Smoothed Temporal Atlases of Age-Gender All-Cause Mortality in South Africa

Cancer mortality distribution in South Africa, 1997-2016

Introduction

The mortality data in South Africa (SA) have not been widely used to estimate the patterns of deaths attributed to cancer over a spectrum of relevant subgroups. There is no research in SA providing patterns and atlases of cancer deaths in age and sex groups per district per year. This study presents age-sex-specific geographical patterns of cancer mortality at the district level in SA and their temporal evolutions from 1997 to 2016.

Methods

Individual mortality level data provided by Statistics South Africa were grouped by three age groups (0-14, 15-64, and 65+), sex (male and female), and aggregated at each of the 52 districts. The proportionate mortality ratios (PMRs) for cancer were calculated per 100 residents. The atlases showing the distribution of cancer mortality were plotted using ArcGIS. Spatial analyses were conducted through Moran's I test.

Results

There was an increase in PMRs for cancer in the age groups 15-64 and 65+ years from 2006 to 2016. Ranges were 2.83 (95% CI: 2.77-2.89) -4.16 (95% CI: 4.08-4.24) among men aged 15-64 years and 2.99 (95% CI: 2.93-3.06) -5.19 (95% CI: 5.09-5.28) among women in this age group. The PMRs in men and women aged 65+ years were 2.47 (95% CI: 2.42-2.53) -4.06 (95% CI: 3.98-4.14), and 2.33 (95% CI: 2.27-2.38) -4.19 (95% CI: 4.11-4.28). There were considerable geographical variations and similarities in the patterns of cancer mortality. For the age group 15-64 years, the ranges were 1.18 (95% CI: 0.78-1.71) -8.71 (95% CI: 7.18-10.47), p < 0.0001 in men and 1.35 (95% CI: 0.92-1.92) -10.83 (95% CI: 8.84-13.14), p < 0.0001 in women in 2016. There were higher PMRs among women in the Western Cape, Northern Cape, North West, and Gauteng compared to other areas. Similar patterns were also observed among men in these provinces, except in North West and Gauteng.

Conclusion

The identification of geographical and temporal distributions of cancer mortality provided evidence of periods and districts with similar and divergent patterns. This will contribute to understanding the past, present, future trends and formulating interventions at a local level.

Metropolitan age-specific mortality trends at borough and neighborhood level: The case of Mexico City

Understanding the spatial and temporal patterns of mortality rates in a highly heterogeneous metropolis, is a matter of public policy interest. In this context, there is no, to the best of our knowledge, previous studies that correlate both spatio-temporal and age-specific mortality rates in Mexico City. Spatio-temporal Kriging modeling was used over five age-specific mortality rates (from the years 2000 to 2016 in Mexico City), to gain both spatial (borough and neighborhood) and temporal (year and trimester) data level description. Mortality age-specific patterns have been modeled using multilevel modeling for longitudinal data. Posterior tests were carried out to compare mortality averages between geo-spatial locations. Mortality correlation extends in all study groups for as long as 12 years and as far as 13.27 km. The highest mortality rate takes place in the Cuauhtémoc borough, the commercial, touristic and cultural core downtown of Mexico City. On the contrary, Tlalpan borough is the one with the lowest mortality rates in all the study groups. Post-productive mortality is the first age-specific cause of death, followed by infant, productive, pre-school and scholar groups. The combinations of spatio-temporal Kriging estimation and time-evolution linear mixed-effect models, allowed us to unveil relevant time and location trends that may be useful for public policy planning in Mexico City.

Multivariate Bayesian meta-analysis: joint modelling of multiple cancer types using summary statistics

BACKGROUND

Cancer atlases often provide estimates of cancer incidence, mortality or survival across small areas of a region or country. A recent example of a cancer atlas is the Australian cancer atlas (ACA), that provides interactive maps to visualise spatially smoothed estimates of cancer incidence and survival for 20 different cancer types over 2148 small areas across Australia.

