Calculations based on PM/pivotal trial data
-
Is it acceptable to include data in an APS that had to be calculated based on PM/pivotal trial data?
e.g., inclusion of 95% CI error bars in a chart that were calculated using the PM-provided SE and sample mean, or calculating reduction/increase in a primary endpoint from baseline to timepoint X when only the mean change at timepoint X and baseline values are in the PM.
Thanks in advance!
-
Hey @megancouture
In the context of a general question, there are cases where company-generated calculations may or may not be acceptable. For the specific examples presented, a review of the study design/statistical analyses/results would be considered in the assessment. We recommend considering the following Code sections:
-
Code section 5.8 - Methodologies, endpoints and independent review. To be considered as evidence, clinical studies must use established research methodologies and validated endpoints. To aid in the assessment of these study parameters, PAAB looks for evidence that the full study results have been subject to independent review, such as that found by achieving the publishing of study results, including statistical analyses, in a peer-reviewed journal
-
For the second example, we’d also advise to consider Code section 4.2.3 (principles of relative risk reduction [RRR] and absolute risk reduction [ARR]) and Code section 5.9 (statistical significance required for evidence).
-
