Natural history estimation for screening evaluation and planning
Centre of Cancer Prevention
This work complements our research evaluating service screening. Estimates of the parameters of cancer progression can help to resolve important issues in cancer screening and prevention.
- How the beneficial and harmful effects of lung cancer screening depend on screening frequency;
- Individually tailored breast screening intervals based on age, mammographic density, and other risk factors;
- Risk-stratified endoscopy intervals in patients who have had adenomas removed;
- Age to begin, interval and age to cease cervical screening taking into account individual level HPV vaccination and introduction of HPV testing;
We will develop simple natural history models of progression of premalignancy to cancer and of subsequent local and distant spread, taking into account potential complexities including: biological heterogeneity (of cancers and of premalignant lesions); host- and lesion-specific covariates influencing progression rates; and factors affecting sensitivity of surveillance.
Our starting point will be Markov processes. Thereafter, to account for these complexities, it will be necessary to include covariates and relax the strong homogeneity assumptions using frailty methods and other approaches.
Phone: +44 (0) 20 7882 3535