Table 3.

Validation of metastases algorithm comparison of hazard ratio (HR) of metastasis for primary androgen deprivation (PADT) versus surveillance using all claims versus hospital claims.

Cancer risk	PADT events/person-year	Rate per 100	Surveillance events/person-year	Rate per 100	Unadjusted HR (95% CI)	Adjusted HR (95% CI)
Conventional cox multivariate results Metastasis (use all claims)*
Low risk	1147/40804	2.8	1627/100346	1.6	1.77 (1.64–1.91)	1.62 (1.49–1.75)
High risk	1163/21546	5.4	457/12888	3.5	1.53 (1.37–1.70)	1.65 (1.47–1.85)
All risk	2310/62351	3.7	2084/113234	1.8	2.05 (1.93–2.17)	1.66 (1.55–1.77)
Metastasis (use hospital claims only)*
Low risk	309/43083	0.7	428/103651	0.4	1.79 (1.55–2.08)	1.65 (1.41–1.92)
High risk	373/22752	1.6	160/13465	1.2	1.39 (1.16–1.67)	1.55 (1.27–1.88)
All risk	682/65835	1.0	588/117116	0.5	2.13 (1.91–2.38)	1.64 (1.46–1.85)

^*ICD-9 198.5 or 198.3 or 197.0 or 197.7.