Greetings all,

I have a problem.
Customer request: Measurement of uncertainty “as defined by the US Pharmacopeia”
Well, as I read it the USP is only dealing with the standard deviation. It does make reference to the type of weights used (in accordance with E617), however none of the other factors are used. I am aware this is something industry specific.

As I understand it, if I am going to issue a statement giving the MU of a particular device, I must include ALL of the variables.
The FDA in their enforcement of CGMP accept ANSI Z540 as an “acceptable standard.”
I can not find a document to prove it though.
How then do I prove 17025 is also an acceptable standard. I believe the phrase is “this standard is incorporated by USP by reference; other widely recognized standards may be acceptable” This refers to E617 only I think?

In my attempts to educate the customer (given my limited intellectual abilities) they seem to not understand what they are asking for. There also seems to be little or no understanding of 17025 or true uncertainty.

I have been through the USP, The CGMP part 7, CFR title 21 part 820, and sub 58.63.
I cannot find any DOCUMENTATION on what is and what is not Acceptable.

Can anyone help me prove any of this??
I am also in need of the current USP <41> and <1251>. I have not seen any revisions to the version I have regarding these two sections. I guess I need to prove that too.

Thanks…….

I've never heard of Measurement of uncertainty “as defined by the US Pharmacopeia”.

Has anyone else?

Hello Crenfrey!

When you speak of uncertainty, to which of the standards/specifications that you listed are you interested in? You have a lot of them listed there, and they may be slightly different.

I don't think USP 41 has been revised lately (although there is talk to have it revised.). If you are doing FDA work, I would suspect you would already have access to it.

No, I have not heard of MU as defined by USP but then again I don't have much experience with FDA quality standards nor do I know which specification they call upon for calculating MU.

From what I understand, you want to comply with ISO 17025 standards for MU but the client doesn't accept this and instead wants you to be in compliance with USP which is only std deviation? Why not just provide the std deviations then with a disclaimer that the MU provided does not comply with ISO 17025...such and such. Give the client what they want and ensure you've documented a deviation from your standard practice.

If your controlling quality spec does not explicitly call out a standard for calculating MU I would use best practice in your field, you're pretty well covered if you're using 17025 or GUM for your std practice.

I don't think USP 41 has been revised lately (although there is talk to have it revised.).

:confused: The 2007 edition is USP 30. Where did you get 41? :confused:

Since 2002 the USP has had a new revision every year.

Or are you talking about another USP?

Here is a link (http://www.fda.gov/cder/guidance/7260fnl.pdf) that might be useful. There is mention of ISO 17025 on the "useful reference materials" page (25).

Hope this helps some......:(

:confused: The 2007 edition is USP 30. Where did you get 41? :confused:

Since 2002 the USP has had a new revision every year.

Or are you talking about another USP?

Good call for clarification:yes:

What I was speaking of I believe is the USP 41 Section that addresses weights and balances. It's kind of a unique animal. Which is why I was interested what angle the OP was interested in. The last year or so they were going to revise that section, but I don't think they have.

Sorry for any confusion.

I am so sorry it took me so long to get back to this subject.

Yes it is USP section 41.

I have been all over the FDA web site (thanks for the link) and have volumes of documents from there.
I truly have done my due diligence regarding the requirements.

I can find no paper about what is acceptable calibration according to the FDA, 13485, UScGMP etc. Everything states they need a procedure. There is no reference as to what a correct procedure is. They are going by what is stated in the pharmacopeia which is not uncertainty at all.

The real kicker is this customer wants the standard deviation performed with 20mg.
The OEM of the balance requires 200g. This is a nightmare.
I cannot get the customer to understand the limitations of the device.
In my personal opinion, they need the correct tool for the job.
Not a device blown so far out the repeatability is totally in question.

( thanks for letting me vent)

The question remains about customer requirements.
I have given them what they call uncertainty reports.
I have also given them true uncertainty reports on a few of their balances.
I have explained about unreasonable demands on their equipment.
They are having no part in the issue to date.

I believe if they document the deficiencies in their methods, the proverbial can of worms will open with a whole host of problems rippling through their portion of the industry.

I do not wish to cause friction.

I do want to find more documentation to support the argument for them to adopt SOP’s that will better deliver what they truly need.

Yea, I know where you're coming from. That's why it's important to figure out if they are required to follow USP 41.

Basically, you are correct. They need to write their own procedure that covers range of calibration, corner load tests, precision test (if applicable), and most important the uncertainty calculations. There are multiple ways to approach the calculations. They just need to find a way that best fits their application and stick to it. Or, create a chart with four or five categories the balance is classified as (with appropriate tolerance).

Thanks Brad,
It is nice to know I am not going crazy.:o
Prying people out of “this is the way we have always done it” state of mind seems to be my biggest challenge. Many times customers do not fully understand the requirements they are being held to.
My frustration lies in the lack of documentation on acceptable practices.
At least TS 16949 says you need an accredited lab. Leaving the calibration requirements to those who know the equipment.