The Elsmar Cove Wiki More Free Files The Elsmar Cove Forums Discussion Thread Index Post Attachments Listing Failure Modes Services and Solutions to Problems Elsmar cove Forums Main Page Elsmar Cove Home Page
Google
  Web Elsmar.com
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google.

View Full Version : Validation of an excel spreadsheet used for recording test results?


achorste
10th August 2009, 11:20 AM
Good afternoon all,

I was wondering if anyone would be able to help - we recently had an audit from one of our customers. One non-conformity they raised was we were using an unvalidated spreadsheet.

This spreadsheet is used for recording test results and report numbers, no calculations are involved, no macros are used, the only restrictions are based upon the user access to the folder in which the spreadsheet is kept.

What strategy would you use for "validating" the use of such a spreadsheet?

I was going down the lines of developing a combined User Requirements Spec / Functional Spec with the spreadsheet structure etc, but what sort of functional testing can I do with a simple spreadhseet like this?

Yours confusedly,

Achorste

Jimmy the Brit
10th August 2009, 11:44 AM
Coincidentally I received this link to an on-line (http://www.complianceonline.com/ecommerce/control/trainingFocus?product_id=701327&channel=mailer=Alfonso_AUG-11_AUG-10_EM) seminar concerning this very subject this morning and I was thinking of attending to find out the same thing - How do you validate a simple non-calculating spreadsheet? I have no idea if the course is any good, but I have attended a couple of their sessions before and found them to be pretty good value.

If you don't get the answer you need from the Cove maybe it is worth the £249 fee, although the session is tomorrow so probably a bit late! NB I am in no way affiliated with Compliance Online.

Regards,

Jimmy

yodon
10th August 2009, 01:12 PM
I think about all you can do is show that the user restrictions are properly implemented. You should probably also show how the data is immutable, once entered (or if you can't go that far, show rationale for whatever protections that are in place are sufficient).

Sturmkind
10th August 2009, 01:32 PM
Hi! Was the non-conformance concern about the integrity of the data entered? If so, the only way to increase the chance of catching an error such as transposition (entering numbers or data 'backwards') is to deploy multiple visual inspections and/or to institute a reasonable High and Low limit for the cell values. :cool:

Ajit Basrur
10th August 2009, 11:35 PM
Have a look HERE (http://www.spreadsheetvalidation.com/companyinfo/articlesindex.htm)

bio_subbu
11th August 2009, 07:59 AM
Complianceonline is going to conduct a live online training on “Validation and use of Excel spreadsheets in FDA regulated environments”. The training date is scheduled on August 11, Tuesday at 10:00 AM PDT. For more information refer here (http://www.complianceonline.com/ecommerce/control/trainingFocus?product_id=701327&channel=mailer=Alfonso_AUG-11_AUG-11_DM)

jscholen
11th August 2009, 09:02 AM
I had an ex-FDA investigator tell me she didn't care about the use of a spreadsheet if it didn't perform any calculation, so in your case, I would concern my validation around access and audit trails if the data is a part of your quality records.

sorin
11th August 2009, 10:03 AM
http://www.ofzenandcomputing.com/zanswers/609

raghu_1968
12th August 2009, 05:37 AM
Hi Achorste,

If the spread sheet is not involved in any calculation, then it can be considered as a document. And if it is involved in the any official trade, the security access is to be validated. We can test the following;

1) Configuration of the base system,
2) Data source
3) Security of the data and the user
4) Application software specification
5) Functionality of the application software

By testing these parameters, we can declare that the data and functionality has been validated.

Document Methodology:
1) URS preparation
2) Validation Plan
3) Validation test scripts
4) Reporting
5) Testing the URS compliance

Regards / Raghu

achorste
12th August 2009, 09:15 AM
Thank you all for your help, I'm putting a generic protocol for spreadsheet validation together now - I will try to remember to post it up if anyone's interested :D

Thanks again.

Ajit Basrur
12th August 2009, 01:46 PM
Thank you all for your help, I'm putting a generic protocol for spreadsheet validation together now - I will try to remember to post it up if anyone's interested :D

Thanks again.

We are interested :)

zancky
12th August 2009, 02:28 PM
I have only some experience with excel and one of the problem I have seen about non calculating sheet is the input format
if You insert special characher like ^ @ or something that looks like a date and press enter the cell is formatted. You can go back and edit the results but it may happen You have not corrected the format. The worst case is: if You enter @+something, then You You go back to the cell and overwrite a number, You will see the number but for excel is a string!
I have seen an excel spreadsheet at one of my customer (pharmaceutical) and it was protected against any special characheter, invalid date format (only DD/MM/AAAA was allowed), invalid entry (if is required a number only a number can be written) etc.

gar4guv
12th August 2009, 03:22 PM
This spreadsheet is used for recording test results and report numbers, no calculations are involved, no macros are used, the only restrictions are based upon the user access to the folder in which the spreadsheet is kept.

What strategy would you use for "validating" the use of such a spreadsheet?


one way we get around validation of non-calculating spreadsheets (or word docs, or pdfs for that matter) is to print out the file and sign it. then we can argue that 100% verification is being performed.

zancky
17th August 2009, 12:06 PM
I have remembered that someone once addressed me to http://www.npl.co.uk (http://www.npl.co.uk/)
Under http://resource.npl.co.uk/docs/science_technology/scientific_computing/ssfm/documents/ssfmbpg7.pdf You can find some useful informations

achorste
17th August 2009, 12:28 PM
I have remembered that someone once addressed me to http://www.npl.co.uk (http://www.npl.co.uk/)
Under http://resource.npl.co.uk/docs/science_technology/scientific_computing/ssfm/documents/ssfmbpg7.pdf You can find some useful informations

Nice find - that's going straight in my reference docs folder!

Am I the only sad one here who has a folder on his PC crammed with virtually every guidance doc / open standard / sample doc etc he comes across? It's all useful at some point :D