Validation of Software used in Verification Testing


Hello all,

We utilize a few software platforms to gather/summarize/report data during verification testing. For example, we'll have a calibrated piece of equipment (e.g., a temperature sensor) and we use this software to log a data point every second. The actual program that sets the interval, determines where the data gets logged, etc. is custom written and tailored to each individual test. We'll do some quick & dirty checks to ensure the results are as expected, but since the equipment is calibrated, we generally trust what it outputs. Once the test is over, the program spits out a text file with all the data and timestamps for each point.

Does this software need to be validated? It is only used for verification tests; not used in any type of production or for our QMS, which is where our current SOPs draw the line. And if the answer is yes, and you had to validate this type of software - would you perform some type of validation for each program written, in addition to the platform itself?

Thank you in advance!


Trusted Information Resource
Interesting question...
ISO 13485 requires the validation of computer software used in quality management system, production and service provision, and
monitoring and measurement of requirements (sections 4.1.6, 7.5.6, 7.6).
There is no specific requirement for computer software validation of software tools used in during V&V activities. It's a gray area.
Bottom line, your call. And if you chose to do it, I recommend to explore the option of including it as part of the platform itself. This way it would be covered , this way it be covered in masked in one report.



Trusted Information Resource
I am of the opinion that minimally the software as described in the OP needs to have a validation determination documented, and if it is determined that it needs to be validated that you follow your established process for NPS.

It doesn't strike me as being a very different situation than saying "we have this DMM that we use for quick and dirty measurements during design verification, so I don't think it needs to be calibrated." If such a piece of EQ didn't have a calibration determination, I would suspect that something is amiss.
Top Bottom