Nottingham Clinical Trials Unit
The University of Nottingham
University of Nottingham
  • Print
  

Further Statistics for Clinical Trials

Overview

This three day intermediate course in practical statistics for clinical trials is intended for researchers who are, or will be, involved in analysing clinical trials data and are already familiar with descriptive statistics and simple methods for comparing means and proportions such as t-test and Chi-square tests.  The course introduces further methods to analyse randomised clinical trials data including regression methods for data with continuous, binary, or survival (time-to-event) outcomes.  It also provides an introduction to the statistical methods used for the design of clinical trials.  The course combines lectures with practical examples using Stata software.  Examples will be based on real clinical trials datasets, but the statistical methods learnt can be readily applied to other fields including medical, epidemiological, or health services research.  Key statistical components required when conducting randomised clinical trials will be discussed in accordance to the ICH E9 (Statistical Principles for Clinical Trials) guideline.  Delegates must bring a lap-top with Stata installed for the practical sessions.  We can provide temporary licenses for those who do not have Stata.

Prerequisites
Participants should have knowledge of basic statistical methods of at least the level achieved in Introduction to Statistics in Clinical Trials short course.  The practical session will be delivered in Stata though it is not compulsory to have prior knowledge of this software.
 

 

Learning Objectives

By the end of the course participants should:

  • Have knowledge of the principles of simple and multiple regression analyses for continuous, binary and time-to-event data outcomes.
  • Be able to perform regression analyses using Stata
  • Be able to report and interpret the output generated from Stata including the treatment effects, confidence intervals and P-values.
  • Be able to describe the concepts of interaction effects with respect to subgroup analyses
  • Be familiar with the key aspects to consider when planning the analysis
  • Explain selected statistical principles for clinical trials including design techniques to avoid bias such as blinding and randomisation, definition of analysis populations, methods of dealing with missing data, and considerations for multiple primary outcomes.
 

 

Course Outline

During the 3 days of the course, the following topics are covered:

     Selected statistical principles for clinical trials:

  • Blinding and randomisation
  • Choice of analysis population
  • Methods of dealing with missing data
  • Multiple primary outcome variables

    Planning the analysis

  • Statistical analysis section of the clinical trial protocol
  • Preparing statistical analysis plan
  • Dummy/shell tables and figures
  • Interim analysis and roles of data monitoring committee and reporting to DMC
  • Patient flow
  • Baseline characteristics
  • Effectiveness/efficacy evaluation based on predefined outcome(s)
  • Senstiviity analyses
  • Supplementary analyses
  • Analysis of secondary and safety outcomes

    Reporting of clinical trial results

    Analysing continuous outcomes

  • Comparing treatment groups

    Linear regression

  • Simple and multiple regression
  • Interpretation of model coefficients
  • Interactions in multiple regression applied to subgroup analysis
  • Model assumptions

    Analysing binary outcomes

  • Comparing treatment groups including risk differences, risk ratios and odds ratios

    Logistic regression

  • Simple and multiple regression
  • Interpretation of model coefficients
  • Interactions in multiple regression applied to subgroup analysis

    Analysing time-to-event outcomes (Survival analysis)

  • Introduction to survival analysis
  • Failure times and censoring times; hazard functions; survivor functions
  • Kaplan-Meier survival curves and Log-rank tests
  • Regression methods for survival data
 

 

Who is the course intended for?
Anyone intending to be involved in analysing clinical trials data and are already familiar with descriptive statistics and simple methods for comparing means and proportion.
 

 

Course structure
  • Course lectures with worked examples
  • Stata practical sessions
 

 

Course tutors
All sessions on the course will be led by experienced medical statisticians from Nottingham Clinical Trials Unit.
 

 

What does the course include?

Course materials will be provided in delegate packs.

Light refreshments will be provided in the morning before the course begins and for morning and afternoon breaks.  A buffet lunch will also be provided.

 

 

Terms and Conditions
In the event that you need to cancel your booking we ask that you give as much notice as possible.  Participants cancelling more than two weeks before the start of the course will be eligible for a full refund, after that time you will be charged the full course fee.  In the unlikely event that the short course has to be cancelled, our liability is limited to course fees only.  We recommend that delegates have adequate cover to reclaim any travel or accommodation expenses.  If you need to cancel, please email your request to ctu@nottingham.ac.uk.  For further information please contact ctu@nottingham.ac.uk or call 0115 8231600.
 

 

 

UKCRC CTU Logo

Nottingham Clinical Trials Unit

General enquiries
Email: ctu@nottingham.ac.uk
Telephone: +44 (0)115 8231600