Table 2.

Impact of guideline initiation age on cancer screening - adjusted odds ratio of being at or above the guideline recommended initiation age

Dependent variable breast: Received mammogram in the past 2 years Dependent variable colorectal: Received FOBT test, sigmoidoscopy or colonoscopy in the past 2 years Dependent variable prostate: Received PSA test in the past 2 years
	Adjusted Odds Ratio (95% Confidence Interval)
	Breast	Colorectal	Prostate
Atlantic	1.70	0.92	1.22
	(0.86, 3.35)	(0.45, 1.89)	(0.50, 2.96)
Quebec	2.19**
	(1.29, 3.72)
Ontario	1.32	1.12	1.16
	(0.85, 2.05)	(0.74, 1.69)	(0.63, 2.12)
Saskatchewan		3.62*
		(1.09, 12.05)
Prairies	1.31
	(0.71, 2.42)
British Columbia	1.21	0.99	1.01
	(0.60, 2.43)	(0.52, 1.89)	(0.48, 2.12)

*** p <= .001, **p <= .01, * p <= .05. Notes: Each coefficient comes from a separate logistic regression which includes controls for age, age² and age³; race (white/non-white), marital status (5 categories), education (4 categories), household income quartiles, and sex (for colorectal).

Marginal effects for the age cutoff variable (breast): Atlantic 12.95, Quebec 19.34, Ontario 6.87, Prairies 6.76, British Columbia 4.69

Marginal effects for the age cutoff variable (colorectal): Atlantic -0.82, Ontario 1.48, Saskatchewan 12.51, BC -0.11

Marginal effects for the age cutoff variable (prostate): Atlantic 4.37, Ontario 3.14, British Columbia 0.21

Sample sizes (breast, women ages 40-60): Atl N = 2,745, QC N = 4,885, ON N = 6,341, Prairies N = 3,849, BC N = 2,324

Sample sizes (colorectal, adults ages 40-60): Atl N = 5,005, ON N = 12,112, SK N = 1,149, BC N = 4,722

Sample sizes (prostate, men ages 40-60): Atl N = 1,533, ON N = 3,353, BC N = 2,230

Note: Additional file 1, available in an attached document, presents the full set of regression coefficients for each of these models.