Process Monitoring using SPC software

Hi,

I'm new here. Hello.
I'm employed in Quality Assurance and I'm trying to determine if verification is required for SPC software (QC-Calc) within a Medical Device company. The software is being used out of the box for process monitoring and trending of data for validated manufacturing processes. No-one is asking that this be validated in any way but I have an expectation that there should be some work done in relation to this, even just a comparison of source data (from process) to report data (output to SPC software). Reviewing ISO9001 and ISO13485 (see below) would also lead me to believe that there would be an expectation that any methods used for process monitoring should be verified as fit for use.

Does anyone have an opinion on whether verification should be completed for SPC software or experience where accuracy of SPC reports was questioned?

Thanks,
Sidhe

ISO9001:2015 Clause 9.1
9.1.1 - “the organization shall evaluate the performance and the effectiveness of the quality management system”
9.1.3 - “The organization shall analyse and evaluate appropriate data and information arising from monitoring and measurement”

ISO 13485:2016+A11:2021 - Medical devices. Quality management systems: Clause 8.2.5 states that and “the organization shall apply suitable methods for monitoring and, as appropriate, measurement of the quality management system processes.”

ISO 9001:2015 Clause 7.1.5.1 relates specifically to validity of data generated.
The organization shall determine and provide the resources needed to ensure valid and reliable results when monitoring or measuring is used to verify the conformity of products and services to requirements.
The organization shall ensure that the resources provided:
a) are suitable for the specific type of monitoring and measurement activities being undertaken;
b) are maintained to ensure their continuing fitness for their purpose.

The organization shall retain appropriate documented information as evidence of fitness for purpose of the monitoring and measurement resources.

Since it is easy to do the calculations by hand, or Excel spreadsheet, it would not be that big of a deal to get a few example sets of data and run them manually through Excel and through your software. Sometimes different softwares use different methods for estimating the standard deviation of the data for control limits, so it is not a bad exercise to at least determine what assumptions the software is using. If it was me (but I don't find doing a manual SPC to be that hard) I'd do it to at least see what the inner workings are of the automated tool. There are plenty of example data files out there on the internet supporting various textbooks. Even just create a set of data with a random number generator and see if both methods give the same result.

Sidhe_G said:

trying to determine if verification is required for SPC software

Click to expand...

I would say "yes." (But actually, it's "validation" which encompasses a bit more than verification.)

Steve: Although I agree in principle with your statement, using one software package to validate another software package is just a circular arguement. Either do it manually, or don't do it. I have never been questioned about the validation of 'mass market' software (I am thinking of Excel and Minitab specifically right now). The intent (to my understanding) is that you validate software which you have specifically written/modified to perform some type of task for you.

yodon said:

I would say "yes." (But actually, it's "validation" which encompasses a bit more than verification.)

Click to expand...

Would you be talking about validation of the software or validation of the specific usage?
If you were using standard functionality of COTS software, what would need to validated? I was suggesting verification of calculations or software output through confirmation that input meets expected output. Yes, I agree this would fall under "validation" activity.

In terms of validation of "mass market" software, I would usually suggest only validating specific usage and not the software itself. This is the approach I would take for Excel or Minitab. I'm not too familiar with QC-Calc so I'm not sure this would be equivalent to Excel or Minitab. Therefore, I'm trying to determine if additional validation outside of confirming calculation / output is required.

In my experience certain regulators (medical/pharma) tend to require validation of any software used to directly effect product quality or accept product for release. Things to validate:

• The software correctly implements the mathematics and intended functions
• Proper recording of results (data integrity)
• Security for changes.

You would need to look to your regulator’s or Customer’s requirements.
I have found that this type of validation is in your best interests regardless of requirements. I have continually been surprised by what programmers and engineers or scientists will try to do in software.

SPC software is typically not used as above. However, Miner is correct that you should validate that the software is using the correct data and calculations for the limits as some software allows you to use the total standard deviation or Sd of the mean to calculate the limits Instead of hte within subgroup SD. Some software can also be set to continually recalculate the limits. Som will also let you set the limits at 2 or 4 Sd instead of 3SD…

I also caution that chart type, rational subgrouping and sampling frequency are things to monitor, review and/or approve

Sidhe_G said:

only validating specific usage

Click to expand...

Yes, exactly.