ESRC funded PhD studentship

Methods to reduce selection bias in health economic evaluation.

Economic evaluations often use observational data where the major concern is selection bias. Traditional methods for dealing with this bias suffer from important limitations, and can lead to misleading inferences for policy-making. The project will examine a novel non-parametric technique (Genetic Matching) that makes more appropriate assumptions. While Genetic Matching has been tested on a single dataset in labour economics, this research will further examine the technique using a series of health economic evaluations. The case studies are primarily from a developed country context (mainly the UK, and US) and include evaluations of clinical and public health interventions. This context offers an excellent opportunity to compare estimates from high-quality observational studies with those from pragmatic RCTs. This research will provide rigorous evidence on the technique's merits, which can ultimately help evaluations provide a stronger basis for policy-making. The PhD thesis will be undertaken within a larger project funded by ESRC/NIHR which involves collaboration with colleagues at Universities of California, Harvard and MRC Biostatistics Unit, Cambridge.

SUPERVISORS AND ENVIRONMENT

The project will be supervised by Dr Richard Grieve (senior lecturer in health economics). The successful applicant will also be able to draw on expert guidance from senior colleagues with expertise in statistics (Dr James Carpenter) and econometrics (Prof Jas Sekhon, University of California, Berkeley). As part of the studentship, the applicant will receive advanced statistical training, for example by taking study modules from the highly regarded Master course in Medical Statistics, at LSHTM. LSHTM provides a highly supportive environment for Research Degree study, and has a strong tradition for providing training for PhD students and junior health economists.

This project will be based in the Health Services Research Unit, in the Department of Public Health & Policy.