Table 3. Cox regression analysis for repeat PSA test (N = 555, 160 with repeat PSA test). Uni- and multi-variate Cox regression, using genetic risk score, awareness of having a genetic test, PSA level at inclusion, perceived risk of PC, perceived risk of dying from PC, and intention for a repeat PSA were investigated as potential explanatory variables for actually having a repeat PSA test within 2 years.

	Univariate	Multivariate
Variable	Characteristics	HR (95% CI)	P value	C-index ^a	HR (95% CI)	P value
Genetic risk	Normal versus high	8.11(5.83 to 11.29)	< 0.001	0.66	5.99(4.09 to 8.79)	< 0.001
Awareness of having a genetic test	No versus yes	1.23(0.90 to 1.69)	0.193	0.53	1.32(0.95 to 1.83)	0.099
PSA at inclusion	Continuous	1.24(1.05 to 1.46)	0.0097	0.56	1.16(0.98 to 1.38)	0.083
Perceived risk for PC	1–5	2.13(1.68 to 2.71)	< 0.001	0.62	1.2(0.83 to 1.74)	0.331
Perceived risk for dying of PC	1–5	1.77(1.38 to 2.28)	< 0.001	0.59	1.11(0.79 to 1.57)	0.553
Intention for repeat PSA test	1–4	1.35(1.18 to 1.54)	< 0.001	0.62	1.2(1.05 to 1.37)	0.007

^aC-index, Harrell’s concordance index. HR = hazard ratio. PC = prostate cancer. PSA = prostate specific antigen.