Principles of data integrity in pharmaceutical industry

Principles of data integrity in pharmaceutical industry.

Data integrity is the protection of data from unauthorized and unaccountable changes. ALCOA is the concept to implement data security and integrity in pharmaceutical industries.

Data integrity is highly important in any industry, but especially so in the pharmaceutical industry where any data error could mean serious consequences. Data integrity is defined as maintenance and assurance of data consistency and accuracy throughout its life-cycle. Keeping data consistent (unchanged from the very start to the end) is also a matter of data security and even though data integrity and data security overlap in their functions, they shouldn't be mistaken for one another.

Principles of data security and integrity had to be standardized in order to regulate them and achieve better processes and higher quality products. This is how the ALCOA principle came to be. 

According to the ALCOA principle, the data should have the following five qualities to maintain data integrity: Attributable, Legible, Contemporaneous, Original and Accurate.

1. Attributable

Each piece of data should be attributed to the person who generated it. This part should include the details of the person who performed the action and when it was performed (a timestamp). This can be done both physically (signing, putting initials and dating a paper document) or electronically (through a digital system). Good documenting practice (GDP) recommends having a signature or alias log so it could be easily determined who changed or recorded new data.

2. Legible

All recorded data should be readable (legible) and permanent. The readable part is fairly obvious - the data will be used multiple times by different people and if only one person can read the actual records then the data is more or less unusable. Permanent means that the data won't be changed accidentally or unwillingly. For the data to be legible GDP suggests using pens with ink which can't be erased, as well as having enough space for the data in the documents and forms.

3. Contemporaneous

This means that the data is always recorded at the actual time the action or work was performed. No piece of data should be recorded retrospectively. Data credibility depends on whether all date and timestamps fall in order because if they don't the data is considered unreliable and should be scrapped. The general advice is to make sure times in all labs are synchronized, or even have a central clock system with which all other computers could synchronize.

4. Original

It is very important to have a medium where the data was first recorded. This could be a form or a protocol, a dedicated notebook or a database, doesn't really matter as long as it is preserved in its original form. Having a standardized recording procedure solves a lot of problems related to the originality of the data.

5. Accurate

Achieving data accuracy means ensuring that the data is error-free, complete, truthful and that it reflects the observations made. Editing data without logging means its accuracy is lost, so it is of vital importance to always record who, when and why changed the data record. When it comes to accuracy, it should be held to a high standard. Witness checking is a technique used when recording critical data to ensure its accuracy. Incorporating accuracy checks inside the electronic system (if there is one) is also a good thing to do.

Data security and integrity should be perceived as a process rather than a one-time factor. Data errors can seriously affect both small and big companies. That is why it is important to implement the ALCOA principle and make the data infrastructure an asset, instead of it being a liability.