The FDA commented in a Warning Letter to Aurolife Pharma that the manufacturer lacked data showing that the process was in “state of control” before batch release. The 483 pointed to the FDA’s guidance document Process Validation: General Principles and Practices, and found the company lacks a state of control of the process, which comes back to change control.
They also found major deficiencies in their OOS and cleaning programs.
Starts with our “Heroes are gone” and “it is time to stand on our two feet.”
Focuses on the time and effort to train people on lean and six sigma, and how many people do not actually do projects. Basic point is that we use the tools in old ways which are not nimble and aligned to today’s needs. The tools we use versus the tools we are taught.
Hacking lean six sigma is along a similar line to Art Smalley’s four problems.
Applying the spirit of hacking to quality.
Covers valuestream mapping and spaghetti diagrams with a focus on “they delays in between.” Talks about how control charts are not more standard. Basic point is people don’t spend enough time with the tools of quality. A point I have opinions on that will end up in another post.
Overcooked data versus raw data – summarized data has little or no nutritional value.
Brings this back to the issue of lack of problem diagnosis and not problem solving. Comes back to a need for a few easy tools and not the long-tail of six sigma.
This talk is very focused on LSS and the use of very specific tools, which seems like an odd choice at an Audit conference.
I appreciate it when the session manager (person who introduces the speaker and manages time) does a safety moment. Way to practice what we preach. Seriously, it should be a norm at all conferences.
Connects with the audience with a confession that the speaker is here to share her pain.
Objective – where we are going. Provide a flow chart of mission/vision (scope) ->establish process -> right direction? -> monitor and measure
Objectives should challenge the organization. Should not be too easy. References SMART. Covers objectives in very standard way. “Remember the purpose is to focus the effort of the entire organization toward these goals.” Links process objectives to the overall company objectives.
Process measures are harder. Uses training for an example. Which tells me adult learning practice is not as much as the QBOK way of thinking as I would like. Kilpatrick is a pretty well-known model.
Process measures will not tell us if we have the right process is a pretty loaded concept. Being careful of what you measure is good advice.
One of the trends in 2019 FDA Warning letters has been cleaning. While not one of the four big ones, cleaning validation always seems relevant and I’m looking forward to this presentation.
Starting with the fact that 15% if all observations on 483 forms related to leaning validation and documentation.
Reviews the three stages from the 2011 FDA Process Validation Guidance and then delvers into a deeper validation lifecycle flowchart.
Some highlights:
Stage 1 – choosing the right cleaning agent; different manufacturers of cleaning agents; long-term damage to equipment parts and cleaning agent compatibility. Vendor study for cleaning agent; concentration levels; challenge the cleaning process with different concentrations.
Delves more into cleaning acceptance limits and the importance of calculating in multiple ways. Stresses the importance of an involvement of a toxicologist. Stresses the use of Permitted Daily Exposure and how it can be difficult to get the F-factors.
Ensure that analytical methods meet ICHQ2(R1). Recovery studies on materials of construction. For cleaning agent look for target marker, check if other components in the laboratory also use this marker. Pitfall is the glassware washer not validated.
Trends around recovery factors, for example recoveries for stainless tell should be 90%.
Discusses matrix rationales from the Mylan 483 stressing the need to ensure all toxicity levels are determined and pharmaceological potency is there.
Stage 2 all studies should include visual inspection, micro and analytical. Materials of construction and surface area calculations and swabs on hard to clean or water hold up locations. Chromatography must be assessed for extraneous peaks.
Verification vs verification – validation always preferred.
Training – qualify the individuals who swab. Qualify visual inspectors.
Should see campaign studies, clean hold studies and dirty equipment hold studies.
Stage 3 – continuous is so critical, where folks fall flat. Do every 6 months, no more than a year or manual. CIP should be under a periodic review of mechanical aspects which means requal can be 2-3 years out.
Investigations of a Dog 