Table 4. The association of manual or computer predicted 2-tiered grade with overall survival in 118 discordant cases.

	Manual Grade			Computer Predicted Grade
	Hazard Ratio	(95% CI)	p-value	Hazard Ratio	(95% CI)	p-value
A. Cases with identical grades between Pathologist 1 and Pathologist 2 (31 events out of 76 cases)
Model A: Crude	0.99	(0.44, 2.23)	0.99	2.05	(1.00, 4.21)	0.05
Model B: Adjusted for Age and Gender	1.04	(0.46, 2.34)	0.93	2.42	(1.13, 5.20)	0.02
Model C: Adjusted for Age, Gender, and Stage	1.18	(0.52, 2.68)	0.69	1.89	(0.87, 4.12)	0.11
B. Cases with different grades between Pathologist 1 and Pathologist 2 (21 events out of 46 cases)
Grade assigned by Pathologist 1
Model A: Crude	0.64	(0.19, 2.20)	0.48	2.49	(0.83, 7.45)	0.10
Model B: Adjusted for Age and Gender	0.63	(0.18, 2.17)	0.46	2.49	(0.72, 7.28)	0.16
Model C: Adjusted for Age, Gender, and Stage	0.59	(0.17, 2.03)	0.40	2.03	(0.62, 6.66)	0.24
Grade assigned by Pathologist 2
Model A: Crude	1.56	(0.46, 5.31)	0.48	NA	NA	NA
Model B: Adjusted for Age and Gender	1.58	(0.46, 5.44)	0.46	NA	NA	NA
Model C: Adjusted for Age, Gender, and Stage	1.70	(0.49, 5.89)	0.40	NA	NA	NA

Confidence Interval, CI