Skip to main content
. Author manuscript; available in PMC: 2013 Nov 5.
Published in final edited form as: Ann Appl Stat. 2013 Sep 1;7(3):10.1214/13-AOAS634SUPP. doi: 10.1214/13-AOAS634SUPP

Table 5.

The constrained maximum likelihood estimators calculated with data from the PCA3 study.

Initial Biopsy Repeat Biopsy
Combined-data Population specific Combined-data Population specific
ROC(t)
t = 0.1 Est 0.266 (0.209, 0.339) 0.274 (0.205, 0.368) 0.266 (0.209, 0.339) 0.256 (0.166, 0.369)
Eff★ 1.57 1.00 2.53 1.00
t = 0.3 Est 0.591 (0.518, 0.671) 0.600 (0.510, 0.704) 0.591 (0.518, 0.671) 0.578 (0.462, 0.697)
Eff 1.57 1.00 2.41 1.00
t = 0.5 Est 0.772 (0.710, 0.831) 0.778 (0.703, 0.854) 0.772 (0.710, 0.831) 0.764 (0.670, 0.847)
Eff 1.64 1.00 2.21 1.00
t = 0.7 Est 0.886 (0.836, 0.927) 0.889 (0.827, 0.939) 0.886 (0.836, 0.927) 0.882 (0.810, 0.940)
Eff 1.67 1.00 2.19 1.00
t = 0.9 Est 0.967 (0.943, 0.982) 0.967 (0.937, 0.986) 0.967 (0.943, 0.982) 0.966 (0.930, 0.987)
Eff 1.67 1.00 2.32 1.00
CDFR(p) p = 0.65 p = 0.20
Est 0.855 (0.711, 1.000) 0.843 (0.698, 1.000) 0.418 (0.276, 0.564) 0.406 (0.211, 0.586)
Eff 1.12 1.00 1.60 1.00
Risk(y) y = 60 y = 20
Est 0.654 (0.567, 0.731) 0.659 (0.569, 0.749) 0.208 (0.147, 0.270) 0.214 (0.135, 0.292)
Eff 1.22 1.00 1.59 1.00

Eff★: efficiency relative to the population-specific estimator