Table 5.

Overall and age-specific discrimination of the Navarre and Gail prediction models among women in the Navarre Breast Cancer Screening Program cohort, 1996–1998 to 2005

Age (years)	No. of invasive breast cancers	No. of comparable pairsa	Navarre prediction model		Gail prediction model		Discrimination difference (95 % CI)c
			No. of concordant pairsb	Discrimination index (95 % CI)c	No. of concordant pairsb	Discrimination index (95 % CI)c
45–49	49	44,932	25748.5	0.573 (0.497–0.649)	25879.5	0.576 (0.501–0.651)	−0.003 (−0.015 to 0.009)
50–54	185	378,725	206573.5	0.545 (0.502–0.589)	209957.5	0.554 (0.512–0.597)	−0.009 (−0.031 to 0.013)
55–59	202	447,830	240137.5	0.536 (0.492–0.580)	243176.5	0.543 (0.499–0.587)	−0.007 (−0.029 to 0.016)
60–64	193	368,818	199167.5	0.540 (0.496–0.584)	195754.5	0.531 (0.488–0.574)	0.009 (−0.001 to 0.019)
65–69	166	271,740	145556.5	0.536 (0.490–0.581)	145508.5	0.535 (0.490–0.581)	0.000 (−0.011 to 0.011)
70–74	40	31,467	19753.0	0.628 (0.531–0.725)	19683.0	0.626 (0.534–0.717)	0.002 (−0.038 to 0.042)
Overall				0.542 (0.521–0.564)		0.544 (0.523–0.565)	−0.002 (−0.011 to 0.007)

To correct for the optimistic bias arising from assessing discrimination of the Navarre prediction model on the same data used to fit the model, a 10-fold cross-validation was used in which separate discrimination indexes were estimated for each 10 % random subcohort based on hazard ratios estimated from the remaining 90 % of women in the Navarre Breast Cancer Screening Program cohort and the resulting estimates were combined over the 10 subcohorts (see statistical appendix in Supplementary Material 1)

^aNumber of woman pairs in which their actual times to invasive breast cancer in the corresponding age interval can be ranked (two different event times or a censoring time equal to or longer than an event time)

^bNumber of comparable pairs in which the woman with shorter event time had a higher relative risk of developing breast cancer in the corresponding age interval according to the Navarre or Gail prediction model. When predicted risks were identical for a woman pair (the same risk factor pattern), a value of 0.5 was added to the number of concordant pairs

^cAge-specific discrimination indexes for the Navarre and Gail prediction models were estimated as the proportion of concordant pairs over the total number of comparable pairs in the corresponding age interval. Jackknife methods were used to compute 95 % confidence intervals (CIs) for each model’s age-specific discrimination as well as for the difference in age-specific discrimination between models. Overall discrimination indexes were calculated as the average of age-specific discrimination indexes with weights proportional to the number of comparable pairs in each age interval