USFDA 483 / Intas / May 2023 /Manipulation of Visual Inspection records, Particles in injection vials.

Observation 1

Batch records do not have complete information related to Production controls

A. Manual visual inspection records of parental line routinely manipulated to record values just below the limits for Particle defect categories – black particles, white particles, glass particles, fibers, since at least 2021. Two patterns of manipulations were observed:

1. 1. Change the reported counts for individual particles (black particles, white particles, glass, fibers) and reporting these as other categories or removing the reported counts. The originally recorded values were crossed out and rejects moved out to different categories This would mean there is no further investigation and inspectors need not reinspect the batch. 
   2. Visual Inspection records show a pattern where visual inspection records for multiple operators were identical. Nine different manual visual inspectors for different set of trays of an injection drug product reported identical number of particles for all listed categories. Similar practices observed in several batches. The personnel (interviewed by the Firm) explained carrying out the practice to make it appear there are no operator performance issues, by getting together and dividing what to record, resulting in everyone reporting the same thing.

The repeated patterns of altering /manipulating visual inspection records were observed since at least 2021 with several inspectors and supervisors involved. Quality Unit used these reports to release several batches.

B. Process controls for tablet compression are inadequate

Tablet rejection settings set by operator at beginning of campaign (no specific justification for the set parameters). Production personnel have access to change the settings during manufacture of a batch; changes were not documented in the electronic Batch records (BMR). The PLC change report (which record the changes) is not attached to the BMR or reviewed by Quality. Inprocess checks are not always documented when changes are made to ensure change did not impact quality. 

Manipulation, creation, misrepresentation of records is a fundamental GMP deviation. It is a basic crime in all spheres of life, not just cGMP and Pharma world. For Organisations it is about setting the right Quality culture across the company, different manufacturing sites. But what goes into setting the Quality culture? It is not only about training and lecturing people about Quality, cGMP and Data Integrity; These are of course required at a basic activity. But setting a Quality culture has a much larger perspective:

• Drive by example : Senior leadership at manufacturing sites, Top Management.
• Zero tolerance : A no nonsense, zero tolerance approach to handing integrity issues at all levels, in all areas. It is about who fits in and who don’t with the organisation.
• Appropriate reward and recognition mechanism / Right Ecosystem : People should be able to own up mistakes and open up. It is about an ecosystem where people are not scared about opening up about failures fearing repercussions. Owning up errors and working on the remediation are behaviours which to be applauded and rewarded rather than chided.
• Checks and Balances: Supervisory controls at different levels; Adequate oversight by “Independent” Quality unit; Mechanisms to detect issues – regular plant rounds, witnessing operations and activities, meaningful review of documents, meaningful Internal audits and IPQA (Inprocess Quality Assurance) checks. Why is it that issues that is observed by people outside the Firm like the FDA investigators (who have very limited information about the specific sites being audited), were not identified in internal audits, IPQA processes. Is it because of lack of knowledge, understanding, competency or wilful neglect and lack of oversight. Address the issues here. 
• Awareness build up: Do people understand the impact of their actions. Impact on patients & patient safety, Impact on Organisation. There should be regular interventions, lectures and trainings, case studies, small forum workshops where people are encouraged to come out and speak on issues, involving people at all levels.

(Also review the post on USFDA 483 / Sun Pharma, Mohali, India / Aug 3-12, 2022 on backdating of records)

Visible Particulates control in Injectable Products

Dwelling on the issue of Visible Particulates in Injectable products, Particle contamination control is critical as it can jeopardise patient safety. The controls have to be built in through a comprehensive and holistic approach. Elaborating from the USFDA draft guidance on  Inspection of Injectable Products for Visible Particulates a comprehensive approach should address all different phases involved from development to product realisation and distribution to market.

• Inconsistent process performance / spurt in visible particles on an established process and product could be caused by any one or a combination of these factors:
  • Inadequate controls of components, containers, or closures.
  •  A product formulation that is not stable.
  • Uncontrolled changes to the manufacturing process.
  • Equipment and facilities that are not suitable for their intended use.
  •  Personnel practices that generate particles.
    

    (Reference: USFDA draft guidance on  Inspection of Injectable Products for Visible Particulates) 

• Proper identification and characterisation of the particle during investigation of Visible Particulate inspection failures in Injectable products is critical to establish control. Particles Inherent to product (like agglomeration) if part of approved QTPP may not be a safety concern; Intrinsic particles may not have a sterility concern as all components would have been sterilised, but still has concerns on patient safety. The source of particulate material could be input materials, manufacturing equipment, their wear and tear, product-equipment-packing component interactions. Extrinsic particulates could be from inadequate facility controls, air quality, facility cleaning and maintenance processes and their adequacy, operator practices / interventions and these particulates also have concern of microbial contamination & sterility assurance.

Process controls for tablet compression

In tablet compression the compression settings are initially set within the design space / control space established in Process Development report. In regular commercial batches it is usual to start a batch with parameters with which the previous batch(es) were run (Compression force, Turret speed, Feeder ratio, fill depth etc.). With the initial settings, machine is run, tablets collected and tablet characteristics checked – appearance, weight, size and thickness, hardness, friability etc. Based on this further the batch is run with periodic adjustment in control parameters based on inprocess checks. This process should be explained in the Standard Operating Procedures (SOP) for tablet compression, also explaining how compression parameters are adjusted to obtain optimum characteristics during the batch run. Adjustments / changes during the run of batch will be necessary due to variabilities in bulk granule profile, environmental conditions and so on which can cause some characteristics of tablets to vary  – like the weight, thickness, hardness during a batch. The key point is to ensure that such adjustments and changes are documented in the Batch records, a process is in place for making those changes (like the doer -checker control by Production Operator and Supervisor). Review configuration of Batch records / eBMRs for interlocks within the BMR to avoid overriding inprocess tests, attaching PLC records etc. In Batch review records, acknowledge setting changes, summarise rationale for the changes. The key cGMP concern is whether the Quality unit is evaluating the changes while reviewing the batch. If the data is not recorded in the Batch record, or the associated documents like PLC record are not attached to Batch record, it means that Quality didn’t review the data while releasing the batch. This will be a concern for regulators and inspectors as changes beyond defined or validated ranges could impact the product quality through product shelf life (e.g. dissolution failures, build up of impurities and so on).   

• Impact of Data Integrity / discrepancies citings could be disastrous. Companies would do well to evaluate their systems and controls, review and enhance the Quality oversight, build checks and balances (IPQA / Internal Audits), Awareness build up. Continuously work on enhancing the Quality culture. Take up comprehensive actions for enhancing Quality culture in the organisation – Basic trainings and workshops, appropriate R&R (rewards and recognition) systems, review personnel competencies and credentials, redeploy personnel where necessary. Consider automation wherever feasible, as this reduces variability due to human element; implementation of electronic records (with appropriate compliances) – Work out a plan for the same with defined timelines.
• Evaluate and assess products for historical data for Visible particulates, trends, investigation of atypical values. If there are gaps, investigate the discrepancies comprehensively to identify the potential sources of particles, assess the controls on materials, facility and equipment, specific activities and operations including cleaning and maintenance, personal practices, Establish corrective actions. 
• Review the Master Batch Records of all products; If there is no provision for recording the changes in parameters during the batch run, and the data was not being reviewed, implement the same with provision for review of the data during batch review for Batch release. Implement system for attaching the change records from PLCs of equipment wherever the same is possible. For batches manufactured and distributed and within valid shelflife, if the data is still available in the PLCs / equipment memory prints can be taken /data reviewed and annexed to Batch record on current date. Perform a risk assessment of the potential impact of not reviewing the data and document the same.