835 [95 % CI (0.752, 0.919)]. CA aneurysms occurred in 6.7 % of IVIG responders and 32 % of IVIG-resistant children (p