Measuring mammography compliance: lessons learned from a survival analysis of screening behavior.

Data from a telephone survey of women participating in a federally funded screening program were used to demonstrate the sensitivity of mammography compliance estimates to varying definitions of the time interval within which women are considered compliant with screening guidelines and what constitutes a true screening (as opposed to diagnostic) mammogram. The survival analysis approach used reveals patterns concealed by other approaches to measuring mammography behavior and provides a means for quantifying the impact of various definitional choices on compliance estimates. The results suggest that, although variations in defining and excluding potential diagnostic mammograms lead to differences in compliance measures no greater than 6%, differences as small as 1 month in the screening interval definition used can produce differences in compliance estimates as large as 27%. These results call into question the comparability of estimates across studies and suggest that standard measures would greatly facilitate future efforts in understanding how to promote compliance with mammography screening guidelines.

A randomized community trial to increase mammography utilization among low-income women living in public housing.

BACKGROUND: A randomized trial was conducted to evaluate the impact of a community-based intervention on mammography use among low-income women living in public housing. METHODS: All 41 public housing high-rise buildings were randomized to treatment and delayed treatment (control) conditions. After a cross-sectional baseline survey, an intervention called Friend to Friend was conducted in the treatment buildings by American Cancer Society and building resident volunteers. The intervention consisted of a health professional talk, small group discussions, and an opportunity to request assistance in obtaining a mammogram or mammogram reminder. A second cross-sectional survey was conducted to measure differences in screening rates between the study groups. RESULTS: Participation in the intervention averaged 27%. The study groups were equivalent at baseline. At follow-up, the proportion of women age 50-79 years who reported mammography screening in the previous 15 months was significantly higher in the treatment group (64%) than in the control group (52%). Breast cancer knowledge, attitudes, and beliefs did not differ between groups. CONCLUSIONS: These findings suggest that a multidimensional intervention which reaches women within their social environment and uses community volunteers can increase mammography utilization among women in public housing.