Response to Vidal et al. regarding “External validation of the STARZ score highlights structural limits of neonatal AKI prediction models”