FDA Findings on Data Integrity Decreased
According to a RAPS article, FDA inspections have become more remote and data integrity citations have decreased.
This article summarizes a speech given by FDA consumer safety officer Maan Abduldayem at the 46th International GMP Conference.
Trends in Data Integrity Findings
- FY2021：Noted in 5 Warning Letters (W/L)
- FY2020：Pointed out at 24 W/L
- FY2019：Pointed out at 32 W/L
After all, remote inspections seem to have limitations.
Once the corona disaster is contained, the FDA will resume on-site inspections.
If data integrity cannot be verified, patient safety and product quality
For example, if data falsification, whether intentional or negligent, occurs during quality testing in a QC lab patient safety will be at stake.
In other words, there is a risk of shipping poor quality products that should not have been shipped. It is important to note here that while intentional falsification is a must, unreliability of data due to negligence (human error) is also a problem.
Whether intentional or negligent The reason for this is that both patient safety is affected when data is unreliable. In general, negligence-based data integrity lapses account for 80% of all data integrity lapses. Specifically, these are transcription errors, calculation errors, and manipulation errors.
In addition, intentional does not necessarily mean dishonest, and a person may change raw data, etc., believing it to be correct.
These can be caused by misunderstandings, misconceptions, memory errors, insufficient understanding of SOPs, and ambiguous SOPs.
On the other hand, product quality is threatened by inadequate manufacturing and validation records
Thus, if data integrity cannot be verified, the FDA will not be able to confirm patient safety and product quality and will not be able to approve shipments.
In the case of a Pre-Market Inspection (PAI), if data integrity cannot be confirmed, the FDA cannot determine whether to accept the submitted materials, which will affect the approval/denial decision within the time limit.
]]>
Comment