Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.

Risk Analysis Updates due to complaints

#1
Currently I am having issue trying to get everyone (Quality, Regulatory, Engineering) on the same page regarding the risk analysis. We basically use a FMEA approach for our risk analysis then list mitigation, risk acceptability, risk benefit analysis, you get the point.

We have it set up so the hazard is identified, then the sequence of events (Failure mode),, the hazardous situation and then we list the harm. We will break it out if multiple harms are possible.

Currently, when quality receives a complaint in which there is no patient harm, they wants us to update the risk file for a new line item in which there was a device non conformance but it was noticed prior to use and no harm occurred. To me, adding in line items in the risk analysis in which the sequence of events did not lead to the hazardous situation and thus no harm occurred does not make sense to me. They insist they need a specific bucket to place the complaint into.

For example:
if you are operating and a device were to fracture and thus fall into the open patient, creating an unintended laceration of maximum severity - Minor.

Vs

You receive a device and it is already broken. It is noticed and so the device is discarded, and a complaint is filed.

Do we need to include product non conformancies that occur prior to the involvement of the user in the risk analysis?

When answering the complaint I would think the correct thing to do would be to evaluate what would happen if the device failure were to recur and if it would be likely to cause or contribute to a death or serious injury for a recurrence, then you would report it to the FDA.

I would not create an additional line item in the risk analysis of negligible severity for the user to notice the defect prior to use and discard it. Tis seems as though then when the complaint is being analyzed, they are not considering the maximum harm if the failure to recur. TO me this would actually just be a mitigation for the scenario, saying to inspect devices prior to use for any damage or wear.


I hope some of this makes sense!
 

try2makeit

Quite Involved in Discussions
#2
To me even with no harm done, the device broke and you should do a root cause of why and how to prevent it from happening and then add that to your FMEA.
 

yodon

Staff member
Super Moderator
#3
To me, failure to meet specification would not necessarily result in an update to the risk file, even if there's a high likelihood of recurrence. If, though, having to discard (and presumably get another) results in a delay of therapy that could result in harm, you should update (so there are plenty of things that could occur prior to involvement with the user that potentially impact the risk file. I think you have to consider each case on its own merits.

A complaint could be anything - even something as simple as the number of units received was not the number ordered. To me, if the investigation does not indicate that anything in the risk file is missing / inaccurate, there would be no need to touch it (and the conclusion nicely documented in the complaint investigation).
 
#4
To me even with no harm done, the device broke and you should do a root cause of why and how to prevent it from happening and then add that to your FMEA.
but I would think this would fall under device functionality severity and not patient safety severity. This particular document is all based on risk to the end user, not just business/ quality risk. If there's no resultant harm to me it wouldn't make sense to update the file.
 
#5
To me, failure to meet specification would not necessarily result in an update to the risk file, even if there's a high likelihood of recurrence. If, though, having to discard (and presumably get another) results in a delay of therapy that could result in harm, you should update (so there are plenty of things that could occur prior to involvement with the user that potentially impact the risk file. I think you have to consider each case on its own merits.

A complaint could be anything - even something as simple as the number of units received was not the number ordered. To me, if the investigation does not indicate that anything in the risk file is missing / inaccurate, there would be no need to touch it (and the conclusion nicely documented in the complaint investigation).
Thanks for the reply. What you are saying seems to make sense. Like Yodon said, everything is pretty much a case by case basis.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#6
I think it depends on what is already in the risk document.
it would trouble me that the device arrived broken in the box. is this an indication of how easy it is for the device to break? can it break during use? if so, how could that hurt the patient? does this increase your estimate of the probability of failure or the occurrence rate?

every complaint is additional data that can better inform your risk assessment.
 
#7
I think it depends on what is already in the risk document.
it would trouble me that the device arrived broken in the box. is this an indication of how easy it is for the device to break? can it break during use? if so, how could that hurt the patient? does this increase your estimate of the probability of failure or the occurrence rate?

every complaint is additional data that can better inform your risk assessment.
My example might not have been a good one. I guess I mean for any instance in which a device non conformance is noticed by the user prior to use and is then returned. Since there is no hazardous situation that occurs and therefore no resultant harm, how do you incorporate this complaint back into your risk analysis.
 

yodon

Staff member
Super Moderator
#8
You have to ask questions (and, of course, document the answers). Will this result in delay of therapy? What if the user didn't notice and tried to use it? [There may be others.] If none of the questions point to any risk or impact any of the estimates in the risk file, you don't, IMO, update the risk file. And bear in mind that risk is not only to the patient. You need to consider anyone along the line and the environment.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#9
Aside form the really good advice to look at every complaint and see fi it provides informative value to improve yoru risk assessmentAND the product performance, I am really struggling with this line and your question:
Currently, when quality receives a complaint in which there is no patient harm, they wants us to update the risk file for a new line item in which there was a device non conformance but it was noticed prior to use and no harm occurred. To me, adding in line items in the risk analysis in which the sequence of events did not lead to the hazardous situation and thus no harm occurred does not make sense to me. They insist they need a specific bucket to place the complaint into.
!

what is it that Quality is actually trying to do? it sounds like they are either just trying to record the complaint somewhere OR they trying to say "see it isn't so bad, nobody got hurt". What are they trying to bucket?
 
#10
You have to ask questions (and, of course, document the answers). Will this result in delay of therapy? What if the user didn't notice and tried to use it? [There may be others.] If none of the questions point to any risk or impact any of the estimates in the risk file, you don't, IMO, update the risk file. And bear in mind that risk is not only to the patient. You need to consider anyone along the line and the environment.
THis is also my understanding
 
Top Bottom