Analyzers really need a two point calibration to be truly trustworthy. Calibrating in air is great, but unless you can calibrate with another known gas (100% O2 being the easiest), you're really just sort of trusting that the cell is linear.
The best thing to do is get your own analyzer and do a two point calibration regularly so that you can be sure your analyzer is linear all the way up to pure gas (ex. 100% O2) or a known quantity in the case case of ppm-sensing cells. If you can get a hold of calibration gas, that is bump gas of a specific quantity of whatever gas you are testing for, you can easily test throughout the range as opposed to just the extremes.
Fortunately, 1 atmosphere O2 and CO sensors tend to last a while, so while you should calibrate regularly, it should be safe to assume that you should get decent life out of your cells. However, this is not always the case so it's really a "trust but verify" sort of thing.
Remember that an O2 sensor is a galvanic fuel cell, an electrochemical transducer which generates current that is, theoretically, proportional and linear to the partial pressure of the oxygen to which it is exposed. As such, they essentially have a limited lifespan based on the amount of oxygen the sensor "sees." As the anode is consumed or in the event of something like an electrolyte leak, the reading will either become current limited at higher partial pressures of oxygen, or simply be grossly incorrect. Because analyzer cells typically see low percentages of oxygen, they tend to last a rather long time, compared to things like rebreather cells which will see significantly higher partial pressures of oxygen.
In that regard, you can see how analyzer cells should theoretically last for an extended period of time, but it is always best to verify with a known gas, preferably at high partial pressures or higher ppm's of said gas, in order to ensure that the sensor is operating correctly within the parameters in which you will be using it.
