Dr B Taylor, Dr T Keegan
No more applications being accepted
Competition Funded PhD Project (European/UK Students Only)
About the Project
With over 300,000 new cases diagnosed per year in the UK, and with a 50% lifetime risk of developing the disease, cancer is one of the biggest heath challenges faced by the UK population and those of other developed countries (CRUK, 2016a). Understanding the changing profile of cancer incidence and survival over time is crucial for monitoring the success of treatment advances and also planning healthcare and infrastructure provision for our ageing population (CRUK, 2016b). This is pertinent as cancer survival prognoses in the UK have recently been shown to lag behind those in other developed economies, such as Italy’s (Allemani et al., 2015; Guardian, 2015).
This project will develop statistical models, methods and visualisation tools for analysing UK cancer registrations (approx. 7 million records) at a fine spatial-scale (approx.. 70,000 spatial units) over a 20-year period. A central goal will be to improve the presentation of UK cancer statistics: rather than commonly presented summaries such as raw, or age-adjusted rates at a coarse ecological level, the student will produce model-based summaries of incidence, survival and risk at the Lower Super Output Area (LSOA) level.
Data for this project will be obtained from the National Cancer Data Repository (NCDR).The NCDR is a gold-standard resource for cancer intelligence in the UK: with 20 years of data between 1990 and 2010, it maintains individual-level information on all cancers registered in the UK and is spatially resolved to the LSOA level (NCIN, 2016).
The aims of this project are:
• To develop statistical methodology for the analysis and presentation of large-scale spatio-temporal survival datasets.
• To analyse the NCDR data by major primary cancer site (e.g. lung) and quantify fine-scale spatio-temoral variations in survival prognosis/incidence.
• Having adjusted for individual, socio-economic and environmental risk factors, to identify (small-scale) areas of the United Kingdom where cancer outcomes (survival/incidence) are worse or better than expected.
• To forecast incidence and survival rates for cancer over the next 10 years at the LSOA level and with national coverage.
• To develop state-of-the-art visualisation tools to allow epidemiologists to interact with model outputs in a useful and easily interpretable way.
The student will work with statisticians and epidemiologists; be encouraged to develop relationships with organisations such as Public Health England, and will be encouraged to attend relevant training courses and international conferences.
Funding Notes
Awards are available for UK or EU students only for a maximum of three years full-time study. Awards will cover University Fees and Doctoral Stipend (2017-2018: £14,553).
References
Allemani, Claudia et al. (2015). Global surveillance of cancer survival 1995–2009:analysis of individual data for 25 676 887 patients from 279 population-based registries in 67 countries (CONCORD-2) The Lancet , Volume 385 , Issue 9972 , 977 – 1010
CRUK (2016a) Cancer Statistics for the UK http://www.cancerresearchuk.org/health-professional/cancer-statistics
CRUK (2016b) Cancer Incidence by Age http://www.cancerresearchuk.org/health-professional/cancer-statistics/incidence/age
Guardian (2015) UK cancer survival rates trail 10 years behind other European countrieshttp://www.theguardian.com/society/2015/mar/24/uk-cancer-survival-rates-trail-10-years-behind-those-in-european-countries
NCIN (2016). National Cancer Data Repository. http://www.ncin.org.uk/collecting_and_using_data/national_cancer_data_repository/
Taylor (2015) Auxiliary Variable Markov Chain Monte Carlo for Spatial Survival and Geostatistical Models. Benjamin M. Taylor. Submitted.
Taylor and Rowlingson (2016a) spatsurv: an R Package for Bayesian Inference with Spatial Survival Models. Benjamin M. Taylor and Barry Rowlingson. To appear in Journal ofStatistical Software.
Taylor (2017) Spatial Modelling of Emergency Service Response Times. Benjamin M. Taylor. To appear in Journal of the Royal Statistical Society Series A.