By definition, data integrity is the “generation, transformation, maintenance and assurance of the accuracy, completeness and consistency of data over its entire life cycle to be in compliance with applicable regulations.”
It is important to understand the fact that data integrity has a wide spectrum of applications therefore, this article will be limited to requirements of the pharmaceutical Quality Control Laboratory environment. Data integrity in the quality control laboratory has become a widespread global topic, although integrity of data is inherently essential to the pharmaceutical quality system.
The negative impact such as, the company’s reputation, the high cost of mass recall, and above all, the effect on the patients who do not have access to safe medication, shows that data integrity is crucial for the accuracy and efficiency of business processes. The approach of the pharmaceutical companies towards data integrity should be equally focused on paper-based and electronic data generation, management and storage.
In February of this year, Health Canada issued a letter to Drug Established License Holders, to emphasize their responsibilities and obligations with respect to Good Manufacturing Practice, including the high importance of data integrity.
Health Canada and US-FDA have assessed data integrity issues in numerous pharmaceutical manufacturers’ facilities domestically and globally.
This is a categorical summary of some of the warning letters issued by North American regulatory institutions:
- “Failure to maintain complete data derived from all lab tests conducted to ensure compliance with established specifications and standards”
- “Failure to record and justify any deviations from required lab control mechanisms”
- “Failure to investigate and document out-of-specification results”
- “Failure to record activities at the time they are performed”
- “Failure to validate analytical methods used to test APIs”
- “Failure to exercise appropriate controls over computer or related systems to assure that only authorized personnel institute changes in master production and control records, or other records”
Ultimately, failure to make the right decisions at every level.
Among other observations, US-FDA and Health Canada also found the following major issues:
- Not reporting failing results
- Conducting unofficial analysis
- Trial analysis
- Re analyzing failing results until passing results are obtained
- Deleting electronic data
- Back Dating/Postdating/Missing signature
- Disabling audit trails in electronic data capture system
- Finding mismatch between reported data and actual data
- Fabricating training data
- Having unofficial analytical batch sheets and analytical reports
- Releasing failing products
Taking into consideration the above mentioned facts, there is a question for which many pharmaceutical manufacturers are looking for answers.
What can be done to recover the products that are already affected by breaches of data integrity?
A few solutions on how to verify the historical data and related processes are presented below. Data integrity categories are organized in a cause and effect diagram which includes seven major data integrity issues.
The approach however, should be executed in accordance to each pharmaceutical companies based on individual needs.
This particular structure helps to visualize each area where data integrity might be a problem.
From the above diagram, we take the “Laboratory Documentation and Procedures” scenario as an example and present different strategies.
- Review the SOP(s) and Work Instruction(s) linked to documentation
- Review analytical notebooks
- Examine all records for accuracy and authenticity
- Document the raw data throughout the project
- Note if raw data is missing or if the records have been rewritten
- Paper-based raw data should be documented only in books or analytical sheets for which there is accountability (e.g. pre-number pages, binder type notebook)
- Verify all the dates to reflect time of actual occurrence and to be in synchronism with the computer data base
- Review mobile phase preparation, standards and sample preparation (weights, dilutions, calculations), conditions of the instruments
- Cross reference the data done in accordance with GMP
- All cross outs must have explanation, initials and current date as per GMP
- Examine chromatograms and integrity of the data in Empower and LIMS
- Review the integration parameters in Empower, etc
The above list is just an example of how to analyze, to review and to address data integrity issues in accordance to the existing regulations. Data integrity has become a category on its own in the eyes of regulatory authorities reflecting its importance.
As a global effort, continuous assessment of data integrity through internal audits, external review programs, acceptance of a third-party data integrity audit are important to evaluate and to reinforce the ongoing compliance and the need for change.