Reduce occurrence rating based on the PMS data and customer complaint data

[mithilesh]

HI

I am conducting PMS for class IIB DEVICE per EUMDR requirement. Last year I had a change in occurrence for a defect in my risk file based on customer complaint data. This shifted the occurrence to a higher level which resulted in a medium level or risk resulting in a benefit risk assessment.

based on the PMS data and customer complaint data..our occurrence levels have actually gone down without us taking any corrective actions. The occurrence level calculation is based on a complaints per million rate. If I use the lower occurrence to do my risk evaluation, I get a low risk and my risk is now acceptable.
My question is: can I use this occurrence data to now reduce my risk to an acceptable level which was unacceptable an year ago?

thanks

 

[yodon]

Let me be sure I understand.

You did an initial risk analysis and assigned what you thought would be a reasonable likelihood of occurrence to a particular risk. You got some complaints that led you to believe you needed to increase the likelihood of occurrence but you conducted a benefit-risk analysis and concluded no changes (no additional controls / no corrective actions) were necessary. Subsequently now, you are not seeing data to indicate that you should have raised the likelihood in the first place?

Without knowing more, it would be, well, risky, to advise that the score be lowered. It's not completely clear that you have enough data to justify. Have you done a thorough analysis of the failures? We don't have enough data to understand why you apparently saw a spike (causing you to raise it) and then it hasn't occurred (or is occurring at a much lower rate).

 

[Bill Hansen]

Is it possible that the original increase in Occurrence was due to a "special cause", or something that does not reflect an increase in failure rate? A few years ago, my company had a "spike" in occurrence based on complaints, but when we dug deeper we found that one customer was entering a complaint each time a problem occurred, on the same unit. There was one defect, but it was reported multiple times. In the end, though, only one actual failure. (It's important to recognize "device failure" and "opportunity for harm" can be slightly different...)

As @yodon said, more data may be necessary. Maybe telling the whole story will give you good justification to return the occurrence rating to the lower value.