Verification of lab data is outlined is CFR section 21 CFR 211.194(a)(8) which states: “The initials or signature of a secondary data reviewer shows that the original records have been reviewed for accuracy, completeness, and compliance with established standards.” But what is “original data” for an electronic system like an HPLC? According to FDA Guidance Document: Data Integrity and Compliance with Drug CGMP, chromatographic systems generate dynamic records. These are records that allow an interaction between the data reviewer and the record content. Because these records can be altered after gathering the data (i.e. change the baseline or reprocess the chromatographic data to make the peaks appear larger or smaller), it is not acceptable to only review the paper print outs if the data is generated to support a GMP requirement.
A company recently received the following observation:
…. Appropriate controls are not exercised over computers or related systems to assure that changes in master production and control records or other records are instituted only by authorized personnel.
Specifically, your data acquisition software and workstations, including the systems used for your chromatography instruments, have not been suitably qualified to ensure that the data generated for release and stability sample testing of drug products, including … and …, are reviewed, complete, and attributable. For example:
Your firm uses validated laboratory software, including the chromatography software …, to acquire and process release and stability testing data. As part of your firm’s review process of the data collected, your procedure SOP…, “Peer Review”, states “the peer reviewer will do a 100% check of all data and calculations” and as part of peer review check list it further states that the audit trail will be peer reviewed. However, your Associate Manager explained that the original raw data and audit trails within the chromatography system are not reviewed during the peer review of data, instead only a printout of the data is reviewed. The printout is considered a copy of the data, not the original data. The raw data is not electronically signed, locked, or routinely archived to ensure no additional changes can be made to the data. In addition, any conditioning sequences or potentially unreported data are not reviewed to ensure all the data generated for a particular project is complete.
To view more observations pertaining to second review, refer to the following GMP Trends® Issues: 1099, 1075, 1034, and 1029. To view more observations pertaining to data integrity, refer to the following GMP Trends® Issues: 1114, 1109, 1104, and 1100.
Stay in Compliance! Subscribe today!