STANDARD OPERATING PROCEDURE R&D SOP 605 Sample Size Calculation

Joint Arrangements for Research
STANDARD OPERATING PROCEDURE
R&D SOP 605
Sample Size Calculation
Version date
Effective date
Number of pages
Review date
Version 1.2
17/07/2014
18/07/2014
5
01/07/2016
Author
Role
Professor Lee Shepstone
Head of Statistics, Norwich CTU
Approved by
Role
Signature
Date
Rachael Mold
Head of NNUH UEA Joint Research
Rachael Mold
17th July 2014
Authorised by
Role
Signature
Date
Professor Marcus Flather
R&D Director
Marcus Flather
17th July 2014
R&D SOP 605 v1.2
Effective Date: 18/07/ 2014
Page 1 of 6
Joint Arrangements for Research
It is the responsibility of all users of this SOP to ensure that the correct version is being
used.
All staff should regularly check the NNUH R&D website for information relating to the
implementation of new or revised versions of SOPs. Staff must ensure that they are
adequately trained in the new procedure and must make sure that all copies of
superseded versions are promptly withdrawn from use.
The definitive versions of all Joint NNUH/UEA health care research SOPs appear
online. If you are reading this in printed form please check that the version number and
effective date is the most recent one as shown on the NNUH R&D website.
TABLE OF CONTENTS
1 ABBREVIATIONS
3
2 INTRODUCTION
3
3 SCOPE
3
4 PRODUCING SAMPLE SIZE CALCULATIONS
3
5 DOCUMENTATION OF SAMPLE SIZE CALCULATIONS
4
6 RELATED DOCUMENTS
5
7 LIST OF APPENDICES
5
Appendix 1 SOP Change Control and Review Form
R&D SOP 605 v1.2
Effective Date: 18/07/ 2014
6
Page 2 of 6
Joint Arrangements for Research
1
ABBREVIATIONS
CTIMP
CTU
GCP
ICH
NNUH
PI
R&D
SSC
REN
SOP
TMF
UEA
Clinical Trial of an Investigational Medicinal Product
Clinical Trials Unit
Good Clinical Practice
International Conference for Harmonisation
Norfolk and Norwich University Hospital
Principal Investigator
Research and Development
Sample size calculation
Research and Enterprise Services UEA
Standard Operating Procedure
Trial Master File
University of East Anglia
2 INTRODUCTION
This SOP defines the procedure for the production of Sample Size Calculations
(SSC) for clinical trials. Whilst the detailed procedure of this SOP relates to
clinical trials, the principles of this SOP should be adopted for non-clinical trials
healthcare research by the member(s) of the research team with responsibility for
statistical design and analysis.
3
SCOPE
This SOP applies to all healthcare research sponsored by NNUH or UEA which
falls within the scope of the Research Governance Framework (2nd edition
2005). Where additional legislation applies, for example the Medicines for
Human Use (Clinical Trials) Regulations 2004 (and amendments) or the Medical
Devices Regulations 2002, required procedures will be indicated. External
sponsors may require use of their own SOPs and this will be specified in site
agreements. It is the responsibility of the local PI to ensure that study specific
SOPs can be operated without conflict to this SOP and in accordance with all
organisational polices related to research.
4
PRODUCING SAMPLE SIZE CALCULATIONS
A sample size calculation (SSC) should be carried out by an appropriately
qualified individual, usually the nominated trial statistician, and using appropriate
software. It should be based upon information provided by the trial Chief
Investigator and other members of the trial team.
R&D SOP 605 v1.2
Effective Date: 18/07/ 2014
Page 3 of 6
Joint Arrangements for Research
SSCs should be considered at a very early stage of clinical trial development.
They are required for funding and governance applications and need to be
justified in detail. It is recognised that there are differing approaches to SSCs
depending on trial types and objectives. Generally, an SSC will be based around
a degree of precision in a parameter estimate, often involving the notion of
statistical Power and statistical significance (or, equivalently, Type II and Type I
errors). However, for some studies, such as feasibility or pilot studies, such an
approach will not be appropriate but the sample size still needs to be justified on
alternative grounds.
The basis of the SSC will often be the agreed primary efficacy end-point.
Typically, a minimal clinically relevant effect will need to be agreed upon amongst
the trial team. Ideally, there should be some justification that this is achievable.
An estimate of between group variation will typically be required. The source for
this needs to be documented. Statistical errors (Type I and Type II) to be
tolerated need to be stated and where ‘not typical’ should be justified. Typical
values would be a ‘two-sided’ Type I error of 5% and Type II error of 80% or
90%, preferably the latter.
5 DOCUMENTATION OF SAMPLE SIZE CALCULATIONS
It is recognised that there may be a degree of iteration and different versions of
SSCs should be documented (SOP 800 Documentation Management).
Documentation for SSCs should clearly state:
• the date of the calculation,
• the individual responsible,
• any software used,
• the general approach to the calculation,
• assumptions made,
• values included in the calculation
• where possible, references to supporting literature.
• Date and individual responsible for independent quality checks (when
appropriate see below)
Quality
When an application is being made for funding a quality check on the SCC
should be made prior to the submission for funding. If changes to the sample
size are made as a result of peer review during the funding application, an
additional quality check should be made prior to the submission for Ethical
approval.
R&D SOP 605 v1.2
Effective Date: 18/07/ 2014
Page 4 of 6
Joint Arrangements for Research
6 RELATED DOCUMENTS
SOP 800 Documentation Management
7 LIST OF APPENDICES
Appendix 1: Change Control, Revision and Review Sheet
R&D SOP 605 v1.2
Effective Date: 18/07/ 2014
Page 5 of 6
Joint Arrangements for Research
Appendix 1: Change Control, Revision and Review Sheet
CHANGE CONTROL, REVISION and REVIEW SHEET: SOP 605
Version No
1.1
Change Date
01/04/2014
Reason for Change
Extend time scale
1.2
18/07/2014
Minor amendments as part of
the bi-annual review including
the addition of statistical
significance.
Reviewer:
Designation :
Date:
A M Swart
Director, Norwich CTU
17.06.2014
Reviewer :
Designation :
Date:
Allan Clark
Senior Statistician,
Norwich CTU
17.06.2014
Reviewer :
Designation :
Date:
R&D SOP 605 v1.2
Effective Date: 18/07/ 2014
Page 6 of 6