Validation of Kawasaki MATCH on a Latin American cohort: application to the REKAMLATINA network

Tremoulet A.H.; Lam J.Y.; Ulloa-Gutierrez R.; Burns J.C.; Nemati S.; Gardiner M.A.; Mesa-Monsalve J.G.; Díaz-Díaz A.; Martínez-Ramírez R.; Rubio-Pérez N.; Passos S.D.; Hernández-Magaña R.; Soza G.; Pujadas M.; López P.; Díaz-Maldonado A.; López-Gallegos D.; Yamazaki-Nakashimada M.A.; Daza C.; Assandri E.; Lovera D.; Borzutzky A.; Melgar M.; Izquierdo G.; Grazioso C.F.; Collia A.; Franco L.; Pérez-Camacho P.; Gámez-González L.B.; Pirez M.C.; López-Medina E.; Marques H.H.S.; Alvarez-Olmos M.I.; Viviani T.; Camacho-Moreno G.; Rodríguez-Quiroz F.J.; Luciani K.; Dueñas L.; Martínez-Medina L.; Saltigeral-Simental P.; Chacon-Cruz E.; Cofré F.; Beltrán S.; Miño-León G.; Faugier-Fuentes E.; Gómez V.; del Aguila O.; Estripeaut D.; Garrido-García L.M. · Pediatric Research (2026) · DOI: 10.1038/s41390-026-05190-2

Resumen

Impact statement: Key message: In a large multinational validation study, the Kawasaki MATCH machine-learning clinical decision support tool accurately identified patients with Kawasaki Disease (KD) using data from the REKAMLATINA network. What this study adds: This is the first international validation of Kawasaki MATCH. Prospective and retrospective validation of the model across numerous diverse Latin American clinical settings demonstrates consistent performance despite differences in laboratory availability, data completeness, and practice patterns. Impact: These findings support the use of AI-assisted decision support to improve recognition of KD, reduce diagnostic delay, and potentially prevent coronary artery complications in children across varied health systems. © The Author(s), under exclusive licence to the International Pediatric Research Foundation, Inc 2026.