Skip to main content
. 2022 Mar 24;7(6):1210–1218. doi: 10.1016/j.ekir.2022.03.016

Table 2.

Concordance statistics, calibration slopes, NRIs, and IDIs of full models in our cohort

Variables Full model with race Full model without race
Harrell c index 0.845 (95% CI: 0.810–0.879) 0.845 (95% CI: 0.811–0.879)
AUC at 5 yr 0.838 0.819
AUC at 2.8 yr 0.881 0.891
Calibration slope 1.234 (95% CI: 0.973–1.494) 1.211 (95% CI: 0.954–1.468)
Calibration at 5 yr 1.211 1.211
Calibration at 2.8 yr 1.234 1.211
5-year performance compared with the full model without the race
NRI 0.222(95% CI: 0.058–0.383)
NRI (events) 0.444 (95% CI: 0.326–0.567)
NRI (nonevents) −0.222 (95% CI: −0.295 to −0.144)
IDI 0.010 (95% CI: −0.005 to 0.029)
2.8-year performance compared with the full model without the race
NRI 0.347 (95% CI: 0.230–0.482)
NRI (events) 0.451 (95% CI: 0.324–0.579)
NRI (nonevents) −0.105 (95% CI: −0.144 to −0.068)
IDI 0.021 (95% CI: 0.008–0.035)

AUC, area under the curve; IDI, integrated discrimination improvement; NRI, net reclassification improvement.