Royal Statistical Society
RSS HERTS/BEDS LOCAL GROUP MEETING
"Problems and Solutions in the Use of the Proportional Hazards Model"
RICHARD KAY (Biostatistics & Programming, PAREXEL International)
On THURSDAY 13th JANUARY 2000 at 6.30pm at SmithKline Beecham (South Site),
Harlow. Tea at 6.00pm. Refreshments will also be available following the
meeting.
ALL WELCOME - There is no charge, but please inform Suzanne Clark (contact
details below) of your intention to attend by 6th January 2000.
"The Proportional Hazards Regression Model (Cox, 1972) has been the model of
choice for the analysis of censored time to event data for over 20 years.
Its appeal has been widespread and relates to a number of aspects; its
partially non-parametric form, its large sample efficiency when compared
with simple parametric models and its modelling flexibility to allow for
multiple events, for example, to list just a few.
Recent work, however, in the area of frailty has indicated that the model
may have certain drawbacks that are not directly apparent. Additionally, its
unthinking use in the area of multistate modelling has led to inappropriate
interpretation of data. This presentation will discuss the model and its
application, problems that arise and their avoidance and/or solution."
If you wish to attend this meeting, or for further information, directions,
etc., please contact Suzanne Clark (tel: 01582 763133 x2742 or e-mail:
[log in to unmask]) by 6th January 2000.
================================================================
Dr. Michaela J. Cottee E-mail: [log in to unmask]
Statistics Group Phone: +44 (0)1707 284377
University of Hertfordshire Fax: +44 (0)1707 284799
College Lane, Hatfield
Hertfordshire, UK. AL10 9AB
================================================================
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|