Performance statistics of the generic classification tree model.

Note: Accuracy indicates the proportion of correctly classified cases, with associated confidence interval; No information rate (NIR) is 1-[proportion of CAST≥7] in the sample; Difference between model accuracy and NIR tested using a one-sided binomial test; Sensitivity = [number of correctly classified CAST ≥7]/[number of CAST ≥7 in sample]; Specificity = [number of correctly classified CAST <7]/[number of CAST <7 in sample]; Positive predictive value = [number of correctly classified CAST ≥7]/[all classified CAST ≥7]; Negative predictive value = [number of correctly classified CAST <7]/[all classified CAST <7]. The country datasets contain both participants from the training and the validation dataset. Performance statistics of the generic classification tree model.