Overview

This one day intermediate course delves into the statistical theory underpinning sample size calculations within the hypothesis testing framework

It equips participants with practical insights for determining sample sizes across a diverse array of trial designs and outcome distributions.

Topics covered include factorial, non-inferiority, group sequential, and clustered randomized clinical trials. 

Real-world examples illustrate sample size calculations using industry-standard software tools such as Stata, SAS, and R.

Each lecture will include a practical session to consolidate learning the concepts covered.

Who this course is for

This course will be of interest to those working on trials and study designs, which includes trialists, trial statisticians, clinicians and other scientists. 

Familiarity with the hypothesis testing framework is essential (i.e., type I error rate, power, and various effect sizes).

Familiarity with sample size calculation for simple/basic designs is desirable but not essential.

Course content

The day will be structured as follows:

• 09:45 -10:00 – Introduction  
• 10:00 -11:15 – Lecture 1: Introduction to statistical theory for sample size calculations: continuous, binary, and ordinal categorical outcomes (Professor Ian White)
• 11:15 -11:30 – Break
• 11:30 -12:30 – Lecture 2: Designs with time-to-event outcomes (Dr Babak Choodari-Oskooei)
• 12:30 -13:30 – Lunch break  
• 13:30 -14:30 – Lecture 3: Factorial and cluster randomised trial designs (Professors Ian White & Andrew Copas)
• 14:30 -14:45 – Break
• 14:45 -15:45 – Lecture 4: Designs with interim stopping looks: Groups sequential designs (Dr B. Choodari-Oskooei)
• 15:45 -16:00 – Break
• 16:00 -17:00 – Lecture 5: Dealing with uncertainty in design parameters (Dr Matteo Quartagno)

Learning outcomes

By the end of the course you should be able to:

• Derive standard sample size formula
• Perform standard sample size calculations for trials with continuous, binary, ordinal categorical and time-to-event outcomes.
• Project the power and events of trials with a time-to-event outcome into the future.
• Perform sample size calculations for group sequential, factorial, cluster randomised and noninferiority trials, and how to account for uncertainty in the design parameters.
• Plan simulation to determine sample size and understand advantages and disadvantages of doing sample size calculations by simulations.
• Display an understanding of the underlying assumptions for sample size calculations and the impact of violation of those assumptions to said calculations, along with strategies to mitigate associated issues.
• Recall an array of sample size re-estimation approaches

Course team

Dr Babak Choodari-Oskooei

Prof Ian White

Prof Andrew Copas

Dr Matteo Quartagno

Course information last modified: 22 Apr 2024, 11:45