OR WAIT 15 SECS
Don DiPalma is director of Quality and Technical Services at LabVantage, firstname.lastname@example.org.
Jeff Vannest is senior director of Product Management at LabVantage.
Integrated computerized systems for data collection improve data security and offer a solution for handling temporary data.
The data that pharmaceutical manufacturers rely on to make quality decisions come from electronic systems, hybrid practices that incorporate both paper and software, or paper records alone. But the pressure is on, from FDA and regulatory agencies worldwide, to ensure the integrity of data supporting the product lifecycle, from drug discovery to product manufacture and, ultimately, release. The World Health Organization, FDA, and the UK Medicines and Healthcare products Regulatory Agency (MHRA) have each issued guidance in the past few years to improve data integrity in pharmaceutical manufacturing.
FDA, in its 2016 Data Integrity and Compliance with CGMP Guidance for Industry (1), noted it was seeing increasing violations of current good manufacturing practices (cGMP) stemming from data integrity infractions. Some of those violations were the result of an inability to audit temporary data that may have been entered, but not saved, into an electronic system. Within the industry, the expectation that data are attributable, legible, contemporaneous, original, and accurate (ALCOA) has been expanded to what is called ALCOA+. This new definition demands that the data also be complete, consistent, enduring, and available.
These “plus” concepts help give a datum context, especially during future reviews and audits. In other words, ALCOA+ takes into account not just the data, and associated metadata, that exist at the time of collection, it also shows a complete picture of the data throughout its life to ensure data integrity. It is no longer appropriate to just demonstrate that data were collected and verified; data must be shown to remain unadulterated, accurate, and attainable. This demonstration can only be done when there are controls around the data to prevent alteration, or measures in place to identify when data have been changed. It is noted that changing data is acceptable for appropriate reasons, such as the wrong entry (i.e., typing mistake) of a test result. It is not acceptable when data are changed just to meet a specification.
Electronic data capture offers significant productivity and security enhancements over paper-based records, but also introduces new challenges for information technology departments, users, and regulators. Challenges include preservation of data, technical aspects of system integration and validation, and acceptance of electronic systems, which require a higher level of technical capability.
For confidence that data meet ALCOA+ standards, pharmaceutical companies and contract manufacturers are increasingly relying on integrated software solutions. Integrating the computerized systems that capture, collect, and archive data is one way to reduce the manual re-keying of data, thus reducing risk and improving integrity. Single data entry, coupled with the automatic transfer of that data across integrated systems, eliminates the possibility of human transcription errors and increases efficiency. Most significantly, it decreases lag time for manufacturing execution systems (MES) and production lines to receive data as they become available in a laboratory information management system (LIMS).
Without confidence in the integrity of data, such as quality and batch release data, there can be no confidence in the integrity of the drugs themselves. In today’s manufacturing environment, paper-based and disconnected data systems don’t have any way to ensure that data are accurate and consistent--that they have integrity--other than by involving extra steps of human review, which is not ideal. Reliance on human intervention increases risk, which can lead to unusable product, product recalls, and lost revenue associated with recalls; warnings, fines, or consent decrees; negative publicity; and most important, possible harm to trial subjects and patients.
Therefore, various pharmaceutical companies are adopting strategies that reduce reliance on human/manual systems and deploying computerized systems instead. Although there is no single software solution for the entire pharmaceutical process, there are integrated systems and platforms that cover specific portions of pharmaceutical processes and that can be interconnected.
Integrated systems reduce risks and enhance product quality by using automated and systematic checks to verify data integrity. For example, solutions that integrate a LIMS, electronic laboratory notebook (ELN), and laboratory execution system (LES) into a single platform can effectively gather and share data on samples, methods of analysis, analytical results, environmental monitoring, and more, between the systems. These systems can also be integrated to enterprise resource planning, MES, and manufacturing control, which further enhances data integrity and efficiency.
One aspect of data collection and integrity that is gaining attention is temporary data, which are data entered into a computerized system but not saved to the database, either accidentally or intentionally. As results are entered into a LIMS data entry page, it remains in a temporary state until the analyst commits it to the database. While in its temporary state, the results can be changed by the user, even multiple times. Such changes might be deliberate in order to make a sample meet certain specifications.
For example, if the amount of product to be filled in a bottle relies on potency, an incorrect potency could cost the company money by adding too much to a bottle (reduced yield), by not adding enough (reprocessing), or by missing stability specifications (product recall). To err is human, and while not rampant, there is evidence of data tampering to fit specifications. Hence the focus on temporary memory.
FDA’s latest guidance (1) recommends placing this temporary data under control so it may be audited. Tracking any changes in temporary memory, therefore, increases the integrity of data and helps pinpoint when data are being manipulated into passing. This control can be achieved using dynamic auditing. Dynamic auditing tracks changes to data once it is entered into a data entry field and prior to the data being saved/committed to the database. Traditional auditing involves auditing of data after being saved/committed to the database.
Integrated platforms can overcome the many potential ways in which data might be modified. Values entered into a connected system, via connected instruments, are not susceptible to transcription errors. It is much more difficult to modify data without authorization within a LIMS, particularly when data come directly from a connected instrument, compared to a paper-based system that relies upon observation and accurate, manual recording. Informatics systems, for example, can be designed to automatically save all entered data, to prompt a user to provide a reason for making a change, and to include all data in the audit review record. In addition, data entered into an informatics system can be checked for accuracy as it is typed; for example, an informatics solution can be configured to recognize that a pH cannot possibly have a value of 15 or a length cannot be negative.
Data from a connected system is still only as accurate and consistent as the interface that connects the component parts. Therefore, thorough verification and validation testing is crucial to demonstrate the reliability of the interface. An integrated LIMS/ELN platform, for example, offers a higher level of protection because no data are transferred between the two systems; instead, each datum is stored exactly once. Removing the double data entry and the duplication of values between systems over an interface between separate LIMS and ELN platforms reduces risk.
Integration with manufacturing systems ensures data integrity by eliminating re-keying of data. The data are automatically transferred between interfacing systems, reducing the time lag between result generation and reporting, which can reduce overall production timelines.
Integrated software solutions can make it much easier to achieve compliance with regulations, such as FDA’s 21 Code of Federal Regulations Part 11 for electronic records and signatures (2) and the European Union’s Annex 11 for computerized systems in pharmaceutical companies (3). These solutions enable human readable and printable audit trails and compliant electronic signatures. Pharma companies should also make sure the platform complies with new guidance on auditing data stored in temporary memory, that it does not allow users to change data that are marked as having come directly from an instrument, and that it features modern, secure, Internet-enabled system-interfacing capabilities.
1. FDA, Draft Guidance, Data Integrity and Compliance with CGMP Guidance for Industry (CDER, April 2016).
2. CFR Title 21, Part 11
3. EC, Guide to Good Manufacturing Practice: Medicinal Products for Human and Veterinary Use--Annex 11: Computerized Systems, The Rules Governing Medicinal Products in the European Union Volume IV, Office for Publications of the European Communities, pp. 139-142 (Luxemburg, January 2011).
Vol. 42, No. 5
When referring to this article, please cite it as D. DiPalma and J. Vannest , "Integrated Systems Aid Data Integrity," Pharmaceutical Technology 42 (5) 2018.
Don DiPalma is director of Quality and Technical Services, and Jeff Vannest is senior director of Product Management, both at LabVantage, email@example.com.