In most countries, who should be screened for bowel cancer is determined by age and family history of cancer. Risk prediction models using genomic (DNA) and/or lifestyle factors to predict a person’s bowel cancer risk have been developed and could be used to split the population into risk categories. Theoretically, if these models were implemented within current population screening programs, more cancers could be detected by screening fewer people, as people at increased risk are more effectively identified. Giving someone their personal risk information may also help them make better decisions about how to live a healthy lifestyle and whether to participate in cancer screening programs.
- Senior Lead: Prof Jon Emery
- Project Lead: Sibel Saya
- UK team member: Dr Fiona Walter
- Others involved: Dr Jenny Walker, Prof Mark Jenkins, Assistant Prof Daniel Buchanan & Dr James Dowty (all University of Melbourne), Prof Ingrid Winship (Royal Melbourne Hospital)
Aims & Objectives
This PhD aims to look at whether a risk prediction model (using a DNA test) administered in general practice, used to guide bowel cancer screening, may be feasible, acceptable, and useful.
- To determine how clinically useful a lifestyle and genomic risk prediction model is to target bowel cancer screening in the Australian population
- To measure the feasibility and acceptability of administering a DNA test for bowel cancer risk in general practice
- To measure the effect of the DNA test result on participants’ risk perception, perceived control over developing cancer, perceived causes of cancer, cancer-specific anxiety, perceived self-efficacy to do bowel cancer screening, and cancer screening behaviour compared with the beginning of the study
Methodology & activity
Part 1: How good risk prediction models are at predicting who will and will not get cancer has been studied extensively. But, the ability of models to discriminate risk (i.e. whether the model helps us find the small part of the population who are at greatest risk) has not been well examined. Additionally, how the risk prediction model might fit in with the existing population bowel cancer screening program has not been explored. This study looked at these two issues for two risk models for bowel cancer in the Australian population.
Part 2: To offer most of the population a DNA test for bowel cancer risk would be done most efficiently through general practice. This study offered 150 general practice patients the opportunity to do this DNA test and we measured how many took it up, and how people made that decision. We also measured whether they were able to make an informed choice about doing the test during a quick consultation that is typical of general practice. We also measured participants’ reaction to the DNA test results and whether it helped them make better decisions about their bowel cancer screening.
Outputs & impact
These studies will provide evidence for the practical next steps of implementing a population bowel cancer screening program that is targeted based on risk.
Study 1 gives real world numbers for how many people would have their screening recommendations changed by using the risk prediction model as well as how many more bowel cancers could be detected in the screening program. These numbers can be directly translated into the Australian National Bowel Cancer Screening Program.
The findings from Study 2 can directly be used by general practitioners (in Australia and also in the UK) to inform how they discuss DNA testing with their patients.
A cost-effectiveness analysis would logically follow on from Study 1, i.e. a study that looks at the balance of costs between providing a risk assessment (e.g. DNA test) to the whole population against the potential gains from targeting screening more efficiently.
From Study 2, a study formally comparing (a randomised controlled trial) the effect of the DNA test on people’s tendency to do the right bowel cancer screening test at the right time will be compared to the generic advice that is currently given in general practice.