Table 3.

Relationship between timing of urine neutrophil gelatinase-associated lipocalin measurements and prediction accuracy

Variable	Before CPB	Within 1 h of CPB	Within 1–3 h of CPB	Within 3–6 h of CPB	Within 6–12 h of CPB	Within 12–18 h of CPB	Within 18–24 h of CPB	Within 24–48 h of CPB
Measurements (n)	155	166	180	115	199	176	80	105
Outcome events (n)	14	11	9	13	16	16	12	14
Urine creatinine–normalized NGAL concentration (μg/mmol creatinine)
ROC area (95% CI)a	0.689 (0.552 to 0.886)	0.736 (0.581 to 0.881)	0.719 (0.541 to 0.897)	0.755 (0.553 to 0.958)	0.762 (0.622 to 0.901)	0.775 (0.662 to 0.888)	0.747 (0.625 to 0.869)	0.441 (0.268 to 0.614)
Absolute urine NGAL concentration (ng/ml)
ROC area (95% CI)a	0.679 (0.512 to 0.847)	0.656 (0.458 to 0.853)	0.675 (0.510 to 0.841)	0.796 (0.632 to 0.960)	0.818 (0.708 to 0.927)	0.783 (0.675 to 0.892)	0.729 (0.572 to 0.887)	0.466 (0.288 to 0.645)
Urine NGAL excretion rate (ng/hr)
ROC area (95% CI)a	0.583 (0.418 to 0.749)	0.725 (0.572 to 0.878)	0.668 (0.454 to 0.882)	0.700 (0.525 to 0.876)	0.672 (0.531 to 0.812)	0.594 (0.445 to 0.743)	0.638 (0.473 to 0.803)	0.548 (0.349 to 0.746)

The outcome events were dialysis requirement and/or in-hospital death. CPB, cardiopulmonary bypass; NGAL, neutrophil gelatinase-associated lipocalin; ROC, receiver-operating characteristic; 95% CI, 95% confidence interval.

^a95% CIs were estimated by boot-strapping with 1000 resamples.