Standards as Design Inputs

[laslo3001]

Hi,

Is it OK to have the following Design Input 'Device must be compliant with ISO 10993' rather than writing out each part of the standard it must comply with?

Thank you

 

[Sidney Vianna]

If there is a specific revision of a standard, it should be noted; otherwise it will imply always the latest version of the document. Also, as written, the expectation is for full (100%) conformity without any exceptions.

 

[Bev D]

Also be aware that many design engineers will have ‘creative’ interpretations of any standard…add that goes double for validation of testing of any requirements…”I designed it to be compliant so we don’t need to validate that”.

 

[Ed Panek]

I would reference standards instead of directly listing the standards as required. Standards are amended or revised in time. As that happens they integrate into industry and regulations. You don’t want to be held to a changing standard that may have no relevance to your product in terms of what has changed since last edition.

 

[yodon]

Recognize that 10993 is more of a family - I think there's like 20+ standards in the family.

You may want to abstract your design input to be "biocompatible" and then have the test lab determine the relevant standards.

 

[laslo3001]

Thanks for your help everyone. That particular standard was just an example. But trying to look at a way, when designing to a standard, to have a simpler way to write a design input rather than pulling out every requirement. Then having a checklist for verification to show how we comply with the standard. Are there any more drawbacks to this?

 

[Tidge]

I'm not a fan of listing standards as design inputs. My perspective is colored by the questions around: What is listed as verification? (of implementation or effectiveness). In general, I think the burden of demonstrating a modified design output still meets some sort of certification to a particular standard gets more complicated (than it needs to be).

I'm more in favor of listing such standards as needs, and then have them filter into requirements that are more specific and directly testable.

 

[laslo3001]

Thanks everyone for your help. I understand that it may not be the best way to do it, but is it OK to do it this way? The reason being, pulling every requirement out of the standards we are following, we are ending up with a huge list of inputs, which are difficult to review and keep track of. Keeping the design input simple and then having checklists in place to verify we are compliant would be helpful to us, but just want to know that it's ok to do this as I'm not used to designing to standards.

Thank you

 

[Ed Panek]

Design inputs should be complete. Not simple. The reason you test to a standard has to do with the requirement and not the standard. The standard just describes agreed upon testing methods and performance criteria.

 

[QuinnM]

In our design documentation traceability for each specification starts at the design inputs document. Each requirement is assigned a number, and this is carried through the design documentation through verification. We do not reference standards. If we referenced a standard, then traceability would be difficult. For us, we could reference a standard, but it is easier to list the individual specifications.