Is it possible to avoid the OQ?

[Kid006]

We receive a new CNC machine from a different plant and since we are certified in iso13485 process validation should be performed,
the other plant is not in the medical device branch so a validation was not performed.
The IQ has been completed but specifically this machine only have a speed parameter which does not have a reference unit of measure it is presented as a percentage, also the parameter was locked by the supplier.
am i able to jump from the IQ to the PQ? (avoiding the OQ with a justification)
Our products are Foam Positioners, Medical dev 1
Any guidance regarding this issue will be very helpful
Thanks

 

[Tidge]

To try to prevent any potential confusion, I will point out that historically "OQ" has meant different things for device and pharmaceutical manufacturers. This thread appears to be (and my response is) relevant to device manufacturing, where OQ is the 'operational qualification' of a process 'at the limits of the process'. The OQ generates evidence that the process makes conforming parts no matter what the machine settings are set at, as long as those settings are within the allowed limits. This data can serve as feedback about the chosen limits for the process.

My opinion:

If the process validation never occurred, then an OQ is called for... even if there is nothing (obvious) that an operator can change.

If the replaced equipment is identical, and the IQ is robust enough (e.g. maintenance and calibration methods) to pin-down potential variables that could exist between the machines; I think it can be rationalized to 'skip' an OQ because you are no longer trying to analyze/establish process limits, provided that the PQ provides the necessary evidence that the operator/material variability still results in a process under control.

Personally, I would still do a 'sanity check OQ' and then rationalize using the data from an (abbreviated) OQ as part of the PQ. It's just my experience, but skipping an OQ tends to raise more eyebrows than a PQ with a slightly wonky study design. I also dislike finding issues in the first step after a 'skipped step'.

 

[Kid006]

The OQ generates evidence that the process makes conforming parts no matter what the machine settings are set at, as long as those settings are within the allowed limits. This data can serve as feedback about the chosen limits for the process.

I think it can be rationalized to 'skip' an OQ because you are no longer trying to analyze/establish process limits,

IQ is robust enough and the only setting in the machine is the cutting speed therefore the variability between doing parts at 1% or 100% (speed) are very minimal that our measuring instruments cannot detect it

skipping an OQ tends to raise more eyebrows than a PQ with a slightly wonky study design.

I was thinking in running the OQ with several combinations of speeds and the same program to desmotrate that the variability is minimum, but i dont know is that will be enough to conclude the OQ

 

[yodon]

I agree with @Tidge - especially the raising eyebrows part! First off, though, what do YOUR procedures say? Do your procedures allow OQ to be skipped?

IQ/OQ/PQ is just *an* approach; not THE approach. You just need to be prepared to defend that whatever results you gather demonstrate the equipment can consistently perform to your needs.

 

[Bev D]

I find it impossible to believe that speed is the only variable setting. Remember that it’s not just about the equipment but the entire process that utilizes the equipment. Are the parts clamped onto a fixture? Is there a torque to that clamping? What are you cutting - whaat type of equipment is this a lathe? A mill? These ar only a few of the questions we could ask...

 

[Tidge]

If the machine is a CNC, depending on the actual machine I wouldn't necessarily try to challenge fixture positioning or material holding in an OQ, but rather allow such things to be considered as part of operator/raw material variability in the PQ. If the PQ results indicate potential issues perhaps it would be necessary to implement specific control mechanisms (with limits) that could be challenged as part of an OQ. Even with limited variables in an OQ, I still recommend doing one, partially because within an OQ you can explicitly collect more data that you might otherwise not collect as part of a PQ (which sometimes can default to pass/fail acceptance criteria). For example, within an OQ you can study the variability of certain dimensions across an entire lot, whereas some typical machining operations can only include relatively simple things like examining the first/last/random-middle pieces.

Story time: I was once involved with the review of a manufacturing process with wide variability in the outputs... there was so much variability in the outputs that only one operator could "do it right". It took entirely too long to convince management that in order to reduce the variability in the outputs that it was necessary to enforce tighter controls in the manufacturing process (based on the 'good' employee's choices), and then challenge those limits to guarantee better output.

 

[somashekar]

We receive a new CNC machine from a different plant and since we are certified in iso13485 process validation should be performed,
the other plant is not in the medical device branch so a validation was not performed.
The IQ has been completed but specifically this machine only have a speed parameter which does not have a reference unit of measure it is presented as a percentage, also the parameter was locked by the supplier.
am i able to jump from the IQ to the PQ? (avoiding the OQ with a justification)
Our products are Foam Positioners, Medical dev 1
Any guidance regarding this issue will be very helpful
Thanks

I have two questions to ask you before going about further...
1. How was your part made earlier.
2. Is the new CNC locked for a specific purpose making it a SPM

 

[Kid006]

The machine is basically a spinning blade guided by a program to cut foam in certain shapes

 

[somashekar]

The machine is basically a spinning blade guided by a program to cut foam in certain shapes

From all that what you say, you would not need to validate this process, if you can verify your output part by a valid means of measurement or use the part if it passes a valid gauge that you have specifically made for this and is duly verified.