Table 7.

Prognostic abilities of pN and mpN stages in the training and validation cohorts.

Variables	AUC (95% CI)	AIC	P value*
Training cohort			0.006
pN stage	0.618 (0.606–0.630)	30,257
mpN stage	0.628 (0.616–0.640)	30,217
Validation cohort			0.012
pN stage	0.587 (0.568–0.605)	12,466
mpN stage	0.601 (0.582–0.619)	12,442

AUC, Areas under the receiver-operating characteristic curve; AIC, Akaike’s information criterion; CI, confidence interval; mpN, modified pathological N; pN, pathological N.

A higher AUC indicates better discrimination and a lower AIC indicates superior model-fitting;

^*P value based on Hanley & McNeil test, AUC_{pN stage} vs. AUC_{mpN stage}.