Geographical disparities, trends and projections in female gender-related cancers in Switzerland

Gender-related cancers have a deep impact on patients and on public health. The psychological and social consequences are further pronounced through their direct link with sexuality, reproduction and gender identity. This thesis focusses on the female cancers. Female gender cancers account for about 42% of all newly diagnosed cancer cases and 29% of all cancer-related deaths in women in Switzerland. Among female gender-related cancers, breast cancer is the predominant one accounting for more than three-quarters of new cases in this group.
This thesis aims to apply and further develop rigorous analytical methods on cancer data collected from cancer registries and the Federal Statistical Office (FSO) in order to quantify the burden of breast and other female gender-related cancers in Switzerland, its spatial disparities and future trends and to investigate the geographical patterns of case management procedures in the country. The results are particularly important for public health policy in Switzerland and in advancing epidemiology for prevention and care. This thesis is structured by addressing broader to increasingly more specific research questions.
Chapter 2 gives an overview of the burden of cancer for the society and for the health system in Switzerland, by estimating the prevalence of cancer, resp. the number cancer survivors by gender, time since diagnosis, cancer type and age group. Cancer survivors are a heterogeneous group with complex health problems. Data concerning its total number and growing dynamics for Switzerland were scarce and outdated. In this chapter back calculation methods (PIAMOD) were applied on nationwide mortality data and relative survival data from Swiss cancer registries, and trends estimated by Joinpoint analysis.
The analysis shows that there is a rapidly growing population of cancer survivors in Switzerland whose needs and concerns are largely unknown. The number of breast cancer survivors more than doubled from 1990 to 2010 and is projected to increase by another 28% from 2010 to 2020. Overall, the number of cancer survivors increased from 2.08% of the population in 1990 to 3.7% in 2010, with the current annual growth rate being 3.3% per year.
In Chapter 3, spatio-temporal disparities in female gender cancer mortality are investigated. Differences in the decrease and in spatial patterns within Switzerland have been reported according to urbanisation and language region and remained controversial.
Using modern Bayesian small area modelling and mapping techniques it was possible to show that all investigated groups of women in Switzerland have benefited from progress in cancer control regardless of place of residence in the past 40 years. Only small differences in the geographical variation of mortality are observed, present on a regional or canton-overspanning level, and different for each cancer site and age group. No general significant association with cantonal or language region borders could be observed.
Chapter 4 provides a more detailed analysis of breast cancer mortality disparities in Switzerland, with a focus on mammography screening programmes and socio-economic differences. In Switzerland, programmes have existed in some regions for over 20 years, while they do not yet exist in others, thus offering the possibility to analyse its effects with modern spatio-temporal methodology. The methodology of Chapter 3 was adapted to measure additional effects of screening and socio-economic disparities. Breast cancer mortality has strongly declined since the 1990s. No important spatial disparities are observed. The analysis shows that moderate geographical differences found are within credible intervals using modern Bayesian techniques. The duration of population-based screening programme and socioeconomic characteristics have no significant influence on breast cancer mortality and have been outweighed by important advances in treatment approaches.
In Chapter 5 breast cancer mortality is projected by canton for 2015-2024. Projections of cancer mortality provide an estimate of the future burden of cancer and are important to assess the impact of novel approaches and future developments in breast cancer care, as well as public health interventions. Projected mortality numbers can support healthcare planning. However, this has not been done for Switzerland so far.
Breast cancer mortality in Switzerland is projected to further decrease within the next ten years. However, the pace is regionally quite different. In more than half of the cantons, the number of breast cancer deaths is projected to increase with the demographic forecast playing a major role.
In Chapter 6 geographical variation of hospital-based case management is assessed by measuring spatio-temporal differences of mastectomy rates by age group and hospital region. The FSO has collected hospital discharge records routinely since 1998, with information on diagnoses, co-morbidities, and treatments. This data with national coverage has not been used to describe disparities in breast cancer care patterns. Data on mammography screening programme duration and surgeon and gynaecologist density were also included.
The analysis shows an important decline of mastectomy rates in Switzerland in 2000-2012 for patients aged 50-69 and 70+, but no change for those under 50. Important geographical differences in rate estimates are present and rates are significantly influenced by age and co-morbidities of patients. Regions with higher surgeon and gynaecologist density have higher rates of mastectomies and regions with mammography screening programmes lower rates. A better insight into the differences in cancer management procedures will be useful for planning future demand and resource allocation for early detection, diagnosis, and treatment.
The main contributions of this thesis are (1) building awareness towards cancer survivors in Switzerland directly prompting several measures from policy to care, (2) identifying areas with high cancer rates and discrepancies of disease burden among areas in order to help health authorities in planning and evaluating cancer control activities (3) statistical methods for studying spatio-temporal patterns and projections of cancer mortality driven by data availability and characteristics in Switzerland that were applied and validated on real datasets, (4) smoothed maps of age-specific patterns of breast cancer mortality over time giving a better insight into the differences in cancer mortality rates between linguistic regions, urbanisation, affluence and cancer management procedures, (5) understanding the effect of differences in the uptake of screening programmes and socio-economic status on breast cancer mortality in the population, (6) estimates of the geographical patterns of breast cancer mortality for the next 10 years, (7) identifying regions in which special attention is required to reduce healthcare inequalities and their impact on the community health, (8) determining the influencing factors for high mastectomy rates and (9) providing information for public health authorities for planning future demands and resource allocation for diagnosis and treatment.