A special challenge in ongoing aggregate evaluation of safety data is application of appropriate statistical techniques that account for clinical considerations, as opposed to strict statistical inference. Exploration and estimation of event rates involve substantial knowledge of quantifying risk and measuring uncertainty, and a proper meta-analytic review should be a routine part of the process. However, this quantitative thinking needs to be combined with a dynamic collaborative process for engaging with scientists from clinical disciplines. A systematic approach to evaluate and assess safety data, guided by medical judgment, would enable investigators to develop clinical as well as statistical understanding of the safety profile. This requires a thorough understanding of existing safety data and interpretational emphasis should be on medical judgment.
This philosophy is captured in the spirit of the FDA IND Safety Reporting final rule:
The important thing is to have a thoughtful process; a system in place to look for clinically important imbalances, applying the best clinical and quantitative judgment, while maintaining trial integrity.
– Jacqueline Corrigan-Curay (2018)
Safety monitoring during clinical development requires a partnership between clinical and statistical scientists. To review accumulating program-level safety information across the drug development continuum, sponsors have implemented product-specific, multidisiplinary safety management teams (SMTs). To leverage the scientific expertise and medical judgment of these teams, we need:
- A multidisciplinary approach
- Assessments that are product-specific
- Quantitative frameworks to measure level of evidence
- Decisions that are driven by medical judgment
Emphasis shifts from analyses with statistical testing and confirming to assessments with learning and decision-making: medical judgment within a quantitative framework.