< 1 min reading time
source: https://www.linkedin.com/groups/78665/78665-6076464952049225729 Marked as spam
|
Meet your next client here. Join our medical devices group community.
Private answer
Karen Boyd, ASQ CQA
David - I believe the NB is asking for / requesting to verify that you (manufacturer) have a process and procedure in place to address a product recall - as preparation and readiness to act on adverse events is required in your QMS.
Marked as spam
|
|
Private answer
Marcelo Antunes
Any procedure should be "simulated" in some way before use to make sure it can perform as intended, if not, how can you be sure if they will work when you actually need them? For example, I've been having problems with clients meeting adverse event reporting deadlines because they created a procedure, but when I asked for them to go thru the procedure step-by-step in a simulated way, they noticed they could not respond in time (it took too for the information to reach some of the key people, and those could wait too long to answer, so in the end the deadlines were not met).
So, although it's not usually requested (which is actually weird, because there's a planning requirement in all management system standard for planning, which would include this type of procedure verification/validation), it's obviously correct that a notified body wants to make sure that the contingency procedure you created was verified that it will work when needed. Marked as spam
|
|
Private answer
Marcelo Antunes
Think about fire safety evacuation training, we don't wait for a fire to happen to verify that the procedure works, otherwise, if it doesn't, we would be fried.
Marked as spam
|
|
Private answer
Ee Bin Liew
This simulation would be a very, very time consuming and costly exercise. please ask the NB for a clear, objective reason why this is necessary before ever agreeing. if it is just a case of running thru who would be notified and take whatever action, then it is as good as not asking - much rather look thru a couple of past examples (if it is available, because if they're none, the NBs cannot make you 'fashion' one).
Cheers, Ee Bin Marked as spam
|
|
Private answer
Karen Boyd, ASQ CQA
It seems to me that "verification" of your product recall / adverse event reporting process and procedure describing / instructing that process is what is required. Is the process feasible and can all required actions be met as illustrated and/or described and intended in your procedure?
It would seem that simulation of the process is "validation" that it actually works and requirements are met as intended. "Validation" would then be a more costly exercise. Is "validation" of product recall / adverse event reporting a requirement? Marked as spam
|
|
Private answer
Leonel Graça
I agree with Marcelo. Plus, maybe the auditor did not requested and just checked you for consistency. I only Audit 9001 and we have to follow the letter of the standard, but sometimes we try to "shake" things for finding consistency.
Very basic 9001 which has less requirements then 13485, PDCA: you planned a procedure. You implemented a procedure. How about checking it? Yes, you have verified it, but how did you validated it? How did you Check for loopholes to Act and improve it? As Marcelo mentioned fire drills, I had that problem with a tornado on the fist months of my job. Civil Defence failed to regularly simulate a catastrophe before I went on-board. Emergency contacts outdated, regulatory changes, overlaping responsabilities, etc. First days were mayhem because it was neither simulated 10 years before when created, or in the years that followed. Marked as spam
|
|
Private answer
Leonel Graça
Ee Bin Liew, maybe that is a pessimist reading, looking at the glass half empty.
Training is expensive? How much does ignorance cost instead? Simulation is expensive? How much does failing the procedure cost? A real life situation should not be more expensive then most types of training. I have been there, when containment actions and emergency plans failed due to lack of revision and simulation. You do not want to be there, when people lives are in stake. Marked as spam
|
|
Private answer
Leonel Graça
And Karen Boyd had said it all already, in short. Spot on.
Marked as spam
|
|
Private answer
Karen Boyd, ASQ CQA
I understand what both Marcelo and Leonel are conveying. However, processes / procedures that are critical to the QMS should not only be reviewed and approved for accuracy, but suitability and effectiveness - much more frequently than 10 years or even 5 years - but as a component of planned Mgmt Review.
A question of suitability and effectiveness then comes down to whether "verification" or "validation" is required, and on what basis would "validation" be necessary? Marked as spam
|
|
Private answer
Ee Bin Liew
yes, Karen, I can see where your question on 'validation' of the procedure is leading to. There're requirements on ensuring that training of personnel is effective, just as there's requirements for ensuring all regulatory requirements are incorporated for field safety corrective actions, but there's no definitive requirement that the validation of a process (i.e. process works as intended) is asked for.
For the same training money, I would prefer to direct it towards making good quality product so that recalls can be minimised, and perhaps beef up the usual awareness training to a classroom style training for the recalls procedure. As for actual product recall situations, there are too many hands on experiences to cite, and some of the fatal ones leaves a personal mark till today. Marked as spam
|
|
Private answer
thankyou for your comments. We have the process and are testing elements like do traceability data extracts work. It is the interaction with customers and other 3rd Party contacts that is the difficult bit to simulate without prior notice. The hassle is the NB have already said that they would not expect to be involved as they already have too high a workload on real events. Not talked to CA yet. We can set up the contracts with couriers to be ready to use. But what about proof that we can pick up from nominated addresses as part of simulated FSN without actually ringing the door bell. The underlying question is how "real should the simulation be?"
I have seen reports that real problems come from when members of first division subject matter experts are not available and 1st or 2nd tier backups are trying to run the process. we are a small company so this is a distinct possibility. Discuss? Regards David Marked as spam
|
|
Private answer
Stuart Engelking
Mock Recalls, pretty normal stuff, I've seen this asked for, for years. You have a recall procedure, step 1. you then pick a product or lot(s) and see if it works. Can you generate a shipment history report? Can you create letters to these customers? Does your process involve reconciliation of returned products?
Marked as spam
|
|
Private answer
Henry Sibun
You need to demonstrate to an auditor (and more importantly to yourselves) that the process for recall could work. I ask for evidence at audits - most companies do mock recalls where they identify where all of the batch is - they do not actually bring the product back (this would be expensive) but trace it to the end customer.
Marked as spam
|
|
Private answer
Alan Barta
Obviously, you need to document every phase of a recall. First comes having a plan on file, then alert intention, then "bounding", which pinpoints defective batches and lots. This begins a clock with milestones to reach by set periods, once FDA is notified. Then isolating and quarantining batches in transit or still in inventory. Then notifying all parties, recovering deployed, arranging for destruction, then certifying destruction. If you haven't followed all these steps, your notified body might demand that you validate your process. Recalls are made necessary by vigilant postmarket surveillance. If you are not right on top of complaints daily, you could be doing irreparable harm to company.
Marked as spam
|
|