Abstract
Background: Kawasaki disease (KD) is the most frequent cause of coronary artery aneurysm (CAA) in children. This study tried to evaluate the accuracy of different KD scores developed for prediction of CAA, in an Iranian population.
Methods: This is a cross-sectional retrospective investigation on pediatric patients with a diagnosis of KD. Clinical manifestations, laboratory, and echocardiographic data were recorded. Five Kawasaki scores, including Kobayashi, Egami, Sano, Nakano, and Harada, were assessed and analyzed in relation to CAA and intravenous immunoglobulin (IVIG) resistance.
Results: During five years, we recruited 121 cases of KD under 13 years of age. The rates of CAA and IVIG resistance were 16.5%, and 13.2% respectively. The IVIG resistance group was significantly younger than responder patients. All five scores had low sensitivity in predicting CAA or IVIG resistant cases; the highest sensitivity pertained to the Harada score with 50% sensitivity and 59% specificity (the area under the curve: 0.545, with a 95% confidence interval: 0.423 to 0.667) in predicting CAA, which is lower than the usual acceptable criteria for a screening test. The specificity of all other scores were more than 85% in predicting CAA or IVIG resistance. Gender, fever before therapy and laboratory data showed no significant difference between the groups.
Conclusion: The Kobayashi, Egami, Sano, Nakano and Harada scores have limited usefulness in the Iranian population to predict high risk patients for coronary artery involvement or IVIG resistance; in our study, age under one year was a risk factor for IVIG resistance.