Jim,

     See my post above re: Apivar--What went wrong?  I sampled my hives and found three that were above my threshold to treat. After treatment I sampled again.  Two of the hives showed one or zero mites.  But one had gone from seven mites to 43 mites.  If I hadn't retested, that hive would have been overwhelmed with mites and died.  I wrote to a Master Beekeeper about this failure.  Here's part of her reply:

"Good for you for checking on your mite counts after a treatment - I recommend that all the time, but I think I get few takers as it probably sounds like busy work.  But your case proves my point: every now and then there is a treatment "failure", or at least, the treatment appears to have been ineffective.  And not knowing that fact can leave a beekeeper with a false sense of security."

Another way to think about checking for mites.  Frist I believe all hives have mites.  Also all hive treated for mites have mites.  So if all hive have mites pre and post treatment the reason to treat in to manage the mite load.  There is a lot of information an acceptable mite load.  Without testing you have no idea of what your hives mite load is.  If it is low don't treat.  Treatment cost money and has some risk to the bees especially the queens.  Post treatment levels are need to establish that you have an acceptable low mite load.