Digitization for Data Integrity and Regulatory Compliance

This article will explore how the move to digitize laboratory processes promises to make data integrity and compliance, an essential goal for regulated laboratories, an easier task. If we go back to the previous century, in 1990 the pharmaceutical industry requested that the US Food and Drug Administration (FDA) permit the use of electronic signatures, which led to the publication of 21 CFR 11 for electronic records and electronic signatures.¹ However, with this regulation in place there was little movement towards working electronically due to the conservative nature of the industry. However, in 2005 Able Laboratories was found to be generating fraudulent paper and electronic data to release products² and this began the emphasis from regulators and industry bodies^3-12 on ensuring data integrity, delivered through enforcement actions and guidance documents.

The road to digitization

The road to digitized processes requires the removal of hybrid systems (signed paper printouts linked to electronic records). The World Health Organization (WHO) guidance on good records and data management practices recommended that hybrid systems are not encouraged and that they should be replaced at the earliest opportunity.⁶ Larger regulated laboratories have started to address the problem with implementation of technically compliant networked informatics solutions focused on laboratory information management systems (LIMS), chromatography data systems (CDS) and electronic laboratory notebooks (ELN).

Interfacing analytical instruments is also a requirement for digitization. There is evidence that the FDA is losing patience with the slow pace of companies’ automation efforts. This is shown in two warning letters from July 2020. Stason Pharmaceuticals and Tender Corporation received identical citations and requirements for remediation:

Technological improvements to increase the integration of data generated through electronic systems from standalone equipment (e.g., balances, pH meters, water content testing) into the LIMS network.^13,14

These two warning letters have been reviewed by the author looking at the costs of compliance and non-compliance.¹⁵ The message from regulatory authorities either in data integrity guidance documents or through regulatory citations is to digitize the laboratory. How best to achieve this goal

Administration of blank paper forms

If paper records are to be used, then the regulatory expectation is that:

• The master template of each blank form needs to be approved and protected
• Each copy must be uniquely numbered
• Each copy should be printed independently by Quality Assurance (QA) personnel
• Records may require printing on colored paper not available in the laboratory or using another security system
• Records should be tracked and traced through the analytical process
• Damaged forms need a good explanation to be replaced and must be kept for inspection^7,8

This is a very labor-intensive process with a high administrative overhead. Using an electronic document management system (EDMS) may be a possible alternative if the application has controlled copy printing functionality that uniquely numbers each blank form, but it must be validated extensively.⁸ However, meeting the track and trace requirement is unlikely to be met by the EDMS. It is important that paper records are eliminated and replaced by electronic processes.

The way it is…

Therefore, in many laboratories we have a situation where:

• Results are done by observation where there is no objective evidence available for review, audit or inspection e.g., color, appearance, odor, melting point or colony counting
• Many analytical instruments are not interfaced to informatics applications and generate paper records e.g., analytical balances, pH meter, melting point apparatus
• Sample preparation in conducted in combination with an analytical balance e.g., a loss-on-drying (LOD) analysis is typically manual in combination with either a calculator or spreadsheet
• Analytical instruments with controlling software are operating as hybrid systems (signed paper printouts linked to the electronic records in the data system) e.g., CDS, UV and IR spectrometers

We will explore how some of these examples can be digitized to reduce or eliminate paper and enhance both efficiency and regulatory compliance in a laboratory.

Results by observation – 1    

Some of the simplest tests in a regulated laboratory are often the most difficult to generate objective evidence, e.g., color, odor or appearance, where the test result is written down on a controlled batch record form or in a laboratory notebook. This result is then entered into a LIMS or other informatics application to collate the results for the certificate of analysis (CoA). The problem with such tests is that no objective evidence exists for a second person to review. To ensure data integrity, the reviewer should repeat the test, which is not difficult, to confirm the result. However, all work is still carried out using paper and the result must be manually typed into a LIMS for the CoA.

