Risk scores for Kawasaki disease, a management tool developed by the KAWA-RACE cohort

Abstract

Introduction/objectives: Asian scores developed to predict unresponsiveness to intravenous immunoglobulin (IVIG) or development of coronary artery aneurysms (CAA) in patients with Kawasaki disease (KD) are not appropriate in Western populations. The purpose of this study is to develop 2 scores, to predict unresponsiveness to IVIG and development of CAA, appropriate for Spanish population.

Method: Data of 625 Spanish children with KD collected retrospectively (2011-2016) were used to identify variables to develop the 2 scores of interest: unresponsiveness to IVIG and development of CAA. A statistical model selected best variables to create the scores, and scores were validated with data from 98 patients collected prospectively.

Results: From 625 patients of the retrospective cohort, final analysis was performed in 439 subjects: 37 developed CAA, and 212 were unresponsive to IVIG. For the score to predict CAA, a cutoff ≥ 8 was considered for high risk, considering a score system with a different weight for each of the eight variables. External validation showed a sensitivity of 22% and a specificity of 75%. The score to predict unresponsiveness to IVIG established a cutoff ≥ 8 for high risk, considering a score system with a different weight for each of the nine variables. External validation showed a sensitivity of 78% and a specificity of 50%.

Conclusions: Two risk scores for KD were developed from Spanish population, to predict development of CAA and unresponsiveness to IVIG; validation in other cohorts could help to implement these tools in the management of KD in other Western populations.

Keywords: Coronary artery aneurysm; Kawasaki disease; Score; Unresponsiveness IVIG.