ALCOA principles: ‘healthy’ guidelines for data integrity - 5 -

Thumbs up for all of you who are still attending our comprehensive, five-part masterclass on data integrity! You have shown the perseverance it actually takes to incorporate a solid approach to data integrity across all levels of your organisation. To wrap it all up, we will now address the final principle of ‘accurate’.

The term accurate means data are correct, truthful, complete, valid and reliable. Basically, data integrity it is all about establishing transparency throughout all processes involved in lab environments. As such, the fifth and final requirement has a lot in common with several other ALCOA principles. Assuring the accuracy of data in paper records and electronic records revolves around three elements: qualification, validation and calibration.

- Equipment, such as balances, pH meters and thermometers that generate printouts, should be qualified (fit for its purpose), calibrated and maintained on a regular basis.
In the pharmaceutical industry, for example, weight and length data are measured by the microgram and micron. Here, even the slightest deviation may have severe consequences.
Staff too needs to be fully qualified and authorised for the job. Education and training certificates should therefore be recorded and continuously updated.

- Computerised systems that generate, process, maintain, distribute or archive electronic records (as well as all related work processes) need to be validated.
For instance, raw data from lab instruments that are entered into a computer are often used to make all kinds of calculations and analyses. To ensure that these are accurate, software should be tested thoroughly to prove that it is fit for its purpose. A risk-based approach may be used here. In case software was developed in-house by the pharmaceutical company, the risk of inaccuracy is considered relatively high.

If the software in question was purchased from a supplier (aka Commercial Off The Shelf software or COTS software) and is used by a multitude of professional users, the risk of inaccuracy is considered relatively low. Nonetheless, software audits remain mandatory and are generally paid for by the pharmaceutical company. The supplier in turn must allow for audits to be executed.

- The validation requirement also applies to analytical methods, production processes and systems that transmit data between/among computerised systems.

Four eyes principle

The entry of critical data into a computer by an authorised person requires an additional check (often referred to as the ‘four eyes principle’) on the accuracy of the data entered manually. This often, but certainly not exclusively, applies to production batches. Before releasing a production batch, a range of analyses is carried out on samples that are entered into a computer system. The reviewer should particularly focus on changes in the captured and processed data. Secure Desktop software or a similar solution implemented on the capture systems can be really helpful here as it only allows users to execute certain predefined actions.

Systematic implementation of all requirements discussed in our blogs will result in a high degree of ALCOA compliance. We hope you enjoyed our masterclass and please know that you can always rely on Vivenics for further advice and on our data integrity ‘tool box’ for quick fixes. Class dismissed!