First, we had a lively panel discussion about how GCP and GMP QA can work effectively together. The struggle is real - lots of GCP QA professionals chafe under a GMP-focused quality management system - but clear communication helps QA teams align on a harmonized approach or agree on separate sets of SOPs.

In another session, FDA inspectors reviewed proposed  scenarios and weighed in on whether they would cite the issue in an inspection finding. For example, what if a sponsor's SOPs state they follow GCP, but they don't conduct a risk assessment? What if a sponsor is using an unvalidated TMF? Is it a problem if a site fails to document that they provided a study subject with a copy of the ICF?  FDA panelists fell back on the "so what" standard - for example, an unvalidated TMF would result in a citation if documents were lost or poorly controlled, but if no ill effects were noted, "so what?" As one panelist dryly remarked, "We are reasonable people."

Computer system assurance, the subject of a new draft guidance from FDA, was discussed in a panel at the end of the day. This risk-based approach looks at the risks of individual "features" and allows more latitude for ad hoc and unscripted testing, but as the presenters noted, "Unscripted does not mean undocumented." The draft guidance gives useful examples of how unscripted testing could be documented.  Looking toward the future, experts predicted that the focus would shift from validation activities to continuous monitoring, with artificial intelligence-enabled surveillance that would identify failures before they happen.