21 CFR 820 QSR 820.25 Statistical Techniques Procedures Requirements

[S Roche Hendrix]

21 CFR 820 21 CFR 820.250 Statistical Techniques Procedures Requirements

I have a question about procedures for Statistical Techniques to meet the requirements of 21 CFR 820 Quality System Regulation.

From an ISO standpoint the auditor will verify statistical techniques are in use but have never asked for a procedure. What type of procedure is FDA expecting? For example, if we create a pareto chart for our monthly management review, would they expect a detailed procedure on how we create the pareto chart or a general procedure that describes what techniques we use and where?

Any advice would be appreciated.

Thanks,
S

 

[Al Rosen]

21CFR820.250 does not require a detailed procedure on how you create a pareto chart. It doesn't require a general procedure just for statistical techniques. You can address your statistical techniques within many procedures. For example, your recieving inspection procedure can address sampling inspection within the procedure or you can have a procedure for sampling inspection that is referenced from it and from other procedures.

The requirement doesn't have anything to do with a chart you use for a presentation in your managment review or anywhere else but in the techniques used for product acceptance and/or process capability and control.

 

[Bwana.Jo]

Hi, Folks

I just dropped for seeking 'Statistical Issue' which can cover QSR 820.250 and also ISO standard, especially [Sec.8 Measurement, Analysis and Improvement].

But I failed to find out those application which I expected.

My study and analysis of ISO 13485;2003 and FDA QSR, they are basically agree with fundamental source and essence of direction, that is, "Process, Process Approach, Improvement of Process, Effectiveness of Process, Efficiency". In addition, we have to think and prepare PDCA cycle for every processes and phases. As you know, we could not skip for measuring or assessment that process outputs and monitor that process.

With this respect, QSR 820.250 & ISO Sec.8 is very important but generally it was paasing over or look over with difficulty to define.

Now, I'm prepare to establish that SOP with my own intention and seek all your opinion (as a expert) on that before finish my work. (70% done, but written with Korean ^^; ). I'd like to introduce main concept and basic theory for that SOP to you.

I prefer to utilize Statistical Approach. What do you think about this? Statistical Approach derives an output from analysis and that output must be desirable output to implement and improve current bad situation or phenomenon. How? I would like to say, Descriptive / Inductive / Expectative Statistical Approach can be explained that question. And also this kind of question would be poured when you conduct FDA inspection, especially during PAPC or 'Process Validation'.

That's all today.. If there exist some reply, then I'll unpack my story. Thanks ! Alex

I like it the pop up expression after complete contribution before.
"You never Know. Your inputs can be other's solutions."

 

[rose24m03]

statistical techniques

you may find ISO/TR10017 guidance on statistical techniques helpful

 

[Steve Prevette]

You are welcome to make use of the procedures I set up to document to the US Government what Fluor does for its statistical analysis at Hanford.

https://www.hanford.gov/safety/vpp/trend.htm

There is a page about Pareto charts and another for Histograms.

 

[Steve Prevette]

Current location for the materials referenced above is (broken link removed)