Objective: To assess the validity of established risk scoring methods for predicting intravenous immunoglobulin (IVIG) resistance in Kawasaki disease in a Caucasian population.
Design: We reviewed the medical records of patients with Kawasaki disease from six medical centers in Israel between 2004 and 2013. The risk scores for each patient were calculated based on the Kobayashi, Sano and Egami scoring methods. The results were analyzed to determine if a higher risk score predicted IVIG resistance in this population. Using our data, we attempted to analyze which factors predicted a lack of response to the initial dose of IVIG.
Results: The cohort included 284 patients, of which 52 (18.3%) did not respond to the first dose of IVIG. The scoring methods for Kobayashi, Sano and Egami were calculated and none were able to reliably predict IVIG resistance. The sensitivities for these scores were 23 to 32% and the specificities 67 to 87%. An attempt to calculate a predictive score specific to this population was also unsuccessful, with a receiver operative characteristic (ROC) of 0.71 (95% CI 0.64-0.77).
Conclusion: The established risk scoring methods created for Japanese populations are not suitable for predicting IVIG resistance in a Caucasian population of Israeli children. Moreover, we failed to create a specific score system that is sufficiently able to predict IVIG resistance for our cohort.
