USFDA 483 / Intas / May 2023 /Failure to review discrepancies, deviations

Failure to thoroughly review any unexplained discrepancy, whether the batch is distributed or not:

A. Visible particles in injectables are not characterised when rejects exceed the defined limits for individual particles (Black particles, White particles, Glass particles) and /or Total particles during visual inspection. Deviations were closed with no identification of root cause, potential sources, particles were not isolated and evaluated, no CAPAs are implemented. (six instances are cited)

B. CpK, PpK values for trend analysis of Visual inspection reject rates showed process is not capable. No investigation performed for batch to batch variation. Limits for individual defect categories were not revised based on historical data.

C. Non-viable particle count (NVPC) excursions in the Grade A filling areas are not investigated to identify the root cause, impact on products being filled; No procedure to remove open vials and stoppers present at the time of machine stoppage due to excursion.

D. No effective measures taken to investigate sources and reduce the occurrence of spore forming organisms in the aseptic processing areas.

E. During manufacture of a tablet drug product batch, operator failed to identify a deviation and proceeded with the batch. The batch was released based in in process results and final batch results. Batch release COA dated 6 Oct 2022, Additional lot samples test were approved on 19 Oct 2022. Investigation did not evaluate the actual parameters set by the Production operator, adequacy of sampling process, sample size selection for additional lot samples and result variation between initial COA and additional lot sample.

F. In a deviation for rough surface observed on tablets, evaluation did not include some of the process parameters; these parameters can be altered as per Operator discretion. It is unknown when the batch samples began trending out of specification for roughness as the operators document “ok” for physical appearance. No assessment of impact on dissolution, homogeneity, efficacy performed. The lots were released for distribution. Batch was not charged for stability. It is not known how many tablets were rejected for rough surface from the total number of reject tablets (7967 as only major defect categories are documented in Batch record.

G. Approximately 206 complaints for odour or smell in tablets between January 2020 and May 2023. Investigations identified (an entity / chemical) in API and its degradation in final product as the cause for odour. But no evaluation of the API or the drug product for amount of the (an entity / chemical). A process change for API was approved by FDA, both API lots (old and new) are available at site. No statistical evaluation of the variation between the API processes.

H. A complaint investigation for an Ointment as “liquified” did not evaluate viscosity of the retainer samples

I. When endotoxin tests are invalidated due to suitability failure, test are repeated without documenting a laboratory Incident report.

Comprehensive procedures for investigation of failures and deviation and robust implementation of the procedures should be in place and implemented, personnel trained in failure investigation tools. Scientific tools for investigation like Brainstorming, Fish bone (Ishikawa) analysis involving cross functional team (CFT) to list out all potential factors and categorising them into different branches (like Men/People, Machine, Method, Measurement, Environment) shall be applied; Further a Fault tree approach to analyze and organize the potential causes can help in systematically eliminating causes that do not align and trace the most probable cause(s). Brainstorming with a experienced CFT helps in identifying and listing all potential causes. Fault tree analysis is a good tool which helps to break down the event /deviation into its various contributing causes and sub-causes in a logical and structured manner. By connecting causes using logical gates (AND, OR, NOT), systematically eliminate potential causes that do not align with the observed Top Event (discrepancy or failure) to arrive at the most probable cause(s); Trace through the AND/ OR logical gates till the most basic events are reached, which represent the root causes for the failure. A Pareto analysis can help in picking out the top 20% of the potential causes, which can be further evaluated by Fault tree analysis to get to the root cause(s) to make the investigation efficient. Once the root cause(s) is identified establish appropriate CAPAs for addressing the causes. With the root causes identified, also perform an impact analysis on other products / batches which could be similarly impacted and take up appropriate remediation actions. Document the entire investigation elaborately which gives the auditors and regulators confidence that the Quality system of the company is robust and if any events happen, the Company will go the depth of the issue and take remedial actions.

Useful links for Fault tree analysis:

• https://www.youtube.com/watch?v=dKtIG0UXS6Y&ab_channel=MisterSimplify 
• https://www.youtube.com/watch?v=ZEShNJX3kcg&t=14s

Each of the failure / discrepancy as cited in the 483 could have several potential causes, which need to be evaluated and investigated by the CFT. For example:

