hi all,
YSS and the East Cumbria local group have teamed up to organize 
the following seminar on medical statistics:


A joint event between the YSS (young statisticians section) and the
Lancashire and East Cumbria local group will take place 3-5.30pm on
Thu 14 June at Lancaster University (Fylde LT1) LA1 4YF. It will be
followed by a wine reception in the Postgraduate Statistics Centre.

We welcome three career young statisticians, working in the area of
medical statistics: Patrick Phillips (MRC clinical trials unit), Tom
Fanshawe (CHICAS, Lancaster Medical School) and Peter Kimani (Health
Sciences, University of Warwick). Titles of talks and abstracts follow
at the end of this email.

We look forward to seeing you here in Lancaster. There is no need to
register, but if you have any queries please contact Matt Sperrin
([log in to unmask]) or Lisa Hampson ([log in to unmask]).



***
3pm: Patrick Phillips: Eradicating the world's oldest infectious
disease: challenges in designing clinical
trials of new treatments for tuberculosis

Tuberculosis (TB) is considered the oldest infectious disease known to
man, but continues to be major global health problem in the 21st
Century. The World Health Organisation declared TB a global emergency
in 1993, and in 2010 there were 8.8 million new cases and over 1
million deaths from TB.

The current standard of care is a 4-drug 6-month regimen that has been
consistently shown to have efficacy of 90-95% in randomised controlled
trials in a variety of settings. These numbers are not reflected in
clinical practice with cure rates unfortunately much lower that those
seen in clinical trials. The rising levels of drug resistance and
co-infection with HIV also contribute to global morbidity and
mortality.

New combination regimens for TB that are shorter, safer and easier to
adhere to are urgently needed. There are a number of new drugs in the
pipeline with two or three now reaching phase III trials. Licensed
drugs are also being repurposed at different doses or for new
indications

In this talk I will discuss some of the challenges in designing and
conducting late phase clinical trials to evaluate new regimens for the
treatment of TB. These include: designing trials of combinations
rather than compounds; the need for non-inferiority trials and the
complexities in defining the margin of non-inferiority; distinguishing
between relapse and new re-infection and the role of adaptive designs.

None of these challenges are unique to the disease area of TB, but I
will show what work has been done to address them in this context and
also what questions are outstanding and where there is work to do!

***
3.45pm: Tom Fanshawe: Multilevel models for cancer progression

Biological studies of cancer progression typically generate multilevel
data: numerical summaries of properties of cells nested within tumour
sites within individuals.  This talk gives an overview of statistical
methods used to analyse data of this type.  Two applications
illustrate the overall theme.  The first looks at identification of
biological markers to detect departure from the cell cycle that
governs the renewal of healthy cells, with a view to improving
early-stage cancer diagnosis.  The second investigates the variation
in properties of lymph cells in response to the presence of uterine
cancer, which provides insights into the sequelae and metastasis of
this cancer type.  Joint models for analysing a continuous outcome
(cell size) and a cluster size variable (number of cells), when these
variables are highly correlated, are also discussed in relation to
this second application.

***
4.30pm: Tea

***
4.45pm: Peter Kimani: Conditionally unbiased estimation in adaptive
seamless designs.

In order to accelerate development of new drugs, adaptive seamless
designs have been proposed in clinical trials. Examples of trials that
use ASDs are the seamless phase II/III clinical trials. Unlike
traditional testing strategy that involve a separate phase II trial
and a separate phase III trial, seamless phase II/III clinical trials
combine phases II and III into a single trial with two stages. Stage 1
is used to answer phase II objectives such as treatment selection and
stage 2 is used for the confirmatory analysis which is a phase III
objective. Although these trials are attractive because the
confirmatory analysis includes phase II data from stage 1 and hence
possible saving on the number of patients required, they can pose
statistical challenges. The inference methods used for trials that do
not have an interim analysis and compare a single experimental
treatment to a control are no longer appropriate. In this talk, I will
consider estimating treatment difference after a seamless phase II/III
clinical trial. I will describe some methods that account for the
adaptive nature of seamless phase II/III clinical trials.

***
5.30pm: Wine reception

As always with the best wishes,
Oliver Ratmann
on behalf of the Young Statisticians of the Royal Statistical Society