Abstract
As an acute systemic vasculitis, Kawasaki disease (KD) could develop coronary artery lesions (CAL) sometimes. However, its etiology was still unidentified. This study was to construct a predictive model based on clinical features and laboratory parameters, and then perform a rapid risk assessment of CAL. We collected clinical and laboratory data retrospectively for all patients with KD who were hospitalized at our hospital from January 2016 to June 2023. All the patients were divided into CAL and non-CAL groups and then randomly assigned to a training set and a verification set. The independent risk variables of CAL were identified by univariate analysis and multivariate logistic regression analysis of the training set. These components were then utilized to build a predictive nomogram. Calibration curve and receiver operating characteristic curve were used to evaluate the performance of the model. The predictive nomogram was further validated in the verification set. In the training set, 49 KD patients (19.9%) showed CAL. Compared with the non-CAL group, the proportion of fever days ≥ 10, C-reactive protein and total bilirubin were significantly higher in the CAL group, whereas age was younger, hemoglobin and albumin were lower. Younger age, fever days ≥ 10, higher C-reactive protein, lower hemoglobin and albumin were identified as independent risk factors for CAL in KD patients. The nomogram constructed using these factors showed satisfactory calibration degree and discriminatory power (the area under the curve, 0.764). In the verification set, the area under the curve was 0.798. Younger age, fever days ≥ 10, lower hemoglobin and albumin levels, higher C-reactive protein levels were independent risk factors for CAL in KD patients. The predictive nomogram constructed utilizing 5 relevant risk factors could be conveniently used to facilitate the individualized prediction of CAL in KD patients.