Forecasting the cancer population in the UK to 2040


Session type:

Samuel Jones1,Hannah McConnell1,James Charnock2,Martina Buoni1
1Macmillan Cancer Support,2Macmillan Cancer Support/National Cancer Registration and Analysis Service (NCRAS), Public Health England



Previous projections have suggested there were 2.5 million people living in the UK with a cancer diagnosis in 2015 and that this could increase to 4 million by 2030.

We aim to update this using the latest outputs from the Macmillan-National Cancer Registration and Analysis Service (NCRAS) UK Cancer Prevalence Project.  and the most recently available public data, to build a model projecting cancer prevalence for the UK up to 2040. We will also explore the feasibility of taking into account specific scenarios which may influence how the cancer population could change in the future.


Year on year cancer prevalence is calculated using aggregate and publicly available incidence and mortality figures for England, Wales, Scotland and Northern Ireland.  Inputs will be split according to gender, age group and cancer type (all cancers, breast cancer, colorectal cancer, lung cancer and prostate cancer). Assumptions around changing diagnosis rates, patterns of staging and survival are applied to each year and lifestyle risk factors which affect incidence or survival are incorporated based on academic research.  By working with site-specific clinical experts we also ensure the most likely scenarios are applied.


Outputs of the model are for the UK cancer population up to 2040. The model will provide estimated prevalence at each phase of the cancer pathway. Outputs will be reported for the UK – and for each nation including a sub-national breakdown - split by age group and gender.  Results will be presented at the conference.


This research allows us to capture and segment the cancer population and assess how this may change over time according to likely scenarios and changing factors. This will help the development of new cancer services for the future. Crucially, this provides us with a model which we are able to update accordingly as to when new baseline data emerge.