The way that oxygen analyzers work is by turning a small fuel source into energy. The more oxygen, the more energy, in a linear curve. But, over time the amount of fuel is less as it is being consumed, so while it is still making more energy as there is more oxygen, the total amount of energy it makes for the same concentration of oxygen will change.
But the sensors are designed to last for 1 year, and many will continue working for more than 1 year, so the amount of electricity generated for oxygen content should not change appreciably over the course of a few hours. And there is a risk of not flushing with ambient air sufficiently, and accidentally calibrating back to air, while the sensor has some nitrox floating around it, so I would not calibrate during the time that you have the sensor in your possession.
If I set an analyzer down, and was far enough away that I could not reasonably be certain no one had fiddled with it, then I would re-calibrate.
But, I also don't bother writing down the digits after the decimal when analyzing, I round... if I get a result further than 1-2% from expected, I would double check with a buddy's analyser or take the time to really recalibrate carefully, but as long as it is close, it's just confirming what the shops analyser on the fill bank says.
The most important is not to stop at oxygen, and follow up with CO. Those sensors are calibrated annually, and many people skip that calibration and just confirm it's working by breathing into it. That's more of a "go/no go" test, and it's designed to keep you safe from a fatal dose of CO, and the breathing check probably works fine. That is, a 1ppm result is not fatal, and the breathing check might not keep the unit calibrated to the point where it accurately detects to 1ppm. But, it should keep it accurate enough that if your gas has 20ppm, then it alerts as something close to 20ppm, say 19ppm, and that's good enough to let you know when you have a fatal amount of CO present.