An alternative approach is to include the requirements for each observation test in a LIMS so that the test can be allocated to the appropriate sample. The instructions in the system will allow the analyst to complete the test electronically by entry of the result directly into the LIMS. This has the following advantages:

• No paper records and blank form controls eliminated
• Attribution of action by unique user identity
• Enforced analysis workflow
• Automatic time stamp, synchronized to a trusted time source, to ensure contemporaneous recording
• Electronic signature by the analyst of the result
• Elimination of manual entry of the result into a LIMS
• Repetition of the workflow for the reviewer, including a check of the audit trail entries (if applicable) followed by the reviewer’s electronic signature
• Automatic incorporation of the result in the CoA

This approach has the advantage of ensuring regulatory compliance and data integrity as well as testing efficiency.

Results by observation – 2    

A melting point or melting range determination is used to assess the purity of a substance and occurs when there is a transition from a solid through softening and melting to a clear liquid. A standard procedure for melting range determinations is USP 741,¹⁶ here the temperature of a heating block is raised by 1 ˚C per minute to ensure consistent heat transfer and avoid thermal lag to the test and reference substances. An analyst records the temperature according to the pharmacopoeial specification. When using a classical instrument, the analyst determines the endpoint and records the temperature or range. This is written down in a laboratory notebook or controlled analytical worksheet and there is no objective evidence for review by the second person.

Modern melting point instruments are more sophisticated and can record the process by video and determine the melting point using software. This permits objective review of the data, but the result is still written down on paper as before. Also, the memory of the instrument may fill with videos that must not be deleted.

To complete the digitization process, the instrument needs to be connected to a networked informatics solution that has:

• Unique user identities for all users, with appropriate access privileges and no conflicts of interest
• Methods for the different substances to be tested
• Has checks to ensure that the instrument is qualified
• Can run any calibration tests either before sample analysis or in parallel with it
• Can download sample information together with the method to be performed
• Can control the instrument; conduct melting point tests then manage and securely store the associated melting point videos on a network drive
• The capacity to generate a report that is signed electronically by the analyst
• The data, including audit trail entries, available for review by a second person
• The ability to produce a report that is approved electronically by the reviewer

This has the advantage of working electronically with objective evidence of any test available for review, audit or inspection as well as eliminating the associated paper records.

 

 

Figure 1: Comparison of Automated and Manual Melting Point (MP) Determination

Loss-on-drying analysis

USP <731> ,¹⁷ the traditional pharmacopoeial loss-on-drying (LOD) analysis, is a slow manual process involving weighing two aliquots of sample, heating them for 3 hours at 105 ˚C, cooling them and reweighing to calculate the water loss by difference using a calculator or spreadsheet. The mean result is compared with a specification to determine the pass or fail status for the material. Similar to the analyses described above, there are paper records plus balance printouts and a hybrid system if a spreadsheet is used.

Digitization of the process requires an informatics solution (e.g., LIMS, ELN or LES) to:

• Generate an electronic workflow for the whole process including second-person review and electronic signatures for authorship and review
• A workflow that is linked to sample information to enable the correct test and specification to be allocated and the reportable result to be collated for the report and CoA
• Interface an analytical balance to the informatics application
• Perform automatic checks for the balance to ensure that it is the right type of balance for the analysis, it is qualified and also calibration checks against calibration masses before starting the analysis
• Calculate the percentage LOD including meaning and rounding need to be incorporated in the informatics application to eliminate the spreadsheet
• Generate of a report for electronic signing by the analyst
• Enable second person data review and electronic signature sign-off

There are still some paper records with the instrument use logbooks of the analytical balance and the oven for recording the work and the recording the sample drying times. However, the electronic workflow will be faster than the paper-based process and with better regulatory compliance and data integrity.

Colony counting

Manual colony counting of Petri dishes, used for environmental monitoring or determining contamination of sterile and aseptic products, is notoriously error prone. If there are relatively few colonies there is reasonable accuracy but when there are larger numbers, accuracy falls.¹² Also, there is no evidence available after the test as the plates cannot be retained, instead, they are destroyed at the end of an analysis.

