Predictors of non-attendance for second round mammography in an Australianmammographic screening programme

Objective-To determine the socioeconomic, cultural, and clinical predictors of non-attendance for second round mammography. Design/participants-Retrospective cohort study of 121 889 women aged 50-69 years who attended for first mammography screening in the BreastScreen Vict oria programme in 1995/1996 and who were recommended to be invited for rout ine biennial mammography. Women were considered to be non-attenders if they had not attended for rescreening within 27 months of their initial screeni ng. Relative risk (RR) was used to compare categories for non-attendance fo r second screening, and a multivariate model was fitted to adjust for possi ble confounding. Setting-BreastScreen Victoria, a population based mammographic screening pr ogramme, which offers free biennial mammography to all women 40 years and o lder. The programme specifically targets women aged 50-69 years. Results-In the multivariate analysis, women from non-English speaking backg rounds were more likely not to attend for second round screening (RR ranged from 1.18 to 1.77). indigenous women (RR 2.02, 95% confidence interval (CI ) 1.61 to 2.54) and women who reported either significant symptoms (RR 1.90 , 95% CI 1.76 to 2.05) or other breast symptoms (RR 2.25, 95% CI 2.15 to 2. 36) at the time of first round screening were also more likely nor to atten d for second round screening. Conclusions-Women from non-English speaking backgrounds, indigenous women, and women who report symptoms at the time of first screening are more likel y to not attend for second round screening. it is important to investigate why these women do not attend for second round screening so that services c an be more appropriately tailored to their needs.