Follow
Data integrity is crucial in any industry, but especially in the pharmaceutical industry where any data error can have serious consequences. Data integrity is defined as the maintenance and assurance of data consistency and accuracy throughout its life-cycle. Keeping data consistent (unaltered from start to finish) is a matter of data protection and although data integrity and data protection overlap in their functions, they should not be mistaken for each other.
Data protection and integrity principles had to be regulated and standardized to achieve better processes and higher quality products. This is how the ALCOA policy came about. According to ALCOA principles, data should have the following five qualities to maintain data integrity: characteristic, readable, contemporaneous, original and accurate.
1. Attributable
Each piece of data should be attributed to the person who created it. This part should include details of the person who performed the action and when it was performed (a timestamp). This can be done both physically (signing, initialing and dating a paper document) or electronically (via a digital system). Good documentation practice (abbreviation: GDP) recommends keeping a signature or alias log so that it is easy to determine who changed or recorded new data.
2. Legible
All recorded data should be readable (text) and permanent. The readability part is fairly obvious – the data will be used multiple times by different people, and if only one person can read the actual records, the data is more or less unused. Persistent means that the data will not be changed accidentally or unintentionally. To read data, GDP recommends using pens with ink that cannot be erased, as well as documents and forms with sufficient space for data.
3. Contemporaneous
This means that data is always recorded at the actual time the action or task was performed. No part of the information should be recorded retrospectively. Data reliability depends on whether all dates and timestamps read in sequence because if they don’t, the data is considered unreliable and should be discarded.
General advice is to make sure all lab times are synchronized, or even have a central clock system that all other computers can synchronize to.
4. Original
It is very important to have a medium where the data was first recorded. It may be a form or a protocol, a dedicated notebook or a database, it doesn’t matter as long as it is preserved in its original form. Having a standardized recording method solves many problems related to data originality.
5. Accurate
Achieving data accuracy means ensuring that the data are error-free, complete, true and reflect the observations made. Editing data without logging means its accuracy is lost, so it is crucial to always record who, when and why the data record was changed. When it comes to accuracy, it should hold a high standard. Witness checking is a technique used to ensure the accuracy of important data while recording it. Incorporating accuracy checks inside the electronic system (if any) is a good thing.
Data security and integrity should be considered a process rather than a one-time factor. Data errors can seriously affect both small and large organizations. This is why it is important to implement the ALCOA principle and make data infrastructure an asset rather than a liability.
