A prediction model for post-treatment presence of coronary artery abnormality before initial treatment in Kawasaki disease in Japan

Le résumé fourni par la source

Background This study aimed to develop and validate a predictive model for the presence of coronary artery abnormality (CAA) after treatment using pre-treatment clinical and laboratory parameters, including coronary Z -scores, in Japanese patients with Kawasaki disease (KD). Method and results A retrospective multicenter cohort study was conducted, analyzing 1,565 patients diagnosed with KD across eight medical institutions within the Wakayama Kawasaki Disease Clinical Research Group, with validation performed at Chiba University and Tokyo Women's Medical University. A predictive model was developed using data from a primary cohort ( n = 970) and validated in both internal ( n = 333) and external ( n = 262) cohorts. Multivariate analysis identified three predictors of post-treatment CAA presence: maximum pre-treatment Z -score ≥1.6 (2 points), albumin level ≤3.1 g/dL (1 point), and age ≤12 months (1 point). A total score of ≥2 predicted CAA with 84.2% sensitivity and 60.8% specificity in the development cohort, with similar performance validated in the internal and external cohorts (area under the receiver operating characteristic curve: both 0.88). Conclusions The developed model accurately predicts post-treatment CAA presence, emphasizing the importance of early coronary Z -score assessment. It could guide intensive initial therapies to reduce CAA incidence, supporting KD management. However, further validation in diverse populations is recommended.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.