Development↦Statistic Methodology↦Statistics in the Protocol↦Data-Analysis-Set
What is it? Why is it important?
A Data-Analysis-Set defines which study participants to include in an analysis.
Commonly used Data-Analysis-Sets are the:
- Full-Analysis-Set (FAS): Includes all randomized study participants, regardless of study protocol adherence or deviations
- Per-Protocol-Set (PPS): Includes only participants who did strictly adhere to a predefined set of protocol requirements (i.e. as defined by the SP-INV)
The FAS aims at an Intention-To-Treat (ITT) analysis. It includes all participants randomly assigned to a group (e.g. intervention or control group) irrespective of their adherence to the treatment protocol. It estimates the effect of being assigned to the treatment versus the control group.
The PPS assesses the performance of an intervention among participants who did closely follow the study protocol. While a PPS analysis can provide valuable insights into the performance of an intervention, one should keep in mind that analysis results may be biased (more information under More).
More
Biased results
In a Randomized Controlled Trial (RCT), the random allocation of participants to different treatment groups ensures that the two groups are comparable. When selecting a subset of randomized participants for the analysis, similarities between treatment groups can no longer be guaranteed. As a consequence, a bias may occur.
Example: In an active treatment vs. placebo study, all participants with an early treatment discontinuation are removed from the PPS. This may exclude participants with severe side effects in the active treatment group. Consequently, the intervention group may look better than it actually is.
What do I need to do?
As a SP-INV, define applicable analysis-set(s) for you study.
- FAS: As all study participants are included its definition is usually straightforward
- PPS: Only include participants who have no major protocol deviations. Define events considered as major protocol deviations (e.g. not meeting specific eligibility criteria, not completed study visits, or visits done outside the allowed timeframe, non-compliance with study treatment)
Document criteria used to define Data-Analysis-Sets in the study protocol and, as applicable, in the Statistical Analysis Plan (SAP).
As required, alternative analysis-sets can be defined. Example: A safety analysis set may be defined as all participants who received at least one dose of the study drug. Analyses of safety outcomes would then be performed on this analysis set.
Where can I get help?
Your local Research Support Centre↧ can assist you with experienced staff regarding this topic
Basel, Departement Klinische Forschung (DKF), dkf.unibas.ch
Lugano, Clinical Trials Unit (CTU-EOC), ctueoc.ch
Bern, Department of Clinical Research (DCR), dcr.unibe.ch
Geneva, Clinical Research Center (CRC), crc.hug.ch
Lausanne, Clinical Research Center (CRC), chuv.ch
St. Gallen, Clinical Trials Unit (CTU), h-och.ch
Zürich, Clinical Trials Center (CTC), usz.ch
Video
References
ICH GCP E6 (R3) -see in particular:
- 3.16.2 Statistical Programming and Data Analysis
- 4.2.6c Data sets prior to analysis
ICH Topic E9 – see in particular
- 5.2 Analysis sets
Swiss Law
ClinO – see in particular article
- Art. 2b Definition intervention
ClinO-MD – see in particular article
- Art. 2a Definition of clinical intervention
- Art. 2a Definition of performance study
HRO – see in particular article
- Art. 3a Definition of research