# Risk Acceptability Criteria - Probability and Acceptability Level

#### alimary15

##### Involved In Discussions
Good afternoon,

My problem/doubt is the following:

We have a medical device ( I will call it System) which is composed by different sub-systems, which are also medical devices.
Each sub-system can be seen as an indipendent product (Software in most cases) which comes with its own risk management plan and risk acceptability matrix and criteria.

The System is looking at the integration of these sub-systems and also for the System a risk matrix, with related risk acceptability criteria, is defined.

My doubt is the following: how to handle at system level different acceptability criteria that might come from different sub-units?

My guts tell me that we always have to consider the most critical acceptability level but I am not sure if this is right.

I believe that the scale of probability and severity should not change , because the probability is always a number from zero to one and should have the same meaning for all the products. However, I wonder how different level of acceptability criteria for two risks that are rated with same P and S can be assessed.

Do you have any experience or suggestion in how such cases could be handled?

Thanks so much for your help!

#### Bev D

##### Heretical Statistician
Super Moderator
Good afternoon,

My problem/doubt is the following:

We have a medical device ( I will call it System) which is composed by different sub-systems, which are also medical devices.
Each sub-system can be seen as an indipendent product (Software in most cases) which comes with its own risk management plan and risk acceptability matrix and criteria.

The System is looking at the integration of these sub-systems and also for the System a risk matrix, with related risk acceptability criteria, is defined.

My doubt is the following: how to handle at system level different acceptability criteria that might come from different sub-units?

hmmm. I always provide an assessment of any failure mode/risk for components at both the component AND system level. could you give us an example of what you are concerned with?

My guts tell me that we always have to consider the most critical acceptability level but I am not sure if this is right.
well yes you need to address the most critical somehow. but I think is in the context that you cannot simply address the lower risks while ignoring the higher risks...some failure modes will have several effects and if we only address the worst effect we are not assured that the slightly less (yet still serious) risks will be addressed. does that make sense?

I believe that the scale of probability and severity should not change , because the probability is always a number from zero to one and should have the same meaning for all the products. However, I wonder how different level of acceptability criteria for two risks that are rated with same P and S can be assessed.

hmm, in general items with the same probability and severity should have the same acceptability criteria. Perhaps if you give us an example of this dilemma it will help. I can see where two items having the same Probability and Severity would land on different sides of the acceptability criteria based on cost to mitigate vs cost of the total number of occurrences...(usually this happens when the probability is low and/or the severity is low...

Replies
5
Views
313
Replies
8
Views
203
Replies
6
Views
646
Replies
1
Views
423