Newcastle has an internationally recognised biostatistics research group who work at the interface of theory and practice. Members of the group have made significant contributions to both statistical methodology and clinical understanding.
One active research area is nonparametric curve fitting and prediction, based on Gaussian process regression for longitudinal/functional data. This is being used, amongst other things, to develop methods for analysing human movement, which are particularly relevant in the rehabilitation of paraplegics. The group also works on missing data problems for longitudinal analysis. This is being extended to missing data in clinical trials, especially crossovers and meta-analyses with missing confounders. There is also interest in both the design of clinical trials and biological assays, and in causal inference from observational data, including dynamic treatment allocation.
There is a strong tradition of collaborative research with colleagues in the Faculty of Medicine and there are links with the Clinical Trials Unit based in the Institute of Health and Society. The many ongoing projects include the analysis of healthcare indicators in HIV infection, dynamic anticoagulation treatment allocation, survival time prediction for hospice patients, cancer epidemiology and prognosis, trial design for haemodialysis in babies, analysis of resurfacing data related to total hip replacement, the effect of lifestyle on blood pressure and the effect of dietary intake on wellbeing in children.
For information on publications, please see the personal web page of each member of the group.
Analysis of longitudinal data with drop-out: Objectives, assumptions and
Diggle P., Farewell D., Henderson R. Journal of the Royal Statistical Society. Series C: Applied Statistics 56, 499–550, 2007.