Breast cancer screening: evidence of benefit depends on the method used

Table 1 Relative risk of breast cancer death before screening start in women likely and not likely to participate in screening, and relative risk of breast cancer death after a period of screening with 75% participation.

		Before screening	After years of screening			Correction according to Duffy *et al.*(2002) [18]
		BC death rate^a	Change in mortality due to screening	Change in mortality unrelated to screening	BC death rate^a	Dr	A = p × RR × Dr	B = 1-[(1-p) × Dr]	Corrected RR (= A/B)
Panel A	Participants	86	0%	0%	86
	Non-participants	142	0%	0%	142
	All women^b	100	0%	0%	100
	RR breast cancer death, participants versus non-participants	0.61			0.61	1.42	0.65	0.65	1.00
Panel B	Participants	86	-20%	0%	69
	Non-participants	142	0%	0%	142
	All women^b	100	-15%	0%	85
	RR breast cancer death, participants versus non-participants	0.61			0.48	1.42	0.52	0.65	0.80
Panel C	Participants	86	0%	-25%	65
	Non-participants	142	0%	-25%	107
	All women^b	100	0%	-25%	75
	RR breast cancer death, participants versus non-participants	0.61			0.61	1.07	0.48	0.73	0.66
Panel D	Participants	86	-20%	-25%	47
	Non-participants	142	0%	-25%	107
	All women^b	100	-15%	-25%	60
	RR breast cancer death, participants versus non-participants	0.61			0.44	1.07	0.35	0.73	0.48

BC: Breast cancer; Dr: rate in non-participants after screening divided by the rate in all women before screening; p: participation rate of 75%; RR: relative risk. ^aPer 100,000 women-year. ^b(rate in participants × 0.75) + (rate in non-participants × 0.25).

ISSN: 1741-7015