EFPSI/PSI Conference 2008
Eddie gave a talk and presented a poster on
"Fitting Straight Lines to Scatter Graphs when Both Variables are Subject to Error".
Earlier Conferences
A talk at the PSI conference in 2007 was on
"Presenting the Uncertainty of Sample Size Calculations".
A talk at the RSS/PSI conference in 2005 (in collaboration with ClinPhone) was on
"When To Close Screening / Run-In".
Two presentations at a Henry Stewart conference on CPMP Points to Consider in 2003 were:
1)
Missing Data: Problems highlighted with the “Points to Consider”
document
2)
Adjustment for Baseline Covariates
[abstracts]
Two talks were given at the PSI conference in 2002:
"Multiple imputation for handling missing covariates" and "A Fresh
Look At ‘Nonparametric’ Estimation For Tied Data". Please
contact Chirostat for more information.
Confidence intervals for multiple
comparison tests
Do you want to do powerful multiple
comparison tests and still quote confidence intervals and p-values? See the Journal of Biopharmaceutical Statistics
(2001: 45-63) for a solution covering Williams'’ and other step-down
multiple comparison tests against control.
[abstract]
|
Pharmacokinetic
Studies
Chirostat has worked with major pharmaceutical and smaller companies on the analysis of PK
studies (including dermato- pharmacokinetics). Chirostat can provide a sample size calculations, data entry,
statistical analysis, report and graphical presentations.
Investigator
training on rating scales
Statisticians can help to ensure that
investigators complete rating scales consistently in areas such as schizophrenia.
Chirostat offers simple methodology supported by an Excel /
Visual Basic program for use at investigator training meetings (ECNP
Congress, 1998).
Lateral Thinking
Eddie can offer a fresh look at
your problem – ‘lateral thinking’. For example, he
raised money for charity by not doing something – fortunately for
everyone’s ears, staff at Knoll were very generous in paying him not to
play the cello.
Greek
letters
As
statisticians, we like to think that we own the Greek alphabet (or
perhaps just share it with mathematicians). The founders of PSI chose
the symbol Y
for its logo. The nice thing about Greek letters is that they
evoke memories of warm holidays in the Greek islands (alternatively
sitting your degree finals in hot and stuffy examination halls!)
|
Multiple imputation
Eddie
has investigated multiple imputation, looking at SOLAS
and SAS and applying the method in two clinical trials. Eddie is offering a course
‘Missing values: CPMP guidance and multiple imputation’.
PSI
activities
Eddie
has always been a keen member of PSI, and has given several talks and served
on the Public Affairs subcommittee.
Employment
history
Eddie has been a freelance statistical consultant
since leaving Knoll Pharmaceuticals in 2001. His earlier career was
mainly in Nonclinical Statistics but in recent years has done a lot of work in
Clinical Statistics. He is particularly interested in practical
statistical problems and supporting other statisticians.
Why Chirostat?
The
three Greek letters which perhaps most represent statistics are c (
c²
test,
c²
distribution),
r
(correlation) and
S
or
s
(variance
or summation). Chiro indicates medical statistics – whether a molecule is
right- or left-handed is often crucial to
its efficacy. –
|