=======================================================================


                                SEMINARS

        MRC BIOSTATISTICS UNIT, CAMBRIDGE -  MICHAELMAS TERM 2003



======================================================================



9th October


Professor Ross Prentice,
Fred Hutchinson Cancer Research Center and University of Washington.



Title: Aspects of the Analysis of Multivariate Failure Time Data.



======================================================================



21st October


Professor A O'Hagan,
Department of Probability and Statistics, University of Sheffield.



Title: Probabilistic Sensitivity Analysis.


Abstract:

Mechanistic models are built in all branches of science
and technology to predict the behaviour of complex systems.
Examples in health include compartmental models for
pharmacokinetics and economic models for cost-effectiveness.
Invariably, users of such models must supply values for input
parameters.  Thus, a PK model may require the transition rates
between connected compartments, while a health economic model
may require relative risks of different drugs or mean resource
usages.  And there is invariably uncertainty concerning the "true"
values of these parameters.  Probabilistic sensitivity analysis (psa)
attempts to quantify the uncertainty in model outputs that is
induced by the uncertainty in model inputs.  There is substantial
interest in psa for all kinds of process models, and in particular the
new draft guidance from NICE emphasises its use in modelling
cost-effectiveness.

This talk will address three key issues in the use of psa:

* How should we specify the uncertainties in model inputs?

* How do we propagate that uncertainty through the model to
deduce uncertainty in model outputs?

* How can we decompose the overall output uncertainty to identify
which uncertain inputs are most influential?




=======================================================================



18th December


Professor Douglas G Altman,
Centre for Statistics in Medicine, Oxford.


Title: Randomised trials should be reported fully and accurately.



Abstract:

Biased results from randomised trials can mislead decision-making in health
care at all levels. Critical appraisal of the quality of clinical trials is
thus essential, both for individual trials and within systematic reviews.

Many RCTs are not reported adequately, however, hindering proper assessment
of their methodological quality. The CONSORT recommendations have been
widely supported but have as yet led to only modest improvement.

I will present evidence of the widespread poor methodological quality and
poor reporting of published trials, with illustrative examples. I will
present recent findings showing that within trial reports selective
reporting is common. Among other implications for systematic reviews and
will argue that all trial protocols should be published.




=======================================================================



9th December


Dr. Andrew J. Copas,
University College London.


Title: The sensitivity of estimates of the change in population behaviour
to realistic changes in bias in repeated surveys.


Abstract:

All surveys may be subject to bias which is not removed by standard
weighting procedures. Consequently when a survey is repeated, and
data from the two surveys compared, estimates of the change in
population parameters might in part reflect a change in the bias
affecting the survey. Typically limited external data are available
to estimate the change in bias directly. However one approach,
which is often possible, is to define in each survey a sample of
participants eligible for both surveys, and then compare the
reporting of selected events that occurred before the earlier
survey time point. A difference in reporting suggests a change
in overall survey bias between time points, although other
explanations are possible. A multivariate approach, considering
several events simultaneously, can be used to make sense of the
evidence. This analysis may help to make inference concerning the
real change in population parameters. We illustrate our approach
through application to the National Survey of Sexual Attitudes and
Lifestyles, 1990 and 2000.



=======================================================================


NOTE:

The seminars start at 2.30pm in the Large Seminar Room, 1st Floor,
Institute of Public Health, University Forvie Site, Robinson Way,Cambridge.


All welcome to attend.



=======================================================================




Daniela De Angelis
MRC Biostatistics Unit                          Tel: + (0)1223 330390
Institute of Public Health
Robinson Way                                    Fax: + (0)1223 330388
Cambridge
CB2  2SR                                        Web: www.mrc-bsu.cam.ac.uk
United Kingdom