Statistician TBD

Posted by Naura | 2:29 AM

Statistician TBD Location of PhD fellow Julius Center of the University Medical Center Utrecht Qualifications

The candidate should have an MD, PharmD or degree in biomedical sciences or related field. Experience in research and training in epidemiology are appreciated. Motivated to work in cutting edge drug development research is a prerequisite

Application
Candidates can apply by sending their CV and application letter to Mirjam Knol (m.j.knol@umcutrecht .nl).
Please indicate in your application letter how you heard of this vacancy.

Employment terms
The candidate will have an appointment at the Julius Center of the University Medical Center Utrecht (www.juliuscenter. nl) and will closely collaborate with people from other institutions. We offer a temporary full time appointment as a PhD student for one year with the perspective of prolongation with another three years.

Contact People for information on project/application

M.J. Knol, University Medical Center Utrecht,
+31 (0)88 755 9380,
m.j.knol@umcutrecht .nl

P.A. van Meurs, Project leader Escher programme
+31 (0)88 755 5196 or mobile +31 (0)621 512631, p.a.vanmeurs@ umcutrecht. nl

Introduction of the Research topic/ background (300 words)
In the past decades the RCT has been viewed as the paradigm study design to determine the intended effects of drug interventions. In addition RCTs are important to assess type (i.e. more common, predictable) unintended effects of drugs. Typically these RCTs are executed in three phases during a period of 10 years prior to drug registration.
Currently, trials are not designed, conducted and analysed in the most efficient and expeditious manner and several improvements in the design, conduct and analysis phase could lead to increased efficiency.
For example, prevention of losses to follow-up and incomplete data, and sophisticated methods to impute missing values if they occur can increase the effective sample size. Sophisticated methods to indicate when there is enough information to stop the trial can shorten the duration of the trial and decrease the number of patients to be included. By using information of earlier trials when analysing a new trial, evidence of efficacy of a drug might be available earlier and less trials have to be performed.

Research Questions (200 words)
How can the design, conduct and analysis of randomized controlled trials on drug effects be improved?
Improvements of design phase:
selection of patient populations
sample size estimation
randomization and blinding
primary, secondary and composite outcomes
Improvements in conduct phase:
recruitment and compliance of study participants
follow-up of study population
stopping trials early
Improvements in analysis phase:
imputation methods for 'lost-to-follow- up' and incomplete data
pooling of studies

Methods (300 words)
Improvements in design phase:
Review of the literature summarizing different strategies to select
patient populations, to perform sample size estimations and to conduct
randomisation and blinding, and analysing which of these different
strategies are most effective. This review can then be used to
formulate a recommendation on designing RCTs.
Comparison of existing and development of new statistical methods to
analyze composite endpoints using empirical data.
Improvements in conduct phase:
Review of the literature on different strategies to recruit and
follow-up study participants and summarizing different strategies to
improve and monitor compliance. This information is used to analyse
which strategies are most effective and to formulate a recommendation
on improving conduct of RCTs.
Comparison of different methods and strategies, for example
conventional and Bayesian analyses, to stop trials early, using
empirical data of already performed trials.
Improvements in analysis phase:
Comparison of different imputation methods of increasing complexity
for random and non-random missing values due to 'lost-to-follow- up'
and incomplete collection of data. Empirical data of trials will be
used to compare these different methods.
Comparison and application of different meta-analytical (Bayesian)
techniques to analyse data of a trial using data from previously
performed trials. Comparison and application of different
meta-analytical (Bayesian) techniques to pool phase III and IV
studies.

Deliverables
(300 words)
Recommendation to improve design phase of RCTs in terms of selection
of patient populations, sample size estimation, randomisation and
blinding, and composite outcomes.
Recommendation to improve conduct phase of RCTs in terms of
recruitment, compliance and follow-up of study participants, and
stopping trials early.
Recommendation to improve analysis phase of RCTs in terms of missing
data and pooling of studies.
5 Papers in international scientific journals
PhD Thesis
Presentations at (inter) national scientific meetings

Timelines and Milestones
(per 6 month episodes)
(200 words)
0-12 months: Reading relevant literature and research training in Epidemiology
12-18 months: Review and analysis of different strategies in design
phase and recommendation of improvements
18-24 months: Review and analysis of different strategies in conduct
phase and recommendation of improvements
24-30 months: Comparison of methods of stopping trials early
30-36 months: Comparison of imputation methods for missing data
36-42 months: Comparison of meta-analytical techniques to pool studies
42-48 months: Finalization of publications and PhD thesis. Most of the
aforementioned periods will be finalized by a manuscript

--
Best Regards,

Grace Wangge

--
Best Regards,

Grace Wangge






0 comments