If you did not buy the spectrometer new and created some reference data with DIY stable color patches you are facing expensive choices anyway. I did the first and not with the expectation that I could avoid a manufacturer's recalibration at some point in time but more to find out when my spectrometer is going bad and has to be recalibrated. A string of opaque acrylic 3mm thick sheet samples measured at day one and the data + samples archived in a black envelope stored cool.
Your white calibration sample that comes with the spectrometer can be aged and you could in theory get another one or create another one if you had a precise spectral measurement of a new one. And no that sample does not have to have the ideal neutral high reflectivity of say a Barite or PTFE tablet, the spectrometer manufacturer may have another opinion on what the better patch white is, per spectrometer model. I have done some cross measurements on two manufacturer's choices and they were different. For saturated colors you can buy (very) expensive ceramic tablets with (very) precise (radio)spectrometer calibration data and it is certain your spectrometer will measure other values. How much and in what direction for several hues and then to know whether that is just the off value that is tolerated between the same instruments or worse and then find a method to compensate that harmonically outside the meter as you will not have access to what a manufacturer does in recalibration (I guess switch parts, lightsources, white patch etc right away and then calibrate as new on very expensive ceramic tiles). Both alternative and manufacturer's methods are not cheap.
There is something you could do and that is checking whether any paper lint, dust etc is in the optical system of the spectrometer. As far as it can be easily taken apart, do not go beyond that stage. I do it sometimes with my Eye1 Pro.
If there is someone with an Isis with a fresh white calibration patch do a cross check with both meters. Check the spectral distribution and the Lab values. I guess it should not be more off than one or two DeltaE.
Met vriendelijke groet, Ernsthttp://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2012, 500+ inkjet media white spectral plots.