MIDLANDS AND NORTHERN UNIVERSITIES STATISTICS GROUP
THIRTY SECOND MEETING
Thursday 23rd April, 1998
Department of Medical Statistics,
De Montfort University, Leicester
The thirty second meeting of the Midlands and Northern Universities
Statistics Group will be held at the Department of Medical Statistics,
De Montfort University, Leicester on Thursday 23rd April, 1998.
Anyone interested is welcome to attend. The meeting will be held in
the Queens Building on the University City Campus.
Copies of this notice, together with the programme, abstracts,
registration forms and a campus plan are being sent to Statistics
Departments for internal distribution.
Car Parking facilities will be available in the Main Staff and
Visitors Car Park marked on the City Campus Plan. Please indicate
on the Registration Form if you require a Car Parking Reservation.
The Conference Fee is 40 pounds (+VAT) for Industrial delegates,
30 pounds (+VAT) for Academic Delegates and 20 pounds (+VAT) for
Full-Time students. This fee includes coffee, tea and buffet luncheon.
If you wish to attend, please complete a Registration Form and
return it as soon as possible.
Closing date for Registrations is Thursday 9th April, 1998
MIDLANDS AND NORTHERN UNIVERSITIES STATISTICS GROUP
THIRTY SECOND MEETING
Thursday 23rd April, 1998
Queens Building, De Montfort University, Leicester
PROGRAMME
10.45 Coffee and Registration
Room Q0.13
11.15 Lecture Theatre Q1.12
David Clayton
MRC Biostatistics Unit, Cambridge
Two-phase sampling in epidemiological
incidence studies
12.30 Buffet Luncheon - Fletcher Suite, Fletcher Building
14.00 Lecture Theatre Q1.12
Chris Jennison
Bath University
Group Sequential Analysis Incorporating
Covariate Information
15.00 Tea
Room Q0.13
15.30 Lecture Theatre Q1.12
Simon Thompson
Imperial College School of Medicine
The analysis of cluster randomised trials
16.45 Meeting Close
ABSTRACTS
David Clayton
MRC Biostatistics Unit, Cambridge
Two-phase sampling in epidemiological incidence studies
The classical epidemiological study designs for detecting association
between exposure and disease incidence are case-control studies and
cohort studies. Cox's partial likelihood method for the analysis of
cohort studies raises the possibility of sampling risk sets to produce
the hybrid "nested" case-control design.
An alternative design is the case-cohort study in which a random sample
of a cohort is studied in detail, after enrichment by all the cases
arising in the remainder of the cohort. For many years the only deviation
from straightforward random sampling to be considered for such two-phase
designs was matching of cases to controls in respect to confounding variables.
More recently more complicated sampling designs have been investigated.
These include "countermatching" in nested case-control studies, and two-phase
case-control studies. This paper will review the potential advantages of
these designs and discuss recent advances in analytical methods.
Chris Jennison
Bath University
Group Sequential Analysis Incorporating Covariate Information
We survey existing results concerning the joint distribution of the sequence
of estimates of the parameter vector when a model is fitted to accumulating
data and we provide a unified theory which explains the independent
increments structure commonly seen in group sequential test statistics.
Our theory covers normal linear models, including the case of correlated
observations, and asymptotic results extend to generalized linear models
and the proportional hazards regression model for survival data.
The asymptotic results are derived using standard methods for the
non-sequential case and they hold as long as these non-sequential techniques
are applicable at each individual analysis. In all cases, the joint
distribution of the sequence of parameter estimates has the same form,
exactly or asymptotically, as that of the sequence of means of an increasing
number of independent, identically distributed normal variables. Thus, our
results provide the formal basis for extending the scope of standard group
sequential methods to a wide range of problems.
Simon Thompson
Imperial College School of Medicine
The analysis of cluster randomised trials
In cluster randomised trials, groups rather than individuals are allocated
to intervention or control, but outcomes are measured on individuals. Such
trials are increasingly being used, especially for addressing public health
issues. In estimating the effect of intervention in such trials, the
between cluster variability in the outcomes has to be taken into account.
We advocate multilevel models for this purpose rather than, for example,
methods based on generalised estimating equations, partly because they
provide direct information on components of variance which are of interest
in their own right. In such methods, it is possible to adjust for baseline
covariates on either individuals or clusters. Trials in which clusters are
paired before randomisation pose particular analytical problems. However,
the analysis can be usefully approached as a random effects meta-analysis
across cluster pairs. Examples of the analysis of both unpaired and paired
cluster randomised trials will be discussed.
MIDLANDS AND NORTHERN UNIVERSITIES
STATISTICS GROUP
THIRTY SECOND MEETING
Thursday 23rd April, 1998
REGISTRATION FORM
______________________________________________________________
In order to register for the meeting please fill out this form
and return via post, fax or email.
Please tick below as appropriate :-
Industrial Rate : 40 pounds + VAT (47.50) [ ]
Full Academic Rate : 30 pounds + VAT (35.25) [ ]
Student Rate : 20 pounds + VAT (23.50) [ ]
Title....... First Name ........... Surname.................
Organisation..................................................
Mailing Address...............................................
..............................................................
....................................... Post Code.............
Tel:.................... Fax: ..............................
E mail........................................................
Please indicate if you will require Car Parking No [ ]
Yes [ ]
Please indicate if you have any special dietary requirements :-
..............................................................
..............................................................
Please also complete the Payment Details below :-
PAYMENT DETAILS :-
Please indicate those payment details that apply
(i) I enclose a cheque for ......... with this registration form.
The cheque should be made payable to
De Montfort Expertise Ltd
(ii) Please send an invoice to the name and address below :-.
....................................................
....................................................
....................................................
....................................................
Signed....................... Date................
Mail Registration To:
Dr Briony A Teather
Department of Medical Statistics
De Montfort University
The Gateway
Leicester LE1 9BH
England
Fax : +44 (0)116 250 6114
EMail : [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
