1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
Dismiss Notice
You must be a registered member in order to post messages and view/download attached files in this forum.
Click here to register.

Need to Validate "off the shelf" Statistical Software Packages ?

Discussion in 'Qualification & Validation (Also 21 CFR Part 11)' started by TheWaeller, Feb 17, 2016.

  1. TheWaeller

    TheWaeller Member

    Joined:
    Jan 13, 2016
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    1
    Hi all,

    a short question:

    Is there a documented need to validate "of the shelf" statistical software packages like Minitab or JMP?
    Is it thinkable or sufficient for - lets say FDA audits - to rely on / to cite the huge numbers of succesful users of these packages? Similar to published biocomp studies in the net?

    Thanks for any hint
    The Waeller
     
    TheWaeller, Feb 17, 2016
    #1
  2. Miner

    Miner Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    577
    Likes Received:
    493
    Trophy Points:
    62
    Location:
    Greater Milwaukee USA
    Miner, Feb 17, 2016
    #2
    QAengineer13 and TheWaeller like this.
  3. TheWaeller

    TheWaeller Member

    Joined:
    Jan 13, 2016
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    1
    TheWaeller, Feb 17, 2016
    #3
  4. ncwalker

    ncwalker Well-Known Member

    Joined:
    Sep 21, 2015
    Messages:
    261
    Likes Received:
    168
    Trophy Points:
    42
    Location:
    North Carolina
    My $.02 - ALWAYS verify.

    Minitab and JMP are indeed "off the shelf." But they are not prevalent. They are prevalent with US, the quality community, but that's still not a lot of seats out there relative to say, Google Chrome or Microsoft Excel. Which means because it is specified and not prevalent, there are bugs. Some are idiosyncratic "features" that are annoying. Others are straight up wrong math in the background.

    We had an "off the shelf" gage management package that flat out improperly calculated Gage R&R results.

    Always take a known (even a simple) data set, run it through the package, and convince yourself it works. The makers of software definitely do not understand the concepts such as "validation" and "early containment" like we, the makers of actual parts, do. :)
     
    ncwalker, Feb 17, 2016
    #4
  5. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    606
    Likes Received:
    664
    Trophy Points:
    92
    Location:
    Maine
    It depends on what your software is doing and where you are in the FDA hierarchy.
    In general the FDA will take a dim view of any 'software' that is not SAS or R. basically because these 2 packages have logging the mathematical validity of any program can be 'easily' audited and verified. GUI based packages such as Minitab and JMP do not have logging. there math is not easily 'seen'. EXCEL is even easier to see the math but for most manufacturers, EXCEL spreadsheets that perform calculations that control quality and compliance must be validated, so JMP and Minitab are in an even worse position than EXCEL. The number of users is irrelevant. The closer you are to have direct human contact that effects safety and wellness the more you will have to validate your software. Of course casual engineering of JMP or Minitab don't have to be validated as long as the final work is in SAS or R. Just my experience
     
    Bev D, Feb 17, 2016
    #5