Skip to main content

Table 3 C-indices for testing of Canadian and Nordic prediction models

From: Validation of prediction models of severe disease course and non-achievement of remission in juvenile idiopathic arthritis: part 1—results of the Canadian model in the Nordic cohort

Prediction modelSevere disease course outcomeNon-achievement of remission outcomeValidation method
Original Canadian model0.85 (0.83–0.87)0.66 (0.63–0.68)External validation (bootstrapping)
Canadian model fine-tuned for Nordic population0.85 (0.81–0.89)0.69 (0.65–0.73)Fine-tuning (repeated random splits)
Nordic model0.90 (0.86–0.92)0.78 (0.72–0.82)aInternal validation (repeated random splits)
  1. C-indices with median interquartile range (IQR)
  2. C-index presented includes patients with systemic JIA, except for athe C-index for the Nordic model and the outcome non-achievement of remission previously published by Rypdal et al. [19]