Screening: the information individuals need to support their decision: per protocol analysis is better than intention-to-treat analysis at quantifying potential benefits and harms of screening

Table 1 Comparison of three plain language syntheses or fact boxes summarising benefits and harms of mammographic screening for breast cancer as reported by the authors (intention-to-treat, ITT) and re-computed according to an individual decision making perspective (per protocol, PP)

	UK independent panel[4] 10,000 women aged 50 screened every three years for 20 years			Euroscreen group[5] 1,000 women aged 50 screened every two years for 20 years and followed up for 30 years			Harding Centre[6], Based on Nordic Cochrane review[13] 2,000 women >50 yrs screened for 10 years
Population time frame
	ITT		PP	ITT		PP	ITT		PP
	Invited	Not invited	Screened	Invited	Not invited	Screened	Invited	Not invited	Screened*
Cancer specific deaths	-	-	-	21-23	30	21-23	7	8	6.6
Prevented deaths	43	-	56	7-9	-	12-14	1		1.4
Cancer incidence	681	552	720	71	67	73
Additional cancer	129	-	168	4	-	6	10		14.2
Ascertainments				100		143
Non invasive				70		100
Invasive				30		43
Psychological distress							200		286

*For the Harding Centre fact sheet the per protocol figures were estimated by applying the average participation rate used by the Euroscreen group [4].

ISSN: 1472-6939