Table 1. Quality rating of included studies.

	Mahoney (1986)	Tate et al (1989)	Rutgers et al (1991)	Snee (1996)	Hussain et al (1995)	Grunfeld et al (1996)	Lees et al (1997)	Jack et al (1998)	Churn and Kelly (2001)	Grogan et al (2002)	van der Sangen et al (2006)	Montgomery et al (2007a, 2007b)
Is the population under study defined (with inclusion and exclusion criteria)?	Yes	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes
Is the original cohort of patients from which those with relapse were drawn defined?	Yes	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	No	Yes
Were all those identified as having relapse analysed?	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	No
Is loss during follow-up specified?	No	No	No	No	No	Yes	No	Yes	No	Yes	No	No
Are the main prognostic factors defined (at least age of patient and stage of tumour)?	No	No	No	Yes	No	Yes	No	Yes	Yes	Yes	Yes (personal communication)	Yes
Is treatment of first tumour specified (including adjuvant)?	No	No	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes
Is mean or median follow-up greater than 5 years?	Not given	No	Not given	Yes	Yes	No	Not given	Yes	Yes	Yes	Yes	Yes
Is the follow-up schedule (including mammographic interval) specified?	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Were methods of diagnosis of relapse prospectively assessed?	Yes	Yes	No	No	No	Yes	No	No	No	No	No	No
Is all relapse, including axillary and new contralateral cancers, included?	Not given	Yes	No	Not given	Not given	Not given	Yes	Yes	No	Yes	No	Yes
Percentage of relapses not analysed due to inadequate information	0	0	2%	0	0	0	0	0	0	0	0	3%
Total score	5	6	4	6	6	8	3	9	7	9	6	7