Objective: To develop a nomogram model to predict the risk of Kawasaki disease (KD) that can inform clinical diagnostic decision making. Methods: A total of 253 children (101 KD and 152 non-KD febrile illnesses) were included in this study. We randomly selected 187 children as the modeling cohort and 66 as the validation cohort for clinical data collection. Univariate and multivariate Logistic regression models were used to analyze the influencing factors of KD in children with fever. A nomogram was developed by R software and validated to predict the risk of KD in children with fever. The area under the ROC curve (ROC-AUC) was used to assess the clinical efficacy of the nomogram. Results: Lymphocyte count

C-reactive protein (CRP)

erythrocyte sedimentation rate (ESR)

platelet count (PLT)

and the percentage of large platelets were the independent factors influencing the risk of KD in children. The internal verification of the nomogram showed that the calibration curve did not deviate significantly from the fit

indicating that the model has good discrimination and accuracy. The ROC-AUC of the modeling group and validation group was 0. 969 and 0. 915

respectively. Conclusion: Based on the risk factors of KD

this study has developed a nomogram model for predicting the risk of KD. The nomogram model with good predictive efficiency can provide reference for screening high-risk populations in the early stage of clinical practice

and formulate intervention strategies as soon as possible.