Issues with data integrity have always existed. With recent egregious cases, the FDA has been taking greater measures to uncover these attempts to misrepresent key data. There have been stories of inspectors looking through the contents of garbage cans to find discarded documentation. Additionally, they are receiving specific training to better identify faulty data.
In its simplest form, reliable data over the life of the product is:
- Complete
- Consistent
- Accurate
The expectation in the regulations required and what investigators will look for evidence of:
- Qualified instruments
- Validated software
- Verified calculations
- Data back up
- Correctly prepped reagents
- Verified methods and approved documentation
An Increasing Problem
Investigators are finding increasing problems with data integrity both in the US and from foreign manufacturers. Some of these issues are unintentional while others are outright fraud. These may include:
- Human error when entering data
- Inadequate training
- Choosing passing results or utilizing multiple analyses
- Resampling
- Changing data
- Hardware/software malfunctions
- Backdating or recording a step before it has occurred
- Creating data
- Using results from previous lots
- Conducting unofficial analysis
- Creating unofficial batch records
- Accepting on a C of A without knowing the reliability of the vendor or never conducting identity tests.
Too Good to be True
The old saying is applicable here. If it is too good to be true, it probably isn’t. If your results are always in specification or you always have zero micro recovery, then it probably isn’t the case. These are obvious and will most likely be flagged. Manipulation of software is more difficult to detect and in some cases may require a higher level of sophistication to execute but it does happen.
Regret Cutting Corners
A culture of quality allows employees to acknowledge mistakes and then provides an acceptable means to document and correct them. Becoming a company known for reputable data and building trust with the agency should be a goal of all companies. It can be difficult to make the claim that an issue was isolated and not part of a systemic practice and affects the overall relationship with them. Additionally, it may result in fines or lost revenue when a submission is questioned or approval delayed due to sketchy data.
The FDA anticipates that mistakes will be made. How firms handle those mistakes is what they look for. Adequate controls and appropriate corrective action goes a long way to building and in some cases, repairing the firm’s reputation.
