The Use of Machine Learning to Create a Risk Score to Predict Survival in Patients with Hepatocellular Carcinoma: A TCGA Cohort Analysis

Table 1

Comparison of demographics by risk score.

	All patients		0 points		1 point		2 points		3 points		p value
Gender											0.026
Female	119	32.80%	15	20.50%	37	33.00%	34	33.00%	33	44.00%
Male	244	67.20%	58	79.50%	75	67.00%	69	67.00%	42	56.00%
Age, mean STD	60	13	60	13	61	12	60	13	56	14	n.s.
Race											n.s.
White	177	48.80%	43	58.90%	56	50.00%	44	42.70%	34	45.30%
Black	17	4.70%	3	4.10%	6	5.40%	5	4.90%	3	4.00%
Asian	157	43.30%	23	31.50%	47	42.00%	52	50.50%	35	46.70%
Unknown/other	12	3.30%	4	5.50%	3	2.70%	2	1.90%	3	4.00%
Etiology											n.s.
No risk factor	88	24.20%	13	17.80%	26	23.20%	30	29.10%	19	25.30%
Alcohol	108	29.80%	30	41.10%	30	26.80%	25	24.30%	23	30.70%
HBV	74	20.40%	13	17.80%	27	24.10%	18	17.50%	16	21.30%
HCV	34	9.40%	4	5.50%	7	6.30%	13	12.60%	10	13.30%
NAFLD	18	5.00%	5	6.80%	7	6.30%	3	2.90%	3	4.00%
Other/unknown	41	11.30%	1	1.40%	6	5.40%	5	5.00%	1	1.40%
Stage											n.s.
1	167	48.70%	34	50.00%	61	58.10%	44	44.90%	28	38.90%
2	85	24.80%	16	23.50%	20	19.00%	24	24.50%	25	34.70%
3	84	24.50%	16	23.50%	23	21.90%	28	28.60%	17	23.60%
4	7	2.00%	2	2.90%	1	1.00%	2	2.00%	2	2.80%
Histologic grade											n.s.
1	52	14.50%	8	11.10%	13	11.70%	22	21.80%	9	12.20%
2	169	47.20%	45	62.50%	45	40.50%	45	44.60%	34	45.90%
3	124	34.60%	18	25.00%	47	42.30%	29	28.70%	30	40.50%
4	13	3.60%	1	1.40%	6	5.40%	5	5.00%	1	1.40%