On September 21, the Engelberg Center for Health Care Reform hosted a one-day expert workshop to discuss the most effective and efficient ways to carry out the signal refinement step of active medical product surveillance. For the purposes of the meeting, signal refinement was defined as the step of active surveillance after signal generation, when a potential association between a medical product and health outcome is identified, and before signal evaluation, where formal epidemiological analyses are implemented. Potential steps in signal refinement are shown in the diagram below.
Experts from academia, the private sector, and FDA discussed several topics including:
- Approaches to building a generalized framework for signal refinement applicable to a broad range of medical products
- Acceptable levels of uncertainty in signal refinement
- Data needs for signal refinement
- Methodological needs and approaches for signal refinement
Two hypothetical scenarios helped to guide discussion about signal refinement data and methodological needs:
- Association between an oral anti-diabetes drug and acute myocardial infarction
- Association between an injectable antibiotic drug (administered in an inpatient setting) and acute liver injury