USFDA 483 / Sun Pharma, Mohali, India / FEI 3002807979 / USFDA Investigators: Justin A Boyd, Teresa I Navas, Jonah S Ufferfilge  / Aug 3-12, 2022 / Observation 3 – Deficiencies in Product Quality Review, CpK, Effectiveness check

Written procedures are not drafted, reviewed and approved by appropriate organizational units and reviewed and approved by the Quality Control Unit.

1. In the APQR, calculated process capability index (CpK)for dissolution attributes indicate, poor consistency of manufacturing process.
2. CAPA effectiveness verification is inadequate (in CAPA proposed to address dissolution OOS trend)
3. Review of APQRs for the year 2020-2021 and SOP 000807 revealed that data from in-process controls quality attributes are not evaluated during the yearly evaluation of process performance capability.
4. In Tablet press, for compression of tablets for batch executed on August 3, 2022, reject level for the process was set up using average compression force from 2018 instead of using the values obtained during set up for the batch currently being manufactured

• Statistical tools used for evaluating data and trends as in APQR should be with adequate understanding and rationale. The tools used should be appropriate and able to bring in additional insights from data and parameters being reviewed. It will be inappropriate to use same tool / indicator across various parameters, just for stake of reporting a statistical indicator.
  • When there are multiple processes that can contribute to variability of an attribute measured, Process Capability Index (CpK) may not be an appropriate measure of Manufacturing Process Capability.  for eg CpK as a measure of process capability with respect to dissolution. Dissolution is a product attribute with intrinsic variability between individual units of a batch (like individual tablets) due to different factors in manufacturing process. This is also a reason the acceptance criteria has 3 levels of testing and requirements. Apart from that other aspects which can contribute to the variability are the analytical method (Method variability or measurement uncertainty), sampling method (representativeness). A more appropriate measure of Process Capability and trends in this case could be an assessment of trend of batches complying in S1, S2, S3 criteria and assess whether the trend is improving (like more batches or less batches comply in S1, or S2 criteria compared to previous time period)
  • A process capability index CpK will be more appropriate for process parameters which are directly measured and/or has limited number of sub processes which can contribute to variability – like Compression force, thickness of a tablet, hardness of tablet. This will help in establishing correlation between process parameters and attribute being measured and take appropriate corrective actions to the process.
  • Select appropriate tools for statistical evaluation of different parameters –Measures of variability like Relative Standard Deviation(%RSD) or standard deviation (%SD), Trend charts, Frequency distributions, Process capability Index (CpK), considering the parameter, magnitude of values, utility of the tool to flag an emerging trend or issue. Document the rationale in the procedure for product quality review.
• There should be a robust process for effectiveness verification of CAPAs, Changes. If certain actions are proposed to improve a factor / attribute, there should be a process to review whether the actions really improved the process. This should be part of the Change evaluation and closure process. The change control documentation should identify whether the change proposed require an effectiveness verification (and if not why). Similarly, when CAPA is getting finally closed, it should also look at whether the CAPA was effective in improving the triggering factor.  And annual product quality review (APQR) is an opportunity to review the same connecting the Change and the CAPA which triggered the change. The procedure for APQR should address the same.
• The APQRs should consider all data points which can indicate process performance capability. In process quality control parameters and test data, process control parameters should be part of Product Quality Review and procedure for APQR should address the same.
• A robust Product Quality Review (PQR) should also identify opportunities for Continuous improvements. The process control parameters could be defined with broader ranges during initial scale up and validation as there is limited data. But as data from more number of batches become available over the years, it can indicate opportunities for narrowing the range of the control parameter, which can help in reducing the process variability. This is the spirit behind the Continuous process verification / validation concept – Improve the processes towards 6-Sigma quality. (e.g. A compression process reject level could be defined more broadly in the beginning; but with more data and learnings over the years the range could be narrowed down with tighter control of process parameters like Compression force, for improving process consistency in product CQAs- Critical Quality Attributes- like dissolution). Procedure for APQR should address the same.

• Review the procedure for Product Quality review, statistical tools applied for different attributes for measuring reviewing process trends. Apply the appropriate tools and document the rationale. And based on this where the PQR flags process performance, consistency issues investigate identify additional actions where necessary.
• Review the procedure for PQR addressing CAPA effectiveness, effectiveness of changes, review of in process quality control parameters and process parameters and identification of opportunities for Continuous improvements.
• Review all processes with respect to process control parameters, inprocess quality control parameters to see whether the ranges, settings employed are consistent with current data (or draw on historical and process validation data heavily). Accordingly identify areas for improving the control parameters.