Statistical methods for colorectal cancer - v1.0
Research type
Research Study
Full title
Improving statistical methods for individualised predictions and understanding population variation in survival for colorectal cancer patients.
IRAS ID
235503
Contact name
Elisavet Syriopoulou
Contact email
Sponsor organisation
University of Leicester
Duration of Study in the UK
2 years, 6 months, 5 days
Research summary
This project aims to develop and apply statistical methods to report and understand variation and eventually address important clinical questions for colorectal cancer (CRC).
At the individual level (for individual patients), a detailed, up-to-date prognostic model for CRC will be developed, specifically tailored to English patients. In prognostic models, specific information about the patient, rather than subgroups of the population, are used. Prognostic models provide scores that enable the prediction of the risk of future events for a patient. For example, the developed model could predict the probability of being alive 10 years following diagnosis for a 57 year old male diagnosed with colon cancer, with advanced cancer stage and a tumour size of 4.5cm. The model will help clinicians and patients to understand patient risk and will aid in clinical decisions such as identifying the best available treatment.
At the population level (for the whole English population), the project will explore the characteristics that influence differences in prognosis among different groups of CRC patients. Understanding variation of cancer at the population level will help develop and improve policy and awareness campaigns to help eliminate the observed inequalities. For example, if differences in prognosis across deprivation groups are largely driven by differences in stage at diagnosis, then policies could be implemented to encourage earlier detection in the most vulnerable groups. This will be achieved through the development of methods known in statistics as mediation analysis.
To ensure the widest possible dissemination of the results a webtool will be developed with a view to improving the clinical and societal management of the cancer.
The study will not be collecting new data, but will use routine NHS data available by Public Health England. The data is anonymised.REC name
Wales REC 7
REC reference
18/WA/0093
Date of REC Opinion
7 Mar 2018
REC opinion
Favourable Opinion