Our research is underpinned by experimental design and data analysis including descriptive statistics, hypothesis testing and bioinformatics.
We have and will continue to contribute to such biostatistical research.
Our objectives are to:
- To develop methodology to better design and analyse studies in cancer screening;
- To estimate the natural history of pre-cancer so as to better plan screening programmes;
- To collate and analyse general cancer statistics for purposes of surveillance and planning and to relate changes in risk factors to changes in cancer incidence;
- To develop tools for the analysis of clinical trials subject to non-compliance and even cross-over;
- To more efficiently design and analyses clinical trials to simultaneously study multiple treatments;
- To apply statistical techniques to provide new solutions to problems in bioinformatics;
- To consider communication of harms and benefits of preventive interventions taking into account strength of evidence;
- To provide statistical support for projects that complement the programme’s research in cancer screening.
Main ongoing projects include:
- Statisticial research for cancer screening
- Natural History estimation for screening evaluation and planning
- Models for multiple events with applications to infections with multiple types HPV types
- Benefit-harm analyses adjusting for event severity with application to chemoprevention and screening
- Multiple comparison problems in micro- array studies
- Pre-specifying order of hypotheses to allow sequential p-value computation
A full list with details can be found under