METHODS

The present study proposes a multivariate Bayesian meta-analysis model, which can model multiple cancers jointly using summary measures without requiring access to the unit record data. This new approach is illustrated by modelling the publicly available spatially smoothed standardised incidence ratios for multiple cancers in the ACA divided into three groups: common, rare/less common and smoking-related. The multivariate Bayesian meta-analysis models are fitted to each group in order to explore any possible association between the cancers in three remoteness regions: major cities, regional and remote areas across Australia. The correlation between the pairs of cancers included in each multivariate model for a group was examined by computing the posterior correlation matrix for each cancer group in each region. The posterior correlation matrices in different remoteness regions were compared using Jennrich's test of equality of correlation matrices (Jennrich in J Am Stat Assoc. 1970;65(330):904-12. https://doi.org/10.1080/01621459.1970.10481133 ).

RESULTS

Substantive correlation was observed among some cancer types. There was evidence that the magnitude of this correlation varied according to remoteness of a region. For example, there has been significant negative correlation between prostate and lung cancer in major cities, but zero correlation found in regional and remote areas for the same pair of cancer types. High risk areas for specific combinations of cancer types were identified and visualised from the proposed model.

CONCLUSIONS

Publicly available spatially smoothed disease estimates can be used to explore additional research questions by modelling multiple cancer types jointly. These proposed multivariate meta-analysis models could be useful when unit record data are unavailable because of privacy and confidentiality requirements.

Multivariate Bayesian meta-analysis: joint modelling of multiple cancer types using summary statistics

BACKGROUND

Cancer atlases often provide estimates of cancer incidence, mortality or survival across small areas of a region or country. A recent example of a cancer atlas is the Australian cancer atlas (ACA), that provides interactive maps to visualise spatially smoothed estimates of cancer incidence and survival for 20 different cancer types over 2148 small areas across Australia.

METHODS

The present study proposes a multivariate Bayesian meta-analysis model, which can model multiple cancers jointly using summary measures without requiring access to the unit record data. This new approach is illustrated by modelling the publicly available spatially smoothed standardised incidence ratios for multiple cancers in the ACA divided into three groups: common, rare/less common and smoking-related. The multivariate Bayesian meta-analysis models are fitted to each group in order to explore any possible association between the cancers in three remoteness regions: major cities, regional and remote areas across Australia. The correlation between the pairs of cancers included in each multivariate model for a group was examined by computing the posterior correlation matrix for each cancer group in each region. The posterior correlation matrices in different remoteness regions were compared using Jennrich's test of equality of correlation matrices (Jennrich in J Am Stat Assoc. 1970;65(330):904-12. https://doi.org/10.1080/01621459.1970.10481133 ).

RESULTS

Substantive correlation was observed among some cancer types. There was evidence that the magnitude of this correlation varied according to remoteness of a region. For example, there has been significant negative correlation between prostate and lung cancer in major cities, but zero correlation found in regional and remote areas for the same pair of cancer types. High risk areas for specific combinations of cancer types were identified and visualised from the proposed model.

CONCLUSIONS

Publicly available spatially smoothed disease estimates can be used to explore additional research questions by modelling multiple cancer types jointly. These proposed multivariate meta-analysis models could be useful when unit record data are unavailable because of privacy and confidentiality requirements.

Modeling and mapping the burden of disease in Kenya

Precision public health approaches are crucial for targeting health policies to regions most affected by disease. We present the first sub-national and spatially explicit burden of disease study in Africa. We used a cross-sectional study design and assessed data from the Kenya population and housing census of 2009 for calculating YLLs (years of life lost) due to premature mortality at the division level (N = 612). We conducted spatial autocorrelation analysis to identify spatial clusters of YLLs and applied boosted regression trees to find statistical associations between locational risk factors and YLLs. We found statistically significant spatial clusters of high numbers of YLLs at the division level in western, northwestern, and northeastern areas of Kenya. Ethnicity and household crowding were the most important and significant risk factors for YLL. Further positive and significantly associated variables were malaria endemicity, northern geographic location, and higher YLL in neighboring divisions. In contrast, higher rates of married people and more precipitation in a division were significantly associated with less YLL. We provide an evidence base and a transferable approach that can guide health policy and intervention in sub-national regions afflicted by disease burden in Kenya and other areas of comparable settings.