Your Easy Guide to Clinical Studies

Upon study conduct, various analysis methods can be used to evaluate and interpret the study data.

 

The choice of the analysis method(s) depends on the study design, research objective(s), and the nature of the data (e.g. quantitative and numerical or qualitative and descriptive)

 

The analysis method(s) selected for assessing each study outcome, must be outlined in the study protocol, and, if applicable, in a Statistical Analysis Plan (SAP).

 

The adherence to pre-defined analyses methods is important as it ensures transparency, thus allowing other researchers to:

• Retrace and understand how study results were obtained
• Reproduce the analysis

 

Pre-defined analyses methods increase the credibility of study results. It avoids the implementation of post hoc decisions (i.e. decision that are made after seeing the data), thereby preventing:

• The selective reporting of favourable results

A (conscious or unconscious) influence by the SP-INV on the data analysis in order to produce desirable outcomes

As a SP-INV, discuss and define together with a statistician the most appropriate analysis methods needed to assess study outcome(s)/endpoint(s).

 

Note:

Apart from outlining the analysis method in the study protocol, a Statistical Analysis Plan (SAP) is usually suggested in more complex studies. These studies require more in depth explanations regarding the statistical aspects of a study (e.g. multi-centre studies, placebo and comparator studies, blinded studies, the inclusion of interim analyses)

References

ICH Topic E9 – see in particular

• 5.1 Prespecification of the analysis