Tuesday, September 18, 2018

Process Verification / Validation

An answer to a query:

Ques:  The  company  (site) that I work for, manufactures intermediate products.  The question that we have is-If a change is made to a process,  should we do validation or is verification sufficient?  All our processes have QC/acceptance criteria.  For example, we got a validated instrument from another site (of our company).   We got IQ/OQ done by the vendor at our site.  We also have the same instrument from another vendor, which will be retired in couple of months.  Can we do verification of the new instrument by writing a technical report ?    In other words, we will  first conduct  a risk assessment and then do experiments to compare the data of the two instruments and record them in a report.  Actually we have already done feasibility studies to show that the new instrument gives the same data as the old one for the same samples .  I heard , in your talk that FDA doesn't require validation if we have the means for 100% checking.  We have similar situations in which we sometimes make a change in the process but we always have means to verify the final product because we have acceptance criteria in place.  Is verification enough for such situations?    Please advise.

Ans:  "FDA doesn't require validation if ..."  The specific reference for my comment for medical device manufacturing is 820.75 "Process validation, (a) "where a the results of a process cannot be fully verified by subsequent inspection and test, the process shall be validated with a high degree of assurance..."  The key exception to this is automated (computerized) processes, which per 820.70(i) must be validated no matter what.

As to the rest of your question, part of the answer depends upon your definitions of verification and validation, technical report, feasibility studies, et al.  Any change requires some level of verification (testing, checking, feasibility studies, etc), or a series of verifications (part of a validation) to prove that the change does what it should and doesn't do what it shouldn't.  While I won't state what you've outlined will satisfy requirements (though on the surface it appears to), if you've satisfied the CGMPs and what I've outlined above, yielding data that nothing has changed and that product quality / specifications are assured after the changes per your data, documented, then you've met the requirements.

Saturday, August 18, 2018

Method Validations

Some Q&A from a recent webinar:

Ques: Do you have any recommended references to create risk categories for method validations?


JVT Article 2004, especially note flow charts; caveat – include potential problems from normal use, not just from a failure mode:


PDA Slides:


Q8, Q9, & Q10 Questions and Answers, U.S. FDA:


U.S. FDA Guidance, 2015: Analytical Procedures and Methods Validation for Drugs and Biologics; note several of the headings may provide risk categories, e.g, Apparatus/Equipment, Operating Parameters, Reagents/Standards, Sample Preparation, Standards Control Solution Preparation, Procedure, System Suitability, Calculations, Data Reporting... :



Thursday, July 26, 2018

CE-Marking and China; Design Inputs / Requirements

Answers to some questions from my recent 3 hour DHF, DMR, DHR ... TD webinar:

CE Mark and China?

The European Union comprises 28 countries that require CE-Marking. Three additional countries (Norway, Iceland, Liechtenstein), although not officially part of the European Union, are signatories to the European Economic Area (EEA). Switzerland is not an EU member nor a signatory to the EEA, but they have transposed the Medical Devices Directives into their national law and these countries require CE Marking. -- https://www.emergobyul.com/resources/europe/where-ce-mark-is-required

 As I mentioned, the China Food and Drug Administration (CFDA) is the administrative body responsible for the regulation of medical devices and pharmaceuticals on the Chinese mainland. The US FDA resident posts in China assisted them in setting up their CFDA.  I personally know that many Chinese companies that manufacture finished devices and components for sale in the EU are certified to ISO 13485 and the CE-Marking of those products.  And those that do so for US destined products are also registered with the US FDA and subject to US FDA periodic inspections.

Writing good design inputs / requirements [for devices, processes, equipment (manufacturing, test)]?
  • List requirements needed / the reason why acquiring the equipment, RM, components in the "requirements"; expectations for it, (see below "functional, performance, interface" also),  Marketing, company manufacturing / engineering inputs; documentation and calibratable instrumentation requirements; spare parts; etc;
  • Also address the following as applicable (from one of my webinar slides): 
           "Virtually every product will have requirements of the following three types:
1.Functional requirements:  what the device does, focusing on the operational capabilities of the device and processing of inputs and the resultant outputs;
2.Performance requirements:  how much or how well the device must perform:  e.g., speed, strength, response times, accuracy, limits of operation, etc. This includes a quantitative characterization of the use environment, including, temperature, humidity, shock, vibration, and electromagnetic compatibility.  Device reliability and safety (see ISO 14971, Device Risk Management)  fit into this category;
3.Interface requirements:   characteristics of the device critical to compatibility with external systems;  characteristics mandated by external systems and outside the control of the developers.  One such important interface is the user and/or patient interface (see Use / Human Factors Engineering IEC 62366-1)."
  • Add required ISO + standards;
  • Add applicable US FDA guidance document requirements or their equivalent;
  • Convert to quantifiable / measurable terms where necessary;
Remember the requirements become the basis for V&V -- Instalation (IQ) and Operational  (OQ, all the remainining requirements) / Performance (PQs;  those requirements that need additional challenging due to allowable "worst case" inputs, e.g., different operator shifts, large volume RM, power fluctuations, other highly variable inputs.

As I mentioned, DIs and DO's during the R&D process is an iterative process, recognized as such by the FDA in it's Design Control Guidance Document.  A DI becomes a DO which in turn becomes a DI at the next level of development, which becomes a DO, until the product reaches the final stage of development.  Generally I recommend that the interim "DO's" be retained under Design Input in the DHF, and only the final DO that outputs to Design Transfer to Manufacturing, be retained under Design Output.  But the company would develop a system that works best for their culture / system, and then specify this in their Design Control SOP and then follow their SOP.