Factors affecting attendance at population-based mammography screening

The aim of the studies described in this thesis was to explore factors affecting attendance at outreach mammography screening, and to assess possible associations between socioeconomic factors and breast cancer survival. The relation of non-attendance at mammography screening to sociodemographic factors, indicators of general health behavior, self-rated health, experience of cancer in others, and own cancer or breast problems was studied in a case-control study with 434 non-attenders and 515 attenders identified in a population-based mammography register in the county of Uppsala (Paper I). Exposure information was obtained through telephone interviews. Being single and being non- employed were the only important sociodemographic predictors of nonattendance. Non-attendance was more likely among women who smoked regularly, had not visited a doctor in 5 years, had never had cervical smear tests, had never used oral contraceptives or hormone replacement therapy, never drank alcohol, and who reported no breast cancer in the family or among friends or history of own breast problems. In the same case-control study the effects of attitudes, beliefs and knowledge were evaluated (Paper 11). Multivariate analysis showed that non-attendance was almost five times higher among women with the highest scores of perceived emotional barriers compared to women with the lowest scores. Both absence of worry about breast cancer and low scores of perceived benefits were associated with non-attendance. Other factors associated with nonattendance were less knowledge about mammography and breast cancer, lack of advice from a health professional to participate, and very poor trust in health care. To obtain descriptive insights into women's rationales for attending or not attending mammography screening, eight focus group discussions were conducted (Paper Ill). Through inductive analysis, six main issues were identified: negative experiences, perceived risk factors, knowledge of one's own body, perceived problems with mammography, political, ideological and moral reasoning, and involuntary non-attendance. The association between sociodemographic factors and non-attendance at mammography screening was further investigated in a record-linkage study with data from the mammography register in Uppsala, the Census of 1990 and the Fertility Register (Paper IV). Multivariate analyses among 4,198 non-attenders and 38,972 attenders confirmed that non-attendance was more common among women living without a partner, and among those who were not employed. Non- attendance was also more frequent among both childless and high-parity women, among those renting an apartment (rather than owning their apartment or house), and among immigrants from non-Nordic countries. The association between socioeconomic factors and survival among 4,645 women with a first diagnosis of breast cancer in 1993 was assessed in a register study (Paper V). Sociodemographic data were obtained from the Census of 1990 and the Fertility Register. Stage of disease at diagnosis was obtained from Regional Cancer Registers. Follow-up through 1998 in the Swedish Cause of Death Register revealed that 772 of the 4,645 women had died of breast cancer. Five-year survival was 59% higher among women of higher socioeconomic status, after adjustment for stage of disease and age