Different survival analysis methods for measuring long-term outcomes of Indigenous and non-Indigenous Australian cancer patients in the presence and absence of competing risks

Table 5 Regression analysis of time trend after diagnosis using cause-specific mortality (Cox proportional hazard regression), relative survival (Poisson regression), and competing risk analysis (Fine-Gray regression), all cancers combined^a, Australia NT, 1991–2009

	Relative survival	Cause-specific	Competing (due to cancers)	Competing (other death)
	HR (95% CI)^b	HR (95% CI)	SHR (95% CI)	SHR (95% CI)
Indigenous^c	1.99 (1.82–2.16)	2.01 (1.85–2.19)	1.94 (1.77–2.13)	2.49 (1.70–3.63)
Female vs male	0.86 (0.79–0.93)	0.84 (0.77–0.91)	0.87 (0.79–0.94)	0.74 (0.53–1.04)
Age at diagnosis^d
Non-Indigenous	1.03 (1.03–1.04)	1.03 (1.03–1.04)	1.03 (1.03–1.03)	1.07 (1.06–1.08)
Indigenous	1.02 (1.01–1.02)	1.02 (1.01–1.02)	1.02 (1.01–1.02)	1.02 (1.00–1.04)
Year of diagnosis	0.97 (0.96–0.97)	0.97 (0.96–0.97)	0.97 (0.96–0.97)	0.97 (0.95–1.00)

^aModel adjusted for cancer site (with colorectal cancer as the reference category for cancer site)
^b HR hazard ratio, SHR standard hazard ratio
^cApplies to the reference categories of the interaction terms (i.e. people of median age 55 years in 2009)
^dPer year of age

ISSN: 1478-7954