MRC Capacity Building PhD Studentship - Predictive Modelling of Infertility Treatment Data and its use in Informing Policy

Faculty of Medical and Human Sciences

University of Manchester

The School of Community Based Medicine is inviting applications for a distinguished three-year MRC Capacity Building PhD studentship commencing from September 2008. The studentship is open to UK/EU nationals only due to the nature of the funding and will attract an annual stipend of £14, 940 in addition to tuition fees. Non-UK EU nationals must have been continuously resident in the UK for a minimum three year period for employment or educational purposes to qualify for the funding support.

The high twinning rates associated with assisted conception (AC) treatments (commonly referred to as IVF) and their associated increased risks of adverse maternal and perinatal outcomes are a growing concern, both from a health and economic perspective. From 2009 clinics will have to meet targets for twinning rates, primarily by greater use of single-embryo transfer and the development of better embryo selection techniques to reduce the incidence of twins. However such changes in treatment policies will lead to reduced success rates. Statistical modelling is being utilised to compare treatment policies and quantify the gains and losses associated with potential changes in treatment and embryo selection protocols - and to develop appropriate treatment selection policies. In the longer term we aim to use modelling techniques to monitor and optimise treatment.

The statistical modelling is complex due to the fact that the outcomes are only partially known and there are a large number of confounding variables and several important predictors act in a non-linear manner. This PhD project will develop and implement modelling methodology that will be of value across a wide range of policy decisions, from risk assessment to economic evaluation.

The aim of this project is to utilise a number of large, rich, datasets along with simulation studies to develop further the statistical methodology associated with modelling AC data and its use in predicting outcomes for treatment scenarios. The project will examine both model structures and their estimation, but also the definition, interpretation and usefulness of alternative target prediction statistics, and the impact of random effects on prediction, both in terms of mean values, but also uncertainty and sensitivity.

The project requires the fusion of a complex statistical modelling approach and statistical techniques for model selection, validation and prediction more commonly used with simpler linear models. It will therefore require the development of good computational skills as the models cannot be fitted directly using standard statistical software.

Applicants should hold, or expect to obtain, a first or upper second-class honours degree in a relevant subject with strong methodological content such as statistics, econometrics or exceptionally social/psychological/physical science. The project will be based within the newly formed Health Methodology Research Group, an integration of biostatistics, health economics and informatics. PhD graduates in this discipline are well placed to progress into the biomedical R&D sector or obtain post-doctoral positions.

Interested candidates should submit a CV and detailed covering letter outlining their suitability for the project to Professor Andrew Pickles at andrew.pickles@manchester.ac.uk. Contact details for two academic referees should also be provided.

Applications will be welcomed up to and including Friday 30th May 2008.

For further details on the study or to confirm your suitability please contact Professor Andrew Pickles (andrew.pickles@manchester.ac.uk) or Dr Steve Roberts (steve.roberts@manchester.ac.uk).

www.medicine.manchester.ac.uk/research/groups/methodology