I currently direct the statistical analysis in the projects listed here. This body of work covers many different elements of analytics – developing new methodologies to deal with the complex problems of heterogeneity in large databases, establishing new national measurement schemes for public health issues, collaborating with other experts to answer interesting and creative research questions, and more….
The Global CoLaboratory – Angiogenic Factor Study for Pre-Eclampsia and Hypertension in Pregnancy
Collaborators – Prof. Chris Redman (Oxford), Prof. Annetine Staff (Oslo), Dr. Samantha Benton (Ottawa), Prof. Peter von Dadelszen (Vancouver)
The Global CoLaboratory aims to improve the health of women and their infants by facilitating research addressing adverse pregnancy outcomes. In this study, merging algorithms have been developed to pool individual patient data of over 16,000 pregnancies worldwide to create the largest known database of angiogenic factor measurements. The information available in this unique resource is currently being used in Phase II of the project – discovery of the angiogenic processes relating to pre-eclampsia and other adverse outcomes in pregnancy.
Comprehensive Geriatric Assessment – cost-effective delivery
Collaborators – Prof. Sasha Shepperd (Oxford), Dr. Mike Gardner (Oxford)
Prof. Sasha Shepperd is the chief investigator for two projects for the National Institute of Health Research: A multi-centre randomised controlled trial of Comprehensive Geriatric Assessment in an admission avoidance hospital at home setting and Implementing Comprehensive Geriatric Assessment in a cost-effective way.
Radiological Protection in Ireland
Collaborators – Dr. Patrick Murphy (Dublin), Environmental Protection Agency (Ireland)
As the second highest cause of lung cancer, high levels of indoor radon concentrations is a significant public health concern – resulting in over 170,000 deaths annually. This project focuses on statistical analyses to improve radon measurement in Ireland and to further our understanding of the environmental factors that play a role in increasing radon levels. In 2010, a new national radon measurement strategy was implemented in Ireland as a direct result of this research.
Time-Series Cross-Sectional Models
Time-Series Cross-Sectional (TSCS) datasets are multivariate time-series data recorded over a set of groups or panels. My methodological research in this area includes development of a statistical framework of TSCS methods along with the development of advanced TSCS models for application to multivariate data. Applications include an analysis of economic openness in OECD countries and a study of the affects of climate and environmental factors on indoor radon levels.
Collaborators – Dr. Alejandra Gonzalez-Beltran, Dr. Philippe Rocca-Serra, Dr. Susanna-Assunta Sansone (Oxford e-Research Centre)
STATO is a general-purpose STATistics Ontology that provides coverage for processes such as statistical tests, conditions of application and inputs for and results from statistical methods
Estimating Risk of Perinatal Asphyxia
Collaborators – Dr. Antoniya Georgieva (Oxford)
Research at the Oxford Centre for Fetal Monitoring Technologies focuses on the relationship between fetal heart rate and fetal health prior to and during labour. This project uses multiple statistical techniques to determine how best to predict risk of asphxyiation during labour.
Other projects include:
Validation of self-reported activity versus accelerometer measures (British Heart Foundation, NDPH, Oxford)
Study of proteinuria and associated factors (Prof. Laura Magee, UBC, Vancouver)
Statistical analyses of the deterioration processes of heritage stones (Katrin Wilhelm, Oxford)
Seasonal Behaviour of Psychiatric Admissions Data (Oxford-Dublin Collaboration)
Hip Arthroplasty Gait Analysis (Dr. Stephen Mellon, Oxford)
Non-response in National Household Surveys (Dr. Patrick Murphy, Prof. Richard Sinnott, Dublin)