GDP: Are Inspectors looking for Data Integrity?
Data Integrity and the governance of all data is still a hot topic in the GMP area. Many observations made in GMP inspections are linked to data integrity problems. But what about GDP? In a recent blog published by the U.K. Medicines and Healthcare Products Regulatory Agency (MHRA), some examples are given from GDP inspections.
When reviewing data, Terry Madigan, who has been GDP Inspector since 2009, asks himself ''does this data matter (i.e. is it critical) and if so, can I trust what I see?''
Sometimes data integrity problems are easy to spot, for example when a thermometer has not been calibrated and staff does know how to use it. So sometimes, the lack of quality of data might be due to poor training or to the use of poor equipment.
More difficult are cases of intended falsification of data. This is often supported by non robust or poorly defined processes. Better written procedures, recording and record review could already help.
MHRA encourages stakeholders to use computerised systems providing they are "appropriately qualified and controlled". Non-trackable software (like WhatsApp) for quality-critical event reporting is not an option (this has been proposed by some wholesalers). Also, "the use of robots and electronic audit trails can lead to a false sense of security in relation to data integrity", especially when "work-arounds are created".
Record Conversion
When paper records are converted in electronic formats or documents into pdf files, "these should be confirmed as being accurate and complete by a person of appropriate seniority". In some cases, both versions might need to be retained.
The use of Excel®
Spreadsheets can be of great help for data management and its summary and presentation. But the "data governance measures need to be rigorous". Special attention should be paid to number-rounding, converting of units for creating graphs. Other problems may occur with multiple users, overwriting of data, version control and wrong calculations and macros.