How consistently do women report lifetime mammograms at successive interviews?

BACKGROUND: Little data exist on the reliability of self-reported regular mammography use measures. We used data from two successive interviews of 892 women aged 50 to 74 years without a history of abnormal mammograms to investigate how consistently women report their lifetime number of mammograms.
METHODS: We added an estimated number of mammograms obtained between interviews to the baseline report to create a revised baseline report for comparison with the follow-up report. We then examined the correlation in paired reports, the level of agreement between paired reports, and factors associated with consistent reporting.
RESULTS: Spearman rank correlation between paired reports was 0.73. Agreement between paired reports dropped with increasing lifetime number of mammograms. After adjustment for mammography use, women's characteristics did not appear to be strongly associated with consistent reporting.
CONCLUSIONS: Self-reported lifetime number of mammograms is a reasonably consistent measure for younger women or women with less mammography experience, but it is less reliable for women with long mammography histories. In these women, it may be useful to distinguish those who obtain regular screening from those who do not. Assessing reliability as well as validity for other measures of regular mammography use will allow additional measures to be identified.