1. Visible particles in Injectables:  Multiple related issues like inconsistent quality and reject rates in Visual inspection, Excursion of Non viable particles in aseptic areas, Incidence of spore forming organisms in aseptic areas, points to systemic issues. When rejects exceed limits for particulates, isolation of the particles and evaluation to identify the type of particles, characterisation of the particle is a must for identifying potential sources of the particles. A thorough brainstorming by a cross functional team of Production, QA, QC, Maintenance & Engineering, R&D of all possible sources like (but not limited to) – Materials (API, excipients, Container closure system – vial, stopper..); Equipment and accessories – Equipment wear and tear, tubing, filters; Personal practices and Operator behaviour and interventions during filling, gowning; Utilities – Air, Ventilation systems, water; Product compatibility-interaction-stability with contact surfaces in equipment, container closure systems; cGMP manufacturing practices and processes  – Cleaning, Maintenance; Inherent characteristics of the product and its stability help in listing down a detailed list of potetial sources which shall be evaluated further. When incidents recur it will be good to have the system thoroughly audited by third party experts, who may identify potential sources other than what the internal team has investigated.
2. Trend analysis / Process Capability issues: Trend analysis of visible particle trends, alerts for out of trend and investigation are critical elements of a comprehensive contamination control strategy. However, the statistical tools used should be appropriate for the purpose. CpK / PpK are tools more suited for evaluating direct product quality attributes with target values and target ranges which are influenced by processing parameters – e.g,-assay, fill volume. For attributes like reject rates due to Visible particles, where the objective is to have zero rejects and potential causes can be many, trend analysis will be more relevant. Nevertheless, investigation into root cause is more important when out of trends or spikes are observed in trend analysis. Also based on trend analysis data the limits for individual particles shall be reviewed periodically. For e.g. for a certain individual particle type if the reject rates are consistently low during actual commercial manufacture than the limits defined initially, the limits ought to be reviewed, and there should be strong rationale documented, if the limits are not revised.
3. Spore Forming Organisms in Aseptic areas / Excursion of Non viable counts: These could be indicative of systemic issues like (but not limited to) effectiveness and thoroughness of cleaning and sanitation practices, disinfection process, personnel practices – personnel movement, gowning practices, material controls (input materials, packing materials, accessories) and / or material movement, HVAC system and their maintenance.
4. Rough surfaces on tablets  / Deviations during compression operation: Deviations like rough surfaces, excursion of process parameters, in process parameters during tablet compression can impact on the quality of batch. The root causes could include (but not limited to) homogeneity of bulk granules, particle profile of granules, compressibility, lubrication issues with potential impact on Content uniformity, Disintegration, Dissolution. It is possible that the issues manifest during product shelf life, even though the tests were well complying during batch release testing. Such events require thorough evaluation of potential impact after identifying root cause for evaluation. Batch documentation should document the changes during manufacture of the batch, audit trials from the equipment should be reviewed as it can help identify the time when deviation started / occurred, the extent of issues, how many tablets within the batch could be impacted.  Loading impacted batch for stability studies will help the Firm to keep monitoring impact and take timely action, should the tablets fail for any parameters during shelf life.
5. Complaints-Odour / Smell in tablets etc: Even when a probable root cause is projected and actions are taken to address the root cause; the investigation should evaluate the issue in detail, to be sure the root cause identified is appropriate and supported by corroborative supporting data. Otherwise, we run the risk of leaving out other causes which could also have a role in the deviation. When impurities / components in the drug product is projected as cause for issues like smell and odour, it should be supported by data. The investigation should also look at other potential causes and this should be documented with adequate rationale if it is concluded that there is no other causes and product / process is in control. 
6. Failure of suitability test in Endotoxin Testing: Recurrence of failure of suitability test in Endotoxin testing could be indicative of issues in the suitability, capability, robustness of the method for specific sample matrices, interference of sample matrix due to factors like colour, viscosity, pH and so on. Such events require investigation to establish root cause; and improvements / correction in the method and should be handled through appropriate quality events like Laboratory Incident Report (LIR). It is not enough to perform a repeat testing without evaluating the incident or else it could be interpreted as testing into compliance, and failures will keep recurring as the fundamental issues are not getting addressed.

It is very well understood in cGMP that failures/ deviations should be investigated comprehensively, root causes established, impact evaluated and corrective actions (CAPA) and remediation measures implemented. But still there could be areas/ events where the failure investigations are inadequate – like visible particles excursions, thoroughness of batch records in recording all batch operational data, excursions in aseptic areas, missing to investigate CpK/ PpK indications of process capability issues. Thoroughly evaluate the quality events / failure investigations for lapses, take up further the investigation where gaps are observed and take appropriate actions based on investigation outcome. Document the investigation in detail. Take up programmes for enhancing the competency of the team in investigations of failures; involve CFT in investigation, train the team in scientific investigation tools like Brain Storming, Bish-Bone or Ishikawa analysis, Fault Tree Analysis, FEMA.

The investigation documentation and investigation process should give confidence to the auditors that the Firm and its Quality unit can be relied on for comprehensive investigation and remedial actions when failures and discrepancies occur.