Age-period-cohort (APC) models are the state of art in cancer projections, assessing past and recent trends and extrapolating mortality or incidence data into the future. Nordpred is a well-established software, assuming a Poisson distribution for the counts and a log-link or power-link function with fixed power; however, its predictive performance is poor for sparse data. Bayesian models with log-link function have been applied, but they can lead to extreme estimates. In this paper, we address criticisms of the aforementioned models by providing Bayesian formulations based on a power-link and develop a generalized APC power-link model, which assumes a random rather than fixed power parameter. In addition, a power model with a fixed power parameter of five was formulated in the Bayesian framework. The predictive performance of the new models was evaluated on Swiss lung cancer mortality data using model-based estimates of observed periods. Results indicated that the generalized APC power-link model provides best estimates for male and female lung cancer mortality. The gender-specific models were further applied to project lung cancer mortality in Switzerland during the periods 2009-2013 and 2014-2018.

Background: Prostate cancer (PCa) is the most frequent cancer and the third leading cause of cancer death among German men. One option for PCa early detection is prostate-specific antigen (PSA) testing, which is still under debate regarding its risk benefits. Besides recommendations on the early PCa detection, daily practice on PSA testing varies in, for example, information communication and usage of the test. This pilot study assessed potential differences between general practitioners (GPs) and urologists in handling PSA testing and guidelines on early detection of PCa. Methods: 172 GPs belonging to the teaching network of the University of Oldenburg in Lower Saxony and Bremen and 128 practicing urologists were included in the online survey focusing on PSA testing. The questionnaire covered 43 questions on topics as the usage of the test, information communication, handling of test results and handling of/knowledge about national and international guidelines on PCa. Wether PSA testing is used in accordance with guidelines was also explored in four standardized case scenarios. Statistical analysis was done at a descriptive level. Results: In total, 65 doctors participated in the survey (response proportion: 21.7%, n=65; 27.9%, n=48 [GPs]; 13.2%, n=17 [urologists]). Results of 41 GPs and 14 urologists were analyzed. The PSA test was judged as useful by all urologists, while almost half of the GPs valued the test as ambivalent or not useful. Urologists showed a more proactive approach of informing men on PSA testing. Regarding guidelines and recommendations on PSA testing, GPs were less familiar with them compared to the urologists. Doctors of both specialties did not always treat men in consistence with the guidelines. This was partially in contradiction to their self-appraisal. Conclusion: This pilot study is highlighting differences in PSA testing practices between GPs and urologists in Germany. Urologists showed a more proactive approach. For further verification, we plan a more comprehensive study covering several German states.

Abstract. Tobacco smoking is a main cause of disease in Switzerland; lung cancer being the most common cancer mortality in men and the second most common in women. Although disease-specific mortality is decreasing in men, it is steadily increasing in women. The four language regions in this country might play a role in this context as they are influenced in different ways by the cultural and social behaviour of neighbouring countries. Bayesian hierarchical spatio-temporal, negative binomial models were fitted on subgroup-specific death rates indirectly standardized by national references to explore age-and gender-specific spatio-temporal patterns of mortality due to lung cancer and other tobacco-related cancers in Switzerland for the time period 1969-2002. Differences influenced by linguistic region and life in rural or urban areas were also accounted for. Male lung cancer mortality was found to be rather homogeneous in space, whereas women were confirmed to be more affected in urban regions. Compared to the German-speaking part, female mortality was higher in the French-speaking part of the country, a result contradicting other reports of similar comparisons between France and Germany. The spatio-temporal patterns of mortality were similar for lung cancer and other tobacco-related cancers. The estimated mortality maps can support the planning in health care services and evaluation of a national tobacco control programme. Better understanding of spatial and temporal variation of cancer of the lung and other tobacco-related cancers may help in allocating resources for more effective screening, diagnosis and therapy. The methodology can be applied to similar studies in other settings.