Meta-analysis in Drug Development and Health Technology Assessment

A course by Stephen Senn

 

 

In this two-day course Stephen Senn will cover the theory and practice of meta-analysis, stressing understanding rather than algebra and covering the practicalities of carrying out such analyses, using case studies where appropriate.

 

The course will be invaluable for all those involved in meta-analysis of pharmaceuticals, whether as consumer or producer, or working for sponsor, regulator, contract research organisation, reimburser or academia.

 

About the presenter

 

Prof Stephen Senn is a well-known medical statistician who has worked in and consults regularly for the pharmaceutical industry. He is the author of over 250 publications including the books Cross-over Trials in Clinical Research, Statistical Issues in Drug Development and Dicing with Death and several papers dealing with the theory and practice of meta-analysis.

 

Topics covered include

 

The practicalities of carrying out this

kind of analysis

-       Different models for analysis

-       When and when not to pool

-       The respective roles of fixed and

-       random effects analyses

Parallels to the analysis of multi-centre

trials

-       The many modes of meta

-       Ordinary least squares versus meta-analysis

-       weighting

-       Handling interactions

-       Type II and Type III attitudes

 

Differences in practice between 'public'

organisations and commercial

organisations such as pharmaceutical

sponsors.

-       The Cochrane Collaboration vs

-       pharmaceutical sponsors

-       Publication bias vs appropriate

-       analysis

 

Graphical methods

-       Scatter and bubble plots

-       Forest plots

-       Galbraith plots

-       Funnel plots

Common pitfalls and issues

-       Double counting

-       Different trial designs

-       Baseline risk

 

Case studies

-       Screening for breast cancer

-       Rofecoxib and heart disease.

-       Formoterol and asthma

Network meta-analysis

-       Basic idea

-       Fixed effects

-       Random effects

-       Mean-mean plots

-       Issues in interpretation

 

 

Typical Timetable for a course

 

Day 1

 

10:00-11:00

1. Introduction to Meta-Analysis

11:00-11:30

Coffee

11:30-11:45

Q&A session

11:45-12:45

2. Some History

12:45-14:00

Lunch

14:00-15:00

3. Fixed effects approaches

15:00-15:15

Q&A session

15:15-15:30

Exercise I

15:30-16:00

Tea

16:00-17:00

4. Regulatory aspects

17:00-17:30

Problems, questions, discussions

 

 

Day2

 

09:00-10:00

5. Graphs and Plots

10:00-10:15

Software

10:15-10:45

Coffee

10:45-11:45

6. Random effects approaches

11:45-12:00

Q&A session

12:00-13:30

Lunch

13:30-14:30

7. Case-studies (Including Exercise II)

14:30-14:45

Break

14:45-15:45

8. Network meta-analysis

15:45-16:00

Discussion

 

Return to StatWise courses