We are inviting applications for a PhD studentship in statistical
epidemiology, funded by GlaxoSmithKline for 3 years, starting at a
mutually convenient date. The studentship is open to EU citizens and
covers tuition fees and a generous living allowance. The project
concerns statistical methods in pharmacoepidemiology, using large
primary care data bases, such as the General Practice Research Database,
as described in the project outline below. The project supervisor is
James Carpenter, and the successful candidate will be collaborating with
statisticians and epidemiologists at both LSHTM and GSK to develop a
stimulating doctoral program of research. Applicants must have an MSc
(or equivalent) in statistics. Some knowledge of epidemiology is
desirable but not essential: an aptitude for applied methodological
research in statistics is more important. Applications, including a CV
and the names of two referees, should be sent to Dr James Carpenter,
Medical Statistics Unit, London School of Hygiene and Tropical Medicine,
Keppel Street, London WC1E 7HT (email [log in to unmask]), from
whom further particulars can be obtained. For an informal discussion
telephone James (020 7927 2033) or Professor Stuart Pocock (020 7927
2413). The closing date for applications is Friday 20 August 2004.
Project outline: Statistical methods in pharmaco-epidemiology using
large general practice databases This project concerns design issues and
statistical methods in pharmaco-epidemiology, i.e. the study of the use
and effects of drug treatment in populations, both as regards
effectiveness and safety. Although this can be viewed as an application
of epidemiologic methods to pharmaceuticals, the nature of the large
routinely collected databases, which typically form the principal source
of information, mean that this field poses challenges that often require
special solutions in both study design and statistical analysis.
Further, the complex and dynamic context of pharmaco-epidemiology gives
rise to fascinating and unique statistical challenges. Such challenges
need to be addressed if the information in such databases is to be
reliably and routinely used. For instance, for any particular class of
drugs (e.g. statins for reducing risk of coronary disease) it would be
appropriate to investigate possible associations with several other
diseases (e.g. Alzheimer's disease, eye cataract, suicide etc). The
case-cohort design seems well suited to such problems by making use of
all disease cases together with a random sub-sample of the whole cohort.
Nested case-control studies for each disease might be a suitable
alternative approach. The analysis of data from large routinely
collected databases also presents unique challenges. Propensity scores
and weighting methods have been proposed in other settings to reduce
bias caused by non-random allocation of treatments. However, their
application to large scale database analyses poses some special
challenges, not least because of the considerable quantity of missing
observations, which are inevitable in a large amount of routinely
collected data. Subjects who have complete data on all exposures and
confounders are likely to be both unrepresentative and a relatively
small proportion of the total. Thus a conventional approach, such as
using only individuals with complete data, is likely to be both biased
and underpowered. We intend to pursue a program of methodological
research to address these questions. In order to focus and enhance the
practical relevance of this research it will be closely linked to and
illustrated by specific potential drug-disease associations of interest
to pharmaco-epidemiologists. The precise sequence of methodological
issues to be tackled will unfold over time but specific potential topics
of interest are as follows: 1. How do case-cohort and traditional
case control designs compare, in terms of efficiency and likely costs?2.
One problem with the case-cohort approach is the difficulty in
taking account of general practice effects since there is no matching of
cases to controls. How much does this apparent deficiency matter?3.
What size of sub-cohort should be selected and how one should restrict
its sampling frame to take account of the characteristics of disease
cases?4. It has been suggested that repeat use of the same
sub-cohort for multiple disease-drug inferences may have some
statistical penalty re non-independence;5. The use of nested case
control studies has the inefficiency of needing a separate selection of
matched controls for each set of disease cases. However, one can easily
match on practice and then the statistical analysis methods are better
established. It will be valuable to compare results (and the effort to
produce them) for nested case-control and case-cohort design for the
same drug-disease questions;6. Development of methods for coping
with missing data in the analysis, including the use of propensity
scores and related approaches when data are missing. This would involve
the exploration and development of multiple imputation methods that take
into account (i) information on the distribution of exposures and
confounders available in the wider epidemiological literature, and (ii)
the hierarchical structure of the data (with patients nested in
practices etc).7. Given the novel approaches required, the
development of appropriate statistical software and experience in
handling such large databases will be useful for future applications.
Our methodological findings, backed up by experience in real-world
pharmaco-epidemiological questions should have a major impact on how
best to make use of such large databases as the GRPD for studying drug
safety and effectiveness in a primary care setting. Stuart Pocock
Medical Statistics Unit
London School of Hygiene and Tropical Medicine
Keppel Street
London WC1E 7HTTel +44 (0)20 7927 2413 (direct)
2230 (secretary)Fax +44 (0)20 7637 2853
