BLOG
MARKET INTEL by PharmaCompass
CONTENT by Suppliers
- Interview #SpeakPharma
- Video #SupplierSpotlight
- Vlog #PharmaReel
- Company Bio #AboutSupplier
- Services Bio #AboutCapabilities
News
Create content with us, ask us
Peter J. Werth, President and CEO of ChemWerth, clears all misunderstandings generated by his Speak Pharma interview, where he said ‘data integrity has no relationship with product quality’
A big thank you to all those who commented on my interview, published on this website July 7, 2016. After reading your comments, criticisms and suggestions, I felt the need to make some clarifications, the first of which is regarding what I meant by ‘data history’. Data history comprises of all relevant data that truly reflects the quality of the product because the analysis was performed properly and the data is all valid.
The data integrity (DI) file contains the data that does not reflect on product quality since it is erroneous data that was caused by some obvious error (gross error). The DI file is a part of the product’s data history, but does not relate to the product’s quality.
I believe the present FDA draft guideline for DI should be rechristened ‘Draft Guideline for Data History’. It should be revised to describe data history, and guide firms on how to maintain good data history that reflects product quality.
How the Jinan Jinda case changed my outlook
Recently, China-based Jinan Jinda Pharmaceutical was inspected by the FDA. It received eight 483 observations related to DI, pertaining to 75 batches. This came as a shock to me. ChemWerth has represented Jinda for over 15 years.
Jinda had never failed FDA inspections and customer audits. I always felt Jinda was a good GMP factory that met all international quality standards.
We reanalyzed 90 batches related to DI deficiency points (which the FDA sees as fraudulent practices). All 90 batches met the specifications. We reviewed the customer complaint file and found no quality-related complaints.
After more research, we discovered there is a relationship between DI and fraudulent practices, but only in the case of manufacturers who produce both the API and finished dosage.
‘API only’ versus API and dosage form manufacturers
The interview was almost entirely related to data history and DI associated with FDA inspections of ‘API only’ manufacturers.
In all cases, where we checked the 483 deficiency points related to numbered batches, we didn’t find even one batch that didn’t meet the specifications. We stopped this process when we reached 200 batches with no failures.
That’s how we arrived at 'zero correlation between the DI (suspected fraudulent activity) 483 deficiency points and product quality’.
However, there is 100 percent correlation between data history (even though deletions were made) and product quality.
Future FDA inspections for ‘API only’ manufacturers should emphasize on GMPs, traceable records, analytical practices, data history and overall plant operations.
How our approach differs from FDA’s?
The primary difference between the existing FDA guidance on lab errors and that of ChemWerth is that “gross errors” are a focus in our SOP (standard operating procedure), which we believe is the cause of most DI issues. As such, our procedure outlines how raw data related to gross errors specifically should be handled in the lab.
The primary action would be to transfer this erroneous data into a separate ‘DI’ location where the data will be maintained (and not deleted). An abbreviated investigation would occur concerning this data. While proper attention would be paid to such an investigation, it would not overburden or distract the laboratories.
This will allow for the more important and product quality-related data to be better maintained as ‘product history’ data. This product history data truly reflects the actual quality of the product.
Why data should not be deleted without documentation
Sometimes I used the word ‘delete’ when I should have said ‘transfer to the DI file’. I did mention that in the past, analysts did delete gross errors, justifying such deletions as ‘gross errors that are not part of data history’.
They also saved analytical time and paperwork by not filling an OOS (out-of-specification) investigation report. While this implies poor lab training and poor laboratory practices in some areas, it is not a fraudulent act that deserves a warning letter.
Why my statements on DI apply primarily to API manufacturers…
After much thought I realized my statements (in the interview) lacked clarity. My comments pertained to ‘API only’ manufacturers who are responsible for the product quality they ship to the dosage form manufacturer.
The dosage form manufacturer, in turn, is responsible for the final quality of the API that goes into the dosage. Any product that does not meet specification will be rejected and returned to the API manufacturer.
There is no economic advantage and no good reason for the ‘API only’ manufacturer to delete or falsify data. They cannot cheat on the analytical testing and produce fraudulent COAs (certificate of authenticity), which makes DI issues related to quality non-existent.
The final dosage manufacturer has advanced equipment and well-trained analysts. The dosage manufacturer will certainly reject poor quality batches.
The misunderstandings…
Many readers misunderstood my comment that there are at least 18 reasons to transfer (they thought delete) data on gross laboratory errors from data history. I did not imply deletion of such data.
I stated that the data should be filed in the DI file, properly labeled and documented. The data history file is 100 percent related to the product quality and should contain all information/data to show a good traceable record. Once again, most of my experience pertains to ‘API only’ manufacturers.
The SOP (that went as a link in the article) clearly states that all laboratory gross errors need to be reviewed and documented by the supervisor before the data is transferred to the DI file. I too believe that no data should be deleted.
Was I attacking the FDA?
No. Rather I was trying to help the FDA so that data history and DI factory inspections are fair and correct. ChemWerth was the first API distributor in the world to hire in-house staff dedicated to improving the GMP level of Chinese API facilities.
In the past, I have personally helped the FDA. I built my business on the premise that we need to have the full trust and support of the FDA, so we can act as a team to ensure high quality. We believe strongly in GMP compliance and want the FDA to enforce the rules and regulations.
Many API firms, particularly in to China and India, are working hard to put DI systems in place. To do this, they need support in hardware/software as well as more training on lab operator data collection practices. Of course, having the right policies and procedures in place in handling DI is also required. Most suppliers want to do this right, and are not looking commit fraud.
What we are doing to improve DI compliance
As a company, ChemWerth is doing several things to improve overall DI compliance of API suppliers.
ChemWerth has committed extensive resources in specifically auditing API suppliers for DI. We are also conducting extensive training and offering advice on what systems are needed to ensure DI can be achieved.
We have also contracted a highly reputable GMP consulting firm that specializes in electronic DI and offers extensive training to all API suppliers in China. Furthermore, ChemWerth holds an annual GMP seminar with multiple API suppliers in China. And the seminar’s focus this year will be DI.
Finally, ChemWerth has shared its DI SOP that focuses on gross errors/data history, which we plan on implementing at our represented API suppliers with the FDA.
The PharmaCompass Newsletter – Sign Up, Stay Ahead
Feedback, help us to improve. Click here
Image Credit : Crystal by storebukkebruse is licensed under CC BY 2.0
“ The article is based on the information available in public and which the author believes to be true. The author is not disseminating any information, which the author believes or knows, is confidential or in conflict with the privacy of any person. The views expressed or information supplied through this article is mere opinion and observation of the author. The author does not intend to defame, insult or, cause loss or damage to anyone, in any manner, through this article.”
