Vendor Column: Why should I perform calculations within my chromatography data system?

Skip to Navigation

Ezine

  • Published: Oct 22, 2016
  • Author: Chris Stumpf
  • Channels: Laboratory Informatics / Chemometrics & Informatics
thumbnail image: Vendor Column: Why should I perform calculations within my chromatography data system?

Another topic that comes up in laboratory informatics workshops that I lead is: “Where is the best place to perform chromatography based calculations?”

I don’t know for sure the percentage of laboratories performing chromatographic calculations in a spreadsheet, but my sense is there are a lot of you out there. It is still quite common for laboratories today to rely on the chromatographic system software to calculate the chromatographic peak area for components and then export the peak areas and concentrations to a spreadsheet where a calibration response factor is calculated with subsequent sample quantities determined (as in an assay).

Some other types of calculations that are often performed in spreadsheets include:

  • System suitability tests and limit testing
  • Incorporation of sample weight and dilution factors, possibly including multiple dilution steps
  • Application of other correction factors (purity, base/acid ratio, etc.)
  • Calculation of percent label claim
  • Calculation of impurities as a percentage of main compound

In a research environment, performing calculations in a spreadsheet, outside a chromatography software package, doesn’t pose a problem. The situation is quite different, however, for a laboratory operating under GxP. In this case, performing calculations in a spreadsheet raises serious data authenticity concerns, and FDA scrutiny, because a standalone spreadsheet almost never contains the 21 CFR Part 11 technical controls to protect the data.

So why is data authenticity of spreadsheet calculations such a problem in laboratories operating under GxP? Here I summarize some of the typical statements made in FDA warning letters and industry presentations that flag spreadsheets as an issue:

  • Computer System Validation is required, as it would be for other software used in GxP laboratories for calculations
  • Spreadsheets must be securely saved for review and submission
  • Spreadsheet files are e-Records and must comply with 21 CFR Part 11
  • Spreadsheets require security and access control which means a “Log on”
  • Microsoft Excel “audit trails” are not sufficient for compliance with 21 CFR Part 11

If spreadsheets are such a concern with regulatory agencies like the FDA, then why are scientists so loathe to part with them? Mainly because spreadsheets have a number of advantages:

  • Spreadsheets are ubiquitous and most people know how to use them
  • Templates can be created and used for all chromatographic systems that a laboratory may use
  • Easy to automate result generation that a CDS can’t handle (bracketing, blank subtraction, etc.)
  • Flexibility of the calculations within the spreadsheet
  • Formulas created for spreadsheets are faster than manual calculations
  • The calculation computations themselves can be validated

But, spreadsheets also have some disadvantages, such as:

  • Manual data transferal from a CDS to a spreadsheet is exceptionally time-consuming
  • They raise the potential for transcription errors
  • There aren’t any direct links from the spreadsheet to the e-records and meta-data (methods, raw data, reports)
  • They make it virtually impossible to comply with 21 CFR Part 11
  • Spreadsheet software is difficult to validate and secure because it’s so flexible
  • Inhibiting built in Excel functionality (like Copy and Paste to a non-secure spreadsheet) is difficult and is an expert skill few people have

So what should you do if you’re currently using spreadsheets in a GxP laboratory environment?

In my opinion, the best course of action is to try to incorporate your calculations in software that provides 21 CFR Part 11 technical controls such as often found in CDS, LIMS, or ELN. The best of these products come with standard built-in features including Part 11 technical controls that provide you with a unique (to every user) login and passwords, audit trail, electronic signatures, and so forth.

A suitably equipped CDS, LIMS or ELN can help in other ways. The 21 CFR Part 11 regulation also requires you to perform a successful Computer System Validation (CSV) to demonstrate "fit for purpose" or "fit for intended use". Validating a Commercial-Off-the-Shelf (COTS) software from a quality supplier which you have already assessed can be much more straightforward than validating a "made from scratch" Microsoft Excel spreadsheet. As you need to validate the CDS, ELN or LIMS to perform functions in your laboratory anyway, extending that validation to include the use of specific calibration, qualititation and correction calculations seems a simpler option.

Historically computer software validation (CSV) followed the software validation “V-model” that could contain many steps and be quite intimidating for bespoke (or GAMP category 5) systems. However, a simplified V-model can now be utilized for COTS systems where the specification and verification steps become combined into single boxes; following GAMP 5 along with the recently published Good Practice Guides (GPG) for Computerized Laboratory Systems from GAMP1. The extent of documentation required for each box in the simplified V-model then depends on a risk-based approach to the quality of the supplier's development processes, determined through a vendor assessment.

Figure 1. Software validation V-model. The left-side of the V-model describes the software’s requirement documentation and the right-side describes the verification step which consists of qualification testing and documentation.


Verifying user requirements is often called qualification, which is a component of the overall system validation lifecycle, ensuring that the computerized system performs correctly now, as well as after any changes, updates, upgrades or modifications - right through to retirement of the application. Qualification testing of the software is the responsibility of the organization using the software because they are ultimately responsible for the validation; however, to make things easier on the end-user the software manufacturer can provide qualification tools and offer services to assist with the various stages of software qualification.

Based on FDA warning letters I’ve read, I think it’s clear that an unsecured spreadsheet is not the best approach for a GxP laboratory. Couple the industry’s scrutiny with workflow inefficiencies such as extra steps for manually transferring data out of the CDS and the potential for transcription errors, and there is a very compelling argument for performing calculations as close to the original source of data as possible – in the case of chromatography this is within the CDS.

The key point to remember is that if you choose to perform calculations within the CDS, the CDS solution should at a minimum have the following attributes:

  • Ability to perform requisite chromatography related calculations
  • Possess 21 CFR Part 11 / Annex 11 technical controls
  • Provide a mechanism or service for some level of software qualification
  • Come with support, training and validation services to supplement your CSV activities and ease that process

If the CDS meets these requirements, then performing calculations at the CDS should deliver some of the following advantages:

  • Easier adherence to 21 CFR Part 11 / EMU Annex 11 compliance
  • Improved workflow efficiency (less manual data transfer)
  • Reduced errors (reduces manual transcription)
  • Easier validation of chromatographic calculations
  • CDS with a relational database provides data traceability and easier report generation because data is easier to find
  • Streamlined data review because all chromatography data is managed in one place
  • More efficient workflow approval through data traceability and the use of electronic signatures

1. GAMP Good Practice Guide, Risk Based Validation of Laboratory Computerized Systems, second edition, International Society of Pharmaceutical Engineering (ISPE), Tampa FL, 2012


The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.

Social Links

Share This Links

Bookmark and Share

Microsites

Suppliers Selection
Societies Selection

Banner Ad

Click here to see
all job opportunities

Most Viewed

Copyright Information

Interested in separation science? Visit our sister site separationsNOW.com

Copyright © 2017 John Wiley & Sons, Inc. All Rights Reserved