Module 14 2022

21/03/2022

Statistical Analysis Plan

• A technical document that describes in detail the planned statistical analysis of a clinical investigation as outlined in the clinical investigation plan (CIP) • Consider timing for the SAP • Justification for control group and, if applicable, placebo/sham • Look out for MULTIPLICITY if multiple tests are planned in a trial (e.g. multiple endpoints, multiple study arms, subgroup comparisons, interim analyses)

The Organisation for Professionals in Regulatory Affairs

Statistical Considerations in Medical Device Clinical Investigations

19

Statistical Analysis Plan

• QUIZ QUESTION: When should the statistical methods used for analysis be determined? • The statistical methods need to be pre-specified in the final SAP • Statistical methods cannot be determined in advance as you need to look at the data to identify specific requirements • As many statistical methods as possible are applied to the data and the final method(s) are selected based on the results for the final report • QUIZ QUESTION: Interim analyses in clinical trials • Should never be done as they lead to biased results • Can be performed any time you like as long as you have access to the data • Can be planned for, but need to be pre-specified in detail in the CIP

The Organisation for Professionals in Regulatory Affairs

Statistical Considerations in Medical Device Clinical Investigations

20

10