As a minimum requirement, there should be objective evidence available for review, audit and inspection as well as a consistent means of colony counting. A colony counter is required for automation of the process plus the provision of objective evidence via a photograph; studies show that automated counting is more reproducible than manual counting. These instruments can either be manually fed or if the volume is sufficient using an automatic plate feeder. Analyst comments can be made into the controlling software so that all work is traceable. Application software must have the ability to connect to other informatics systems such as LIMS to enable the results to be transferred automatically. Some systems offer a CSV file transfer which needs to be secure as the file can be manipulated between the source and destination application.

Most colony counting systems are designed as standalone which means that records are stored on the PC must be backup regularly and verified. This should be done automatically with a backup agent on the workstation that is connected to the network.

Summary

To ensure data integrity and regulatory compliance, analytical procedures involving observation must change from recording observations on paper to direct data acquisition by informatics applications. Three approaches to archive this are discussed here: loss on drying, melting point determination and colony counting. To achieve digitization requires investment and management is often unwilling to do this. However, laboratory digitization is being encouraged by Good Manufacturing Practice (GMP) regulators using a carrot and stick approach. The carrot is contained in regulatory guidance documents and the stick is via regulatory citations and warning letters. If the former approach does not work, the latter certainly will.  

References

1. US Food and Drug Administration. 21 CFR 11 Electronic records; electronic signatures, final rule, in Title 21 1997, US Food and Drug Administration: Washington, DC.

2. Able Laboratories Form 483 Observations. US Food and Drug Administration. https://www.fda.gov/media/70711/download. Published 2005. Accessed August 2021.

3. Medicines and Healthcare products Regulatory Agency. MHRA GMP Data Integrity Definitions and Guidance for Industry 1st Edition. Medicines and Healthcare products Regulatory Agency: London. Published 2015.

4. Medicines and Healthcare products Regulatory Agency. MHRA GMP Data Integrity Definitions and Guidance for Industry 2nd Edition. Medicines and Healthcare products Regulatory Agency: London. Published 2015.

5. MHRA GXP Data Integrity Guidance and Definitions. Medicines and Healthcare products Regulatory Agency. https://www.gov.uk/government/publications/guidance-on-gxp-data-integrity. Published March 9, 2018. Accessed August 2021.

6. WHO Technical Report Series No.996 Annex 5 Guidance on Good Data and Records Management Practices. 2016, World Health Organization: Geneva.

7. FDA Guidance for Industry Data Integrity and Compliance With Drug CGMP Questions and Answers. US Food and Drug Administration. https://www.fda.gov/media/119267/download. Published December 2018. Accessed August 2021.

8. PIC/S PI-041 Good Practices for Data Management and Integrity in Regulated GMP/GDP Environments Draft. Pharmaceutical Inspection Convention/Pharmaceutical Inspection Cooperation Scheme. https://picscheme.org/docview/4234. Published July 2021. Accessed August 2021.

9. GAMP Good Practice Guide: Data Integrity – Key Concepts. Tampa, FL: International Society for Pharmaceutical Engineering. Published October 2018.

10. GAMP Good Practice Guide: Data Integrity by Design. Tampa, FL: International Society for Pharmaceutical Engineering. Published October 2020.

11. GAMP Guide: Records and Data integrity. Tampa, FL: International Society for Pharmaceutical Engineering. Published March 2017.

12. PDA Technical Report 80: Data Integrity Management System for Pharmaceutical Laboratories. Bethesda, MD: Parenteral Drug Association (PDA). Published August 2018.

13. FDA Warning Letter Stason Pharmaceuticals, Inc. US Food and Drug Administration. https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/stason-pharmaceuticals-inc-604889-07082020. Published July 2020. Accessed August 2021.

14. FDA Warning Letter Tender Corporation 2020, US Food and Drug Administration. https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/tender-corporation-599789-07232020. Published July 2020. Accessed August 2021.

15. McDowall RD. Do you really understand the cost of noncompliance? Spectroscopy. 2020;35(11):13-22.

16. USP General Chapter <741> Melting Range or Temperature. Rockville, MD: United States Pharmacopoeia Commission Inc.

17. USP General Chapter <731> Loss on Drying. Rockville, MD: United States Pharmacopeia.