Biostatistics Industry and Regulator Forum
Session 8: Risk-Based Monitoring
- Vice President and Head Biometrics & Information Sciences
- Astrazeneca Pharmaceuticals, United States
- FDA, United States
The recent ICH E6 now includes explicit recognition that centralized monitoring can/should be part of the quality system for trials with the aim of identifying missing data, inconsistencies, outliers, and a lack of variability. The use of statistics is encouraged to identify trends, looking at the range and consistency of data across sites. The current approach of 100% source data verification would thus be replaced by these newer contemporary techniques.
While the principles are clear, implementation requires careful consideration and expert trialists have different views. The session will explore some of the key challenges in fully embedding risk based monitoring. Key topics include:
- Should we monitor primary endpoints directly, indirectly, or not at all?
- Should general rates of AEs by site or country be monitored and evaluated for unusual patterns?
- Should we concentrate of operational dashboards for completeness of data, aggregate data at the patient level, or site/regional patterns?
- What is the role of statistical pattern detection?