Skip to main content
. 2015 Oct 24;17:296. doi: 10.1186/s13075-015-0808-x

Table 3.

ROC classification for best PCR cutoffs reflecting 24hP 0.5 g/day using binary 24hP as the gold standard

Probability level Number of correctly predicted events Number of correctly predicted nonevents Number of nonevents predicted as events Number of events predicted as nonevents Sensitivity Specificity PCR
0.983 65 570 1 597 0.098 0.998 0.138
0.970 66 570 1 596 0.100 0.998 0.128
0.909 66 569 2 596 0.100 0.996 0.108
0.847 552 519 52 110 0.834 0.909 0.098
0.754 579 493 78 83 0.875 0.863 0.088
0.628 604 458 113 58 0.912 0.802 0.078
0.340 634 385 186 28 0.958 0.674 0.058

24hP protein content in a 24-h urine collection sample, PCR untimed sample of urine protein/creatinine ratio, ROC receiver operating characteristic

Boldface type indicates statistically significant values