UCL Institute of Child Health - DOC

Document Sample
UCL Institute of Child Health - DOC Powered By Docstoc
					UCL Institute of Child Health
30 Guilford Street, London, WC1N 1EH



                           Lecturer in Biostatistics

                            Job reference: 1135136
Unit                                     Centre for Paediatric Epidemiology and
                                         Biostatistics
Grade                                    Lecturer in Biostatistics
Salary (inclusive of London allowance)   Grade 7 - £35,415 - £38,411 per annum.
                                         Grade 8 - £39,510 - £46,635 per annum.
                                         Salary will be dependant on skills and
                                         experience.
Reporting to                             Professor T Cole
Duration                                 Permanent
Hours of work                            36.5
Annual leave                             27 days per annum
Probation period                         3 years
Closing date                             Monday 21st June 2010

Background to post

This post is based in the MRC Centre of Epidemiology for Child Health/Centre
for Paediatric Epidemiology and Biostatistics and is a tenured HEFCE funded
post.

There is an exciting opportunity for someone with a strong statistical
background to join a group of senior statisticians and epidemiologists within
the MRC Centre of Epidemiology for Child Health/Centre for Paediatric
Epidemiology and Biostatistics. In particular, the successful applicant will
work as part of the Millennium Cohort Study (MCS) Child Health Group.
He/she will develop their own research programme as well as working with
others in the group, and taking a lead on statistical issues.
The post-holder will also provide statistical teaching and support to
researchers and PhD students within the Centre. There will be opportunities
to contribute to teaching on other short statistics courses within the
ICH/GOSH and for contributing to external courses on specific advanced
topics.
Applications are welcomed from qualified statisticians with experience of
research within a multi-disciplinary environment. Applicants must have a PhD
and a track record of peer reviewed publications in high quality journals.
Experience of teaching statistics to a non-statistical audience is also required.
Interested applicants are invited to contact the Head of the Centre, Professor
Carol Dezateux, Professor Tim Cole or Dr Angie Wade, Senior Lecturer in
medical statistics.



                                             1
Job Specification

The postholder will:

1) Be a member of the Child Health Group of the Millennium Cohort Study,
   and:

   a) Develop an applied statistical research programme related to the
      Millennium Cohort Study. This research may involve development of
      techniques for dealing with missing data, age-related changes in
      outcomes and gaining maximum information from observational data.
   b) Take a lead for the group on statistical and data management issues
      relating to the cohort, including liaising on these issues with EpiLab
      (the Centre’s in-house IT resource) and the research team managing
      the MCS (at Institute of Education).

   c) Collaborate with other members of the MCS Child Health group, on
      general issues and on specific projects.
   d) Liaise with others in the Centre who are working on related issues in
      other cohort studies.
2) Provide training within the Centre at a more advanced level of statistics
   than the basic courses currently running. The postholder will assist with
   providing training across the wider joint institutions of ICH/GOSH and on
   external courses as necessary. It is anticipated that this will take no more
   than 30% of the postholder’s time.
The appointee will be expected to carry out any additional duties as may
reasonably be required within the general scope and level of the post.




                                      2
Person Specification

Requirements      Essential                            Desirable
Qualifications/   PhD in biostatistics, medical        MSc in applied statistics
Training          statistics or related subject, or    or equivalent
                  equivalent qualification             Teaching qualification
                                                       such as membership of
                                                       Institute of Learning and
                                                       Teaching
Experience        Collaborative publications in high   Research publications in
                  quality peer-reviewed journals       statistical methods

                  Contribution to preparation of       Development of written
                  study protocols and grant            training material aimed
                  applications as co- or statistical   at a non-statistical
                  investigator                         audience

                  Teaching statistics and research     Experience of course
                  methods to non-statisticians         development – from
                                                       design through to
                  Small group tutorial training        presentation

                  Experience of working in multi-
                  disciplinary teams on relevant
                  research projects

                  Experience of supervision of
                  research and higher degree
                  students
Skills             Advanced statistical skills in
                   relevant area e.g. multiple
                   imputation, multi-level modelling

                  Familiarity with statistical
                  software used in the department
                  e.g. Stata, R, SPSS

                  Supervisory skills

                   Teaching skills
Personal          Ability to work as a member of a
qualities         team

                  Initiative and leadership

                  Excellent communication skills




                                        3
Background
Centre for Paediatric Epidemiology and Biostatistics
The mission of the Centre is to improve the health and well being of children
through collaborative research that aims to:

 Improve knowledge of the causes and mechanisms underlying child health
  and disease

 Enhance the scientific basis of strategies for the prevention, diagnosis and
  treatment of diseases affecting children and the adults they will become
  and for the promotion of their well being

 Contribute to clinical and public health decision making

 Increase research capacity through the provision of training in
  epidemiology, biostatistics and academic public health applied to children.

We are internationally renowned for our research which is characterised
within six themes:

      Mathematical models and statistical methods
      Infections of the mother, her fetus and child
      Health inequalities and life course epidemiology
      Genetic and developmental disorders of the fetus and child
      Ophthalmic epidemiology
      Research for policy

A common thread in these themes is to understand the genetic, biological,
developmental, environmental and social mechanisms underlying the early life
origins of illness, health and well being in childhood and adult life, and the
longer term consequences of early life exposures and interventions. The
Centre provides an independent source of scientific expertise on children’s
health for the Medical Research Council (MRC), government, the NHS, the
National Institute for Health and Clinical Excellence (NICE) and others, and
contributes to public health and clinical decision-making for children by
syntheses of research evidence relevant to policy and practice.

The Centre hosts the Child Health Group of the Millenium Cohort Study
(MCS) which comprises a random sample of all births in the UK over a period
of 1 year, with oversampling of ethnic minority and disadvantaged children.
The cohort has over 18,000 babies enrolled and is serially assessed, with the
4th sweep at age 7 years. This dataset provides a particularly rich source of
information and has so far yielded 25 peer-reviewed publications from
researchers based within the Centre alone. Topics include physical growth,
breastfeeding, immunisation, inequalities in health, infectious diseases and
methodological papers.




                                       4
The MRC Centre of Epidemiology for Child Health

Our scientific achievements and excellence have been recognised in our
award as a MRC Centre of Epidemiology for Child Health, one of only
three such Centres at UCL. The MRC Centre award adds value to our existing
science by exploiting methodological and collaborative opportunities in four
‘cross cutting’ areas - statistical methods, genetic epidemiology, electronic
health record research and research for policy. Working groups in these areas
are tasked with strengthening the scientific programmes within the Centre
through research within and across the existing themes. New partnerships
have been established to ‘add value’ to the Centre’s programmes of work.
This includes the appointment of Professor Harvey Goldstein as honorary
part-time professor of statistics, and opens a link with the ESRC funded
research node on multilevel modelling
(http://www.ncrm.ac.uk/about/organisation/Nodes/LEMMA/LEMMA.php)

The Centre’s research environment is enhanced by the availability of a
number of large and complex datasets that are ideally suited to the
development of novel statistical methods. They include large national cohort
studies, particularly the British 1958 and Millennium Cohort Studies, several
major infectious disease cohorts, the newborn biobank, and special cohorts of
children with genetic and developmental disorders. In addition the use of large
electronic health and vital statistics record datasets is a feature of all these
themes. A senior information systems consultant is establishing an informatics
team to support the academic work of the Centre and to implement an
advanced computing facility.

Our current statistical interests include both methodological and applied
research and are detailed in appendix 1.

The Centre provides an excellent research, training and career development
environment for statisticians. It is one of the largest of the 25 Academic
Research Units within the Institute of Child Health and currently comprises
seven professors, two readers, five senior lecturers, around 40 researchers
including Wellcome, Department of Health and MRC career scientists, MRC
training fellows, and 12 PhD students. There are monthly Departmental
informal seminars where postgraduate students and research fellows are
encouraged to present their work, formal, monthly Departmental seminars
with external speakers, and a monthly journal club. We are responsible for a
Masters course and run short courses, for example, in evidence-based child
health, statistics and research methods.

The senior statisticians within the Centre (Professors Tim Cole and Harvey
Goldstein and Drs Mario Cortina Borja and Angie Wade) provide supervision
and mentorship to junior statisticians in the department, most of whom are
registered for higher degrees. There are excellent postdoctoral training
opportunities and staff are encouraged to take advantage of UCL’s extensive
programme of courses for specialist and generic skills training and career
development.



                                       5
A monthly Statistics Discussion Group led by Dr Wade brings together all
statisticians employed within our Centre as well as any based in other UCL-
ICH departments. Statisticians within the Centre are actively involved in
formal journal reviews and participate actively in the established UCL
Biostatistics Network. We have direct and strong links with the Royal
Statistical Society (Dr Cortina Borja currently sits on the RSS publications
panel, is a committee member of the General Applications Section and is on
the Editorial Board for two of the society's journals; Professor Cole was
recently a member of the Society's Council; Professor Goldstein holds a Guy
Medal in Silver).

Statistical Support Service (S-cubed)

The Centre provides a Statistical Support Service (S-cubed) for researchers
and students across the joint institutions of ICH and GOSH. The aim of this
service is to contribute to the maintenance of high standards of published
research. The assistance offered covers the whole process of research, from
refining the research hypothesis and helping design the study through to
writing up the results for publication. Researchers are encouraged to seek
advice at the early stages of study design and S-cubed members may
become full project collaborators where their input is high. Professor Tim Cole
is Head of the Statistical Support Service.

Statistics Teaching at ICH/GOSH

The Centre provides statistics and research methods training across the joint
institutions for MSc and PhD students as well as other research staff. It also
provides external courses catering for a wider paediatric audience. Dr Angie
Wade is Director of Statistics and Research Methods Teaching at ICH/GOSH.

Further information about the UCL Institute of Child Health can be found on
our website http://www.ich.ucl.ac.uk/slms/

Full UCL terms and conditions for Research and support staff
http://www.ucl.ac.uk/hr/salary_scales/Support_Research_tcs.php


Informal Enquiries

To find out more about the post please contact Dr Angie Wade (email
a.wade@ich.ucl.ac.uk), Professor Tim Cole (tim.cole@ich.ucl.ac.uk) or
Professor Carol Dezateux (c.dezateux@ich.ucl.ac.uk).

For further information about the Institute of Child Health, please visit our
website at www.ich.ucl.ac.uk

We particularly welcome female applicants and those from an ethnic
minority, as they are under-represented within UCL at these levels. This
is in line with section 48 of the Sex Discrimination Act and section 38 of
the Race Relations Act.


                                        6
Appendix 1

The following statistical methodology topics are relevant to the research
carried out in the Centre.

1. Statistical methods for epidemiology

      Several projects in the Centre are based on large national and
      international birth cohorts, with an emphasis on child to adult growth
      and health trajectories especially in relation to obesity and early life
      influences on health and disease in childhood and adult life.

            Longitudinal data analysis: Apart from the development and
      application of classical methods for longitudinal data, three areas of
      interest in this context are propensity scores and their implementation,
      functional data analysis and visualization methods for complex data
      sets.

            Multilevel: One important component in many analyses carried
      out in the Centre is a hierarchical clustering structure which requires
      multilevel models. Methodological problems concerning multivariate
      and non-normal responses are of interest.

             Imputation: Development and application of these methods is
      of interest in the context of the analyses of data from birth cohorts and
      surveillance studies but also in data linkage problems where attrition or
      missing values occur frequently.

            Evidence synthesis: Bayesian methods for condensing
      evidence from different sources to estimate classic epidemiological
      parameters have been used in research carried out in the Centre
      regarding the HIV epidemic in the UK and it is of interest to extend
      these results to other conditions.

2. Applied statistical methods

      Often, data from projects based in the Centre or the Institute generate
      research into specific applied statistical methods. The following areas
      are currently being developed.

             Finite mixture models: Researchers at the Centre have
      considerable experience in developing latent class regression finite
      mixture models; these have been extensively applied to the analysis of
      infectious disease. Statistical issues currently under investigation are
      diagnostic and model selection procedures, and extensions to
      multivariate responses.

            Models for discrete data: Long-tailed and value-inflated data
      are a feature of many epidemiological and clinical studies. Research in
      the Centre looks at methodological aspects of fitting regression models


                                      7
to discrete data; these include estimation, diagnostics and model
selection procedures.

      Models for seasonal variation: Analyses of cyclical patterns
both as responses and explanatory or mediating effects are carried out
in the Centre in a wide range of epidemiological and clinical studies.
Current research focuses on using copulas to model bivariate
responses and model diagnostics for circular regression models.

      Age-related reference ranges: The Centre has an international
track record in this area and offers many opportunities for related
research. Specific areas of interest are the development of multivariate
standards using copula methods, and the use of probabilistic models
including parameters of location, scale, skewness and kurtosis in the
construction of standards.

      Multiple testing: High dimensional studies into the genetics
components of complex disease pose problems of multiple
comparisons. Research on multivariate permutational tests as an
alternative to Bonferroni-type corrections is carried out in the Centre
and is increasingly important as the Centre will increasingly conduct
genome-wide studies.

 Methods for data linkage: The Centre manages several
surveillance projects which require reliable record matching in data
linkage problems as well as preserving anonymity in all parts of the
study – including dissemination of its results. The two main statistical
aspects of interest in this area are probabilistic record matching and
methods to control data disclosure. The design and implementation of
statistical and computational methods to perform these tasks is a key
part of the Centre’s future development.




                                8

				
DOCUMENT